Currently OpenSSL implementation of TLS in hostapd loads only top
certificate in server certificate file. Change this to try to the
whole chain first and only if that fails, revert to old behavior.
Signed-off-by: Maciej Szmigiero <mhej@o2.pl>
This patch implements shared_freq handler for driver_nl80211.c. If a
"p2p_group_add" happens after legacy STA association, then this patch
will help to start the GO on the same frequency. This is useful when
supplicant is started on multiple interface running over a singly "PHY"
and the station interface is not used for the P2P device operations.
For example, wpa_supplicant -iwlan0 .. -N -iwlan1 ... and wlan0 is used
for station connection and wlan1 as the P2P device interface.
Signed-hostap: Jithu Jance <jithu@broadcom.com>
Without this, it is impossible to tell exactly which
MLME code returned the error.
Also, convert to wpa_dbg so that we get device names
in the messages.
Signed-hostap: Ben Greear <greearb@candelatech.com>
When GO negotation fails the WPS method is currently not cleared, which
can result in GO negotiation being resumed when a GO negotiation request
frame is received from the peer. That is unexpected as locally we
already gave up.
This manifests itself in getting
1319574733.955685: wlan0: P2P-GO-NEG-FAILURE status=-1
1319574733.955723: P2P: Removing pending group interface p2p-wlan0-0
...
1319574736.648378: wlan0: P2P: Starting GO Negotiation with previously
authorized peer
...
1319574736.650115: wlan0: P2P: Sending GO Negotiation Response
...
1319574736.988038: wlan0: P2P-GO-NEG-SUCCESS
1319574736.988233: P2P: No pending group interface
1319574736.988268: P2P: Create a new interface p2p-wlan0-1 for the group
Clear the WPS method to avoid this situation. I wasn't
able to test this though, but given the log I can only
assume this is how the situation happened.
Reported-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-hostap: Johannes Berg <johannes.berg@intel.com>
If P2P device expires while a GO Negotiation is in progress, currently
p2p->go_neg_peer is cleared without indicating GO Nego failure. This
will result in pending group interfaces to be left over. This patch will
indicate GO Negotiation failure and will remove any pending group
interfaces.
This patch addresses a corner case in GO-Negotiation case. Consider the
scenario where two devices A and B are in discovery stage and Device B
vanishes [moves out of range] when a connect is issued on the Device A.
Then Device A keeps on retrying the GO Negotiation Request till the
retry limit is reached. On reaching retry limit, the pending group
interface is removed. But suppose if the peer entry in the device list
expires before the retry limit is reached, then pending group interface
was not removed.
Signed-off-by: Jithu Jance <jithu@broadcom.com>
Invalid use of memcpy instead of memcmp in comparison resulted in the
GO interface address getting set incorrectly if the GO did not show up
in scan results anymore.
Signed-hostap: Jouni Malinen <j@w1.fi>
trace.h may end up including system header files that define offsetof,
so include the compatibility definition from list.h only after this.
Signed-hostap: Jouni Malinen <j@w1.fi>
HT_INFO_OPERATION_MODE_OP_MODE_MASK is used with variables in host
byte order, so it should not be claimed as le16.
Signed-hostap: Jouni Malinen <j@w1.fi>
This is a workaround for Windows 7 supplicant rejecting WPA msg 3/4
in case it used Secure=1 in msg 2/4. This can happen, e.g., when
rekeying PTK after EAPOL-Key Error Request (Michael MIC failure)
from the supplicant.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
This cleans up the code a bit by not having to deal with theoretical
possibility of maximum number of EAP methods to be different between
various components in hostapd.
Signed-hostap: Jouni Malinen <j@w1.fi>
Some static analyzers complain about memset with '0' value. This was
used correctly here, but since use of strdup is about as good an option,
use that to silence the invalid warnings.
Signed-hostap: Jouni Malinen <j@w1.fi>
This avoid incorrect errors from some static analyzers that do not like
memmove with pointers just after the end of a buffer even if the number
of bytes to move is zero.
Signed-hostap: Jouni Malinen <j@w1.fi>
All we really need is the family ID, and we can
get that with genl_ctrl_resolve() and then don't
need to worry about family/cache objects.
Signed-hostap: Johannes Berg <johannes.berg@intel.com>
If a P2P client associates with the group while it is
already associated, two member entries may be added to
the group which also confuses num_members counting.
Deal with this by removing the existing entry first
before adding a new one.
I think the way Reinette ran into this was due to our
tx_sync implementation in iwlagn, mac80211 might have
queued two association frames thinking the first one
just failed, but both only went out after the sync was
really successful (which tx_sync doesn't wait for).
Reported-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-hostap: Johannes Berg <johannes.berg@intel.com>
The P2P_FIND command was failing if it was issued at the moment when
a scan operation was in progress. Avoid returning failure in this
case by scheduling the P2P find to start once the ongoing scan is
completed.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
WEXT is not really changing anymore and more or less all Linux
distros come with linux/wireless.h that is recent enough to
allow the driver wrappers to be build.
Signed-hostap: Jouni Malinen <j@w1.fi>
Commit c2ff13c533 broke this with a
copy-paste typo that ended up adding the Probe Request P2P IE into the
Beacon frame (i.e., Beacon frame had two P2P IEs while Probe Response
had none).
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Avoid zero-length memset at the end of the buffer. This is not really
needed, but it makes the code a bit easier for static analyzers.
Signed-hostap: Jouni Malinen <j@w1.fi>
Avoid zero-length memset at the end of the buffer. This is not really
needed, but it makes the code a bit easier for static analyzers.
Signed-hostap: Jouni Malinen <j@w1.fi>
Explicitly validate data_len so that static analyzers do not get
confused about the padlen validation. This is not really needed, but it
makes the code a bit easier for static analyzers.
Signed-hostap: Jouni Malinen <j@w1.fi>
Explicitly validate seed_len to skip memset call with zero length
of copied data at the end of the buffer. This is not really needed,
but it makes the code a bit easier for static analyzers.
Signed-hostap: Jouni Malinen <j@w1.fi>
data->phase2_method cannot really be NULL if
eap_fast_init_phase2_method() returns success, but this construction
seems to be too difficult for some static analyzers. While this change
is not really needed in practice, it makes it easier to go through
warnings from such analyzers.
Signed-hostap: Jouni Malinen <j@w1.fi>
hapd->time_adv cannot really be NULL if hostapd_update_time_adv()
returns success, but this construction seems to be too difficult
for some static analyzers. While this change is not really needed
in practice, it makes it easier to go through warnings from such
analyzers.
Signed-hostap: Jouni Malinen <j@w1.fi>
md->curlen cannot indicate full buffer size here since the buffered
data is processed whenever the full block size of data is available.
Avoid invalid warnings from static analyzers on memcpy() outside the
buffer length by verifying that curlen is smaller than block size.
Signed-hostap: Jouni Malinen <j@w1.fi>
This function can fail in theory since the SHA-1 functions are
allowed to return an error. While this does not really happen in
practice (we would not get this far if SHA-1 does not work), it is
cleaner to include the error handling here to keep static analyzers
happier. [Bug 421]
Signed-hostap: Jouni Malinen <j@w1.fi>
This has been obsoleted by the more generic Linux WEXT (driver_wext.c)
support. The hostap and madwifi driver wrappers can now be used only
with hostapd. The old station interface remains available in releases up
to 1.x.
Signed-hostap: Jouni Malinen <j@w1.fi>
The driver wrappers broadcom, iphone, osx, and ralink have not been
maintained for a while and it does not look like they will be in the
future either. As such, remove them from the development branch. The
previous versions will be included in older releases up to 1.x.
Signed-hostap: Jouni Malinen <j@w1.fi>
Reassemble partial TLS records to make the internal TLS client
implementation more convenient for stream sockets.
Signed-hostap: Jouni Malinen <j@w1.fi>
The padding validation was done on the last padding-length octets in the
buffer which misses the first padding octet (the last octet is the
padding length). Fix the starting offset for the comparison loop to get
the first octet verified. [Bug 420]
Signed-hostap: Jouni Malinen <j@w1.fi>
Return number of user input bytes from tlsv1_record_receive() to
move this detail into the proper record layer processing. In addition,
ignore unknown content types at record layer and allow processing to
continue after warning level TLS alerts to provide minimal workaround
for closure alerts.
Signed-hostap: Jouni Malinen <j@w1.fi>
Instead of trying to remember to add wpa_msg() calls for every possible
path where a STA becomes authorized or unauthorized, use
ap_sta_set_authorized() to send these events more consistently.
Signed-hostap: Jouni Malinen <j@w1.fi>
wpa_msg() has to use hapd->msg_ctx instead of hapd as the context
pointer to work properly in wpa_supplicant AP mode.
Signed-hostap: Jouni Malinen <j@w1.fi>
Currently, there is no events over the control interface
when a AP disconnects a station due to inactivity. With
this patch, an "AP-STA-DISCONNECTED" event will be sent.
Signed-hostap: Nicolas Cavallari <nicolas.cavallari@lri.fr>
Use a driver_ndis.c specific initialization function to fill in the
wpa_driver_ops information to make it easier to modify struct
wpa_driver_ops in the future. Being able to build driver_ndis.c
with MSVC was the only reason for having to maintain the same order
of function pointers in struct wpa_driver_ops and for having to
update driver_ndis.c for all changes in that structure.
Signed-hostap: Jouni Malinen <j@w1.fi>
The GTK is renewed in the hostapd after a MIC attack dissassociation
without informing the driver, causing decryption failures. This patch
sends the new GTK/IGTK to the driver after it is updated by the hostapd.
Signed-off-by: Yoni Divinsky <yoni.divinsky@ti.com>
When the supplicant acts as a registrar to learn the access point
parameters send the credentials to the wpa_cli interface after
receiving the 7th message. This is needed for proper behavior with
wps_cred_processing set to 1 or 2.
Without this patch, after the 7th message you got the WPS-CRED-RECEIVED
notification without the credentials. This was because the cred_attr and
cred_attr_len were not filled in in the wps structure.
Signed-off-by: Olivier Sobrie <olivier@sobrie.be>
In scan.c, merge a channel's noise value into the scan results. When
comparing scan results, compute the signal-to-noise ratio and use it
when available. Prefer a 5 GHz network if its SNR is really big (> 30)
or if its SNR is relatively close to the other network's.
Some stations have been reported to send EAPOL-Key Error Reports
indicating Michael MIC failures even when the cipher is not TKIP
(e.g., when the network is using only CCMP). Ignore such reports
to avoid starting TKIP countermeasures unnecessarily. This can
prevent certaint types of denial of service attacks by insiders,
but mostly this is to work around invalid station implementations.
Signed-hostap: Jouni Malinen <j@w1.fi>
The eloop timeout to stop TKIP countermeasures has to be canceled
on deinit path to avoid leaving bogus timeouts behind.
Signed-hostap: Jouni Malinen <j@w1.fi>
Converts from a string to a control request enum when input
from a control interface is received. Will be used by a
subsequent patch.
Signed-off-by: Dan Williams <dcbw@redhat.com>
Control requests will be extended for non-EAP uses later, so it makes
sense to have them be generic. Furthermore, having them defined as an
enum is easier for processing internally, and more generic for control
interfaces that may not use field names. The public ctrl_req_type /
field_name conversion function will be used later by the D-Bus control
interface too.
Signed-off-by: Dan Williams <dcbw@redhat.com>
Special processing is needed to handle EAP user request for
identity or password at the beginning of Phase 2 when the implicit
identity request is used. data->pending_phase2_req needs to be set
to an empty buffer in that case to avoid re-processing the previous
part of TLS negotiation when the user enters the needed information.
Signed-hostap: Jouni Malinen <j@w1.fi>
Instead of hardcoding IEEE 802.11b rates to be disabled whenever
P2P support is built in and supported by the driver, do this only
when an interface is set to P2P mode (both when adding a new
interface and when changing the interface mode). This re-enables
use of IEEE 802.11b APs with CONFIG_P2P=y builds.
Signed-hostap: Jouni Malinen <j@w1.fi>
This function was used unconditionally if wpa_supplicant build
includes CONFIG_P2P=y. Adding a separate driver_ops for such use
is not really useful since the driver wrappers can do the same
internally. Remove this driver_ops and move matching functionality
into driver_nl80211.c which was the only driver wrapper using
this driver_ops callback.
Signed-hostap: Jouni Malinen <j@w1.fi>
nl80211_disable_11b_rates() function was supposed to use the 'disabled'
parameter to figure out whether to disable or re-enable 802.11b rates.
In addition, the driver deinit path was now ending up re-disabling the
enabled rates at the end of the deinit operations when the interface
type was forced back to station. Both of these issues are now fixed
and the 2.4 GHz band TX rate index is cleared properly when
wpa_supplicant is stopped.
Signed-hostap: Jouni Malinen <j@w1.fi>
disable_11b_rates() is called on interface addition,
but not on interface type change, resulting in 11b rates
enabled on p2p interfaces.
Signed-off-by: Eliad Peller <eliad@wizery.com>
This can be used to apply the no-CCK rule conditionally depending on
which frame is being sent. The no-CCK rule applies only for P2P
management frames while SA Query and FT use cases do not have similar
restrictions.
Signed-hostap: Jouni Malinen <j@w1.fi>
Add a new attribute to specify whether to use CCK rate or not during
scan and sending management frames. This helps to send P2P probes
at non-CCK rate. This ensures that P2P probe request/response/action
frames are always sent at non-CCK rates.
Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
This was included in the AP configuration parameters, but there has
been no way to set it to anything else than the default automatic
mode. Remove this parameter and just hardcode ERP determination to
follow the automatic mode.
The MS-CHAPv1 and MS-CHAPv2 RFCs specify that the password is a string
of "Unicode characters", which for Windows means UCS-2; thus the
password could be any even-length string of up to 512 bytes.
Instead of making the incompatible change of requiring the incoming
password to be UCS-2 encoded, assume the password is UTF-8 encoded and
convert it before using it in NtPasswordHash and
EncryptPwBlockWithPasswordHash.
Signed-off-by: Evan Broder <ebroder@mokafive.com>
The netlink event processing is delayed since they are only returned
after control returns to eloop. This can result in netlink down events
being processed at a point when the interface has is actually still
there (the first event on new interface is down) and that can result in
odd behavior especially when the resulting interface-disabled event is
delivered to AP mode interface with wpa_supplicant.
Work around this by filtering netlink down events if the matching
interface is up at the time the netlink event is processed. This fixes
an issue brought up by commit 36d84860bb.
hostapd_wpa_auth_conf() is called on uninitialized memory and the
conditional blocks in this function may leave some fields into
uninitialized state. This can result in unexpected behavior elsewhere
since some of the variables may be used without matching #ifdef
blocks. Fix this by zeroing the memory.
TPK will be needed to be able to generate FTIE MIC for the
teardown message, so maintain peer data for a bit longer in
case the teardown FTIE request comes back from the driver.
The Query Response Length Limit is not allowed to be zero when
this is sent by the AP. Use 0x7F to indicate that the limit is
based on maximum number of GAS fragments.
For P2P, the p2p_connect takes in device address argument to make a
connection. However the connected event AP-STA-CONNECTED comes with
interface address. The application listening on events would find it
difficult to map interface address to the p2p device address which is
provided for connection.
Append P2P Device Address to AP-STA-CONNECTED event for P2P Client
connection. This will help applications to easily map the P2P Interface
Address to P2P Device Address on CONNECTED event. For non-P2P case, it
will just print the usual STA MAC address alone.
Signed-off-by: Jithu Jance <jithu@broadcom.com>
The persistent_reconnect configuration parameter was used to decide
whether to accept invitation to re-establish a persistent group.
However, this was not being advertised in the Group Capability bitmap.
Add the Persistent Reconnect bit based on this configuration to GO
Negotiation frames and Beacon/Probe Response frames from the GO.
Use the wpa_tdls_add_peer function to allocate TDLS peer structures.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
Use capability information to decide whether to perform a given TDLS
operation internally or through mgmt-frame Tx.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
Before commencing setup, add a new STA entry to the driver representing
the peer. Later during setup, update the STA entry using information
received from the peer.
Extend sta_add() callback for adding/modifying a TDLS peer entry and
connect it to the TDLS state machine. Implement this callback for the
nl80211 driver and send peer information to kernel.
Mark TDLS peer entries with a new flag and translate it to a
corresponding nl80211 flag in the nl80211 driver.
In addition, correct TDLS related documentation in the wpa_driver_ops
structure.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
Record the capabilities and supported rates of the TDLS peer during
link setup. These are given in the IEs passed in Setup Request and
Setup Response frames.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
Disable the direct connection when a TDLS peer stops responding
to packets, as indicated by the "LOW ACK" event coming from a driver.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
Allow sending a TDLS discovery request as a frame through the driver.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
When a driver does not implement the TDLS_TEARDOWN operation internally,
send an explicit TDLS link teardown frame to the driver.
Change all teardown calls to use these calling semantics.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
Register for the TDLS discovery response public action frame in nl80211.
Print out a debug message when a Discovery Resp frame is received and
validated.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
When a discovery request is received, add the peer to the TDLS peer
cache and send a response containing minimal data. Mandatory IEs in
the discovery response frame will be filled out by the driver.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
Put glue code in place to propagate TDLS related driver capabilities to
the TDLS state machine.
If the driver doesn't support capabilities, assume TDLS is supported
internally.
When TDLS is explicitly not supported, disable all user facing TDLS
operations.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
Allow passing high-level TDLS commands and TDLS frames to kernel
via new nl80211 commands.
Propagate TDLS related nl80211 capability flags from kernel and add them
as driver capability flags.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
A zero dialog-token is considered invalid by IEEE Std 802.11z-2010.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
Cc: Kalyan C Gaddam <chakkal@iit.edu>
This function is actually needed for quite a bit more than just
user space MLME implementation. The old comment here has not been
accurate for a long time, so get rid of it.
These driver_ops functions set_channel, set_ssid, set_bssid,
mlme_add_sta, and mlme_remove_sta were used with the user space
MLME implementation in wpa_supplicant. That was only for testing
purposes and was removed, but these driver_ops were forgotten.
Remove them now to remove confusing driver_ops definitions.
This code was used only with driver_test.c to allow MLME operations
in hostapd to be tested without having to use a real radio. There
are no plans on extending this to any other use than testing and
mac80211_hwsim has now obsoled the need for this type of testing.
As such, we can drop this code from wpa_supplicant to clean up the
implementation of unnecessary complexity.
Netlink sockets can be shared among all driver instances, saving lots
of sockets, spurious log messages, memory, and CPU usage when using
multiple interfaces in a single process.
This offloads the station polling to driver wrappers, which may offload
it again to the driver. The hostap driver wrapper uses "real" data
frames while nl80211 uses null data frames.
Also add a specific event to indicate that a poll was successful for
future use with the nl80211 driver.
Some dualband cards can use more than five seconds to run through
a full scan, so increase the timeout to avoid hitting the missing
scan completed event workaround.
This fixes an issue where WPS run leaves a small ClientTimeout
value (2) configured and the next EAPOL authentication is started
with that small value even for Identity exchange. This can cause
problems when an EAPOL packet gets dropped immediately after
association and a retry of that packet is needed (which may take
more than two seconds).
This function can be used to clean up local UNIX domain socket files
that may be left over from clients that were previously connected to
wpa_supplicant. At least for now, this is only available for Android
builds.
Remove the separate driver_ops functions set_cts_protect(),
set_preamble(), set_short_slot_time(), and set_ht_params(). These
belong into same set of operations as set_ap(), so there is no need
to maintain separate functions that just make the driver wrapper
more complex.
Since these have only been used with driver_nl80211.c, the driver_ops
can be removed immediately instead of maintaining backwards
compatibility period with the old functions.
The AP client isolation parameter is now available through set_ap().
driver_nl80211.c was the only driver wrapper using the set_intra_bss()
call in hostap.git, but some external trees may have used this. Once
those are cleared, the set_infra_bss() driver_ops can be removed
completely. The only remaining use case for it currently is in P2P
GO mode with wpa_supplicant.
AP mode operations were used for P2P device discovery Listen state
only during early experiments. This has now been cleaned up and
ieee802_11_set_beacon() is not called for P2P device discovery.
As such, this hack to skip Beacon configuration can be removed.
This is the first step in allowing SA Query mechanism in hostapd to be
used with drivers that implement authentication and association MLME/SME
(i.e., do not use ieee802_11.c).
This adds the basic mechanism for running through network selection:
scan, ANQP fetch, network selection, and connection. Actual rules for
network selection and the creation of the network block are still
missing, but will be added in separate commits.
"hostapd_cli ess_disassoc (STA addr) (URL)" can now be used to send
an ESS Dissassociation Imminent notification to the STA. This event
is shown in wpa_supplicant ctrl_iface monitors (e.g., wpa_cli):
"WNM: ESS Disassociation Imminent - session_info_url=http://example.com/session/"
The new wpa_supplicant.conf file global parameters interworking and
hessid can be used to configure wpa_supplicant to include
Interworking element in Probe Request frames.
Commit 17f6b90056 moved the concatenation
of WPS IEs, but did not include the validation that the IE buffer is not
NULL. In addition, the concatenation needs to be done based on the full
IE buffer instead of the parsed pointer that includes only a single
WPS IE.
The base64 encoded data cannot included pad characters in the middle, so
we can stop the loop at the first sequence of pad characters. If the
sequence includes more than two pad characters, the encoding is invalid
and we can indicate failure.
Use the SSID filter list passed in the scheduled scan request down to
the kernel driver, so it can use the list to return only the wanted
SSIDs. Some kernel drivers can use this information to offload the
SSID filter to the hardware, helping with reducing the power
consumption.
Signed-off-by: Luciano Coelho <coelho@ti.com>
Pass SSIDs to be matched in scheduled scan results. Only the SSIDs
that are included in the match lists will be reported by the driver,
so the filtering can be offloaded to the hardware and the power
consumption can be reduced.
Signed-off-by: Luciano Coelho <coelho@ti.com>
With scheduled scan support, we may need to pass more than 10 SSIDs in
a single scan request. Some drivers (e.g., wl12xx) support up to 16
SSIDs at once.
Change WPAS_MAX_SCAN_SSIDS from 10 to 16.
Signed-off-by: Luciano Coelho <coelho@ti.com>
In new Linux kernel versions (>=3.0), nl80211 adds scheduled scan
capability. In order to use this feature to its full extent, we need
to support it in the wpa_supplicant core, so that it can also be used
by other drivers.
This commit adds initial scheduled scan support operations and events.
Signed-off-by: Luciano Coelho <coelho@ti.com>
Commit 19ec1f262e tried to fix some
cases for history prev selection, but it broke others. Fix this
properly by using a separate entry for the current edit line that
is not yet in history buffer.
Currently the P2P Interface name[p2p-%s-%d] is reset when the P2P
Interface name reaches the "IFNAMSIZ" limit. Monitor interface name is
derived from p2p interface name with the addition of few characters
[mon.p2p-%s-%d] and hence Monitor interface name hits IFNAMSIZ limit
before P2P Interface name. Rename the monitor interface name to
mon-%s-%d to reduce the length to same with p2p-%s-%d.
Data path for stations that get successfully associated to the
hostapd but fail to get added in the driver/firmware, will not
work. In such cases, hostapd should deauth and disconnect such
stations. In such scenario, hostapd should disconnect the STAs.
Sample output with following patch
wlan0: STA 0c:74:c2:9a:4c:59 IEEE 802.11: authenticated
wlan0: STA 0c:74:c2:9a:4c:59 IEEE 802.11: associated (aid 1)
wlan0: AP-STA-CONNECTED 0c:74:c2:9a:4c:59
wlan0: STA 0c:74:c2:9a:4c:59 IEEE 802.11: Could not add STA to kernel driver
wlan0: STA 0c:74:c2:9a:4c:59 IEEE 802.11: deauthenticated due to local deauth request
Signed-off-by: Yogesh Ashok Powar <yogeshp@marvell.com>
Signed-off-by: Nishant Sarmukadam <nishants@marvell.com>
The password_hash parameter was not copied in case of the integrated
authentication server (but was for RADIUS server). This broke EAP
authentication when the user entry used NT-hash.
This currently unused function would have triggered wpabuf overflows
due to incorrect variable being reset to zero in the case the old
NoA wpabuf was large enough for the new data.
The EAPOL authenticator was previously forcing disconnection in the WPS
use case. However, this can be benefitial operation with any IEEE 802.1X
authentication mechanism and need not be limited to WPS. This helps some
use cases like EAP-FAST where provisioning may require two
authentication runs if the authentication server does not allow the PAC
provisioning step to be used for normal data connection. While the
station would be free to decide to re-associate in such a case, not all
stations do and as such, it helps if the AP does that instead of leaving
the association up with EAPOL state machine in HELD state for 60
seconds.
While EAP-FAST uses protected success notification, RFC 5422, Section
3.5 points out a possibility of EAP-Failure being sent out even after
protected success notification in case of provisioning. Change the
EAP-FAST peer implementation to accept that exception to the protected
success notification. This allows the station to re-connect more quickly
to complete EAP-FAST connection in the case the server rejects the
initial attempt by only allowing it to use to provision a new PAC.
If GO Negotiation Request (or in theory, also GO Negotiation Response)
frame is delivered multiple time for processing, the SSID of the group
could end up getting changed. This could result in possible issues if
the peer ended up using different SSID. To avoid this, make sure the
SSID does not get changed unless the negotiation is for a new group.
Whenever PMK gets changed (e.g., due to re-authentication), all PMKSA
caching entries that were created using the previous PMK needs to be
replaced. Previously, only the entry for the current AP was cleared.
Flush the other entries based on network_ctx matches to get rid of the
OKC entries. These entries can then be re-creating using OKC with the
new PMK.
When polling a station that has been inactive for a while, hostapd currently
always uses a null data frame. This is a bit strange with uAPSD clients
(though it seems to mostly work) since the EOSP bit can never be set in a
non-QoS frame. Make hostapd use QoS null data frames for probing when the
station is a QoS STA.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Wrap self-generated WPS credential for new AP settings and send that to
control interface to provide the needed information in
WPS-NEW-AP-SETTINGS for external processing.
GAS/ANQP is a generic protocol and in no way specific to P2P, so move
routines used to build GAS/ANQP frames to a separate file that can be
shared for other uses than just P2P service discovery.
The new function, p2p_scan_ie_buf_len(), can be used to figure out
how large a buffer needs to be allocated for p2p_scan_ie() use. This
makes it easier to add new data into the buffer without forcing all
callers to be updated to use a larger buffer.
The P2P search mechanism depends on the same scan functionality that
is used for station mode scans. If these operations are being used
at the same time, scan result processing is not handled properly.
Avoid unexpected behavior by delaying station mode scan requests
if a P2P operation is in progress.
Among other things, this allows the station mode connection attempt
to be continued after a P2P find or group formation has been completed
if the interface is available (i.e., when the P2P group uses a
separate virtual interface).
If the initial attempt to start a scan for p2p_find fails, an error
is reported. However, the P2P scan handler and search state was
left behind. That can result in unexpected behavior when the next
non-P2P scan results are indicated. Avoid this by clearing the
P2P search state on failure.
This new flag can be used to change wpa_supplicant behavior in the
default ap_scan=1 mode to move BSS selection into the driver (likely
firmware). This commit is only adding the flag; separate commits
will be used to change the actual connection/roaming behavior.
These protocols seem to be abandoned: latest IETF drafts have expired
years ago and it does not seem likely that EAP-TTLSv1 would be
deployed. The implementation in hostapd/wpa_supplicant was not complete
and not fully tested. In addition, the TLS/IA functionality was only
available when GnuTLS was used. Since GnuTLS removed this functionality
in 3.0.0, there is no available TLS/IA implementation in the latest
version of any supported TLS library.
Remove the EAP-TTLSv1 and TLS/IA implementation to clean up unwanted
complexity from hostapd and wpa_supplicant. In addition, this removes
any potential use of the GnuTLS extra library.
This fixes some build issues in GnuTLS wrapper to be compatible with
at least following GnuTLS versions: 2.2.5, 2.4.3, 2.6.6, 2.8.6,
2.10.5, 2.12.11, 3.0.3.
Instead of using separate bad_record_mac and decryption_failed alerts,
use only bad_record_mac alert regardless of how the CBC decryption
failed. This provides less information to attackers that could modify
packets. In addition, instead of returning immediately on error, run
through the MAC check to make timing attacks more difficult.
When the received data will be decrypted, there is no need to first
copy it and then handle decryption in-place when decryption step can
take care of both operations.
TLS v1.0 and v1.1 RFCs were not exactly clear on the use of the
protocol version in record later. As such, accept any {03,xx} value
to remain compatible with existing implementations and new protocol
versions.
The p2p->drv_in_listen variable is used to track Listen state operations
in the driver. This is cleared when the driver reports that the Listen
state has ended (p2p_listen_end() gets called). However, it is possible
that the driver does not indicate that if the Listen state is canceled.
This can apparently happen in some cases where p2p_connect command is
issues while the Listen state is in progress.
Work around this issue by clearing p2p->drv_in_listen when Listen state
is stopped as part of p2p_stop operation. This allows the P2P module to
process CONNECT_LISTEN timeout in p2p_timeout_connect_listen() to move
to CONNECT state, e.g., when starting GO Negotiation after Device
Discoverability mechanism.
The "too long" buffer (32+1 octets) prevented AP from starting up with
32 octet SSID with WE-21 and newer. Fix this by reducing the
SIOCGIWESSID buffer length.
This adds support for RSN pre-authentication with nl80211 interface and
drivers that handle roaming internally (i.e., wpa_supplicant is not
responsible for requesting scan results that provide the needed
information without these events).
cfg80211 does not currently allow NL80211_CMD_CONNECT when there
is already an association. This can result in wpa_supplicant being
unable to request a connection if the interface is already in
connected state. Work around this by requesting disconnection if
the connect command fails with EALREADY error. This allows
wpa_supplicant to recover from the unexpected state and complete
the connect request on next attempt.
eapol_test command line argument -o<file> can now be used to request
the received server certificate chain to be written to the specified
file. The certificates will be written in PEM format. [Bug 391]
Make the WPS processing of (Re)Association Request frame IEs conditional
on WPS support. This fixes a build issue with wps_is_20() not being
defined and makes the non-WPS build a bit smaller.
os_*() wrappers should be used instead of functions from time.h.
Removing the header from includes.h enforces this. os_unix.c can
include this its uses are valid wrapper calls. wps_upnp.c uses
gmtime() for which there is no os_*() wrapper available yet, so
allow it to use time.h, too. Similarly, allow dump_state.c to
use time.h for ctime().
Currently, the following can happen:
1) P2P state machine requests R-O-C
2) user changes their mind and aborts
3) P2P state machine aborts R-O-C
4) driver_nl80211 rejects abort since there
was no notification about the start yet
5) R-O-C period start notification from kernel
6) P2P state machine requests new R-O-C
7) this overlaps with old R-O-C -- iwlwifi driver
can't handle that and returns -EBUSY
8) state machine stops dead in its tracks
The reason is that the abort isn't going through properly. Instead of
tracking whether a R-O-C is active in driver_nl80211, track whether one
was requested to avoid this scenario.
Reported-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This allows drivers to disable CCK rates from Probe Request frames.
For nl80211, this is currently applying only to the supported rates
element(s), but this mechanism could be extended to address TX rate
control masking, too, to lessen need for global rate disabling.
Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Since the kernel can now advertise P2P concurrent support by advertising
interface combinations, we can take advantage of that and automatically
use P2P_CONCURRENT / P2P_MGMT_AND_NON_P2P for drivers that advertise
support.
Keep driver_param=use_p2p_group_interface=1 for anyone not advertising
interface combinations in their drivers yet.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The default config methods was hardcoded to claim support for
PushButton, Display, and Keypad. While these are supported by
most P2P devices, there may be some cases where it is convenient
to be able to disable a specific config method. Use config_methods
configuration parameter to set the default values for Config Methods
in the P2P Device Info attribute.
The P2P specification (3.1.4.3) disallows use of the Label configuration
method between two P2P devices. This was previously enforced at upper
level, but the obsolete code can be removed from wpa_supplicant. This
adds a bit more strict enforcement of the policy, but should not result
in practical differences since no known P2P implementation uses Label
config method.
If a network configuration block is removed or modified, flush
all PMKSA cache entries that were created using that network
configuration. Similarly, invalidate EAP state (fast re-auth).
The special case for OKC on wpa_supplicant reconfiguration
(network_ctx pointer change) is now addressed as part of the
PMKSA cache flushing, so it does not need a separate mechanism
for clearing the network_ctx values in the PMKSA cache.
If the driver maintains its own copy of the PMKSA cache, we need to
clear an entry from the driver whenever wpa_supplicant is dropping
an old PMKSA cache entry.
The Deauth/Disassoc TX status callbacks were ending up kicking the
station entry from kernel driver when test functionality was used to
inject Deauth/Disassoc frames from the AP with the purpose of leaving
the local association in place. Fix this by using STA flags to figure
out whether there was a pending callback for the frame that we need
to act on.
In addition, add forgotten functionality for the Disassoc TX status
callback to match the behavior with Deauth.
The previous code was trying to figure out which WPA version is
used based on the extra IEs requested for Association Request. That
did not work properly in cases where non-WPA networks are used with
some extra IEs. Fix this by using more robust mechanism for passing
the WPA versions from core wpa_supplicant to the driver_ops
associate().
Instead of hardcoding the Config Methods attribute value in Probe
Request frames, set this based on the configured parameter
config_methods to allow correct set of methods to be advertised.
The device description file may include multiple devices. Improve the
simplistic parser by first trying to find the WFADevice:1 device before
fetching the device parameters. While this is still far from complete
XML parsing, this should address the most common root device
specifications.
This allows non-mac80211 drivers to report received Probe Request
frames to user space and Probe Response and Action frames to be sent
from user space when using AP/P2P GO mode.
This allows the driver to use PS buffering of Deauthentication and
Disassociation frames when the STA is in power save sleep. The STA
entry (and PTK) will be removed from the kernel only after the
Deauth/Disassoc has been transmitted (e.g., when the STA wakes up).
A hardcoded two second timeout is used to limit the length of this
window should the driver fail to deliver the frame (e.g., the STA
is out of range and does not wake up). The kernel STA entry is
marked unauthorized during the wait to avoid accepting Data
frames from the STA that we have decided to disconnect.
This behavior is available only with drivers that provide TX status
events for Deauth/Disassoc frames (nl80211 at this point). Other
drivers continue to use the previous behavior where the STA entry
is removed immediately.
Extend the code that waits for the station to send EAPOL-Start before
initiating EAPOL authenticator operations to cover the case where the
station includes WPS IE in (Re)Association Request frame if that IE
does not include support for WPS 2.0. While this should not really
be needed, this may help with some deployed WPS 1.0 stations that do
not support EAPOL operations correctly and may get confused of the
EAP-Request/Identity packets that would show up twice if EAPOL-Start
is transmitted.
If an ER tries to send a message to a STA that is not in the middle
of WPS protocol, do not try to deliver that. This can help with issues
where an ER takes long time to reply to M1 and another Registrar has
already completed negotiation.
It is possible to get a response for a pending EAP callback after the
EAP state machine has already completed its work or has timed out. For
those cases, make sure that the callback function is not delivered since
it could result in NULL pointer dereferences.
The Config Methods attribute in M2 and M2D messages is supposed to
indicate which configuration methods are supported by the Registrar. As
such, it should not depend on whether PBC mode is currently active or
not. That will only affect the Selected Registrar Config Methods and
Device Password ID attributes.
This change is a first step in better supporting different driver
architectures for AP mode operations with nl80211. This commit in
itself does not add new functionality, but it makes it easier to add
new AP mode parameters to address needs of drivers that have more of
the MLME/SME in firmware or kernel.
Split WPA initialization into two parts so that the Beacon frames can be
configured fully before the initial keys (GTK/IGTK) are configured. This
makes it easier for drivers that depend on the AP security mode being
fully set before the keys are configured.
This avoids a request to the driver to first start beaconing before
the WPA/RSN IE has been generated and then immediately changing the
beacon IEs once the WPA/RSN IE is ready.
Such using params->wpa_ie to figure out whether the connection is for
WPA/WPA2 is not correct since that buffer is used also to add WPS IE. In
case of WPS, do not add NL80211_ATTR_WPA_VERSIONS to avoid confusing
drivers.
Some compilers complain about fwrite calls if the return value is
not checked, so check the value even if it does not really make
much of a difference in this particular case.
The internal TLS implementation assumes that the certificate chain
is ordered by issuer certificate following the certificate that it
signed. Add the certificates to the chain in suitable order when
loading multiple certificates.
Commit e1d526293b added code for verifying
whether the receive Probe Request frame was indicating support for only
802.11b rates, but it missed the for loop for the extended supported
rates element. Add that to fix the validation code for cases where
non-802.11b rates are in the extended supported rates element.
Currently, wpa_driver_nl80211_authenticate() changes the interface type
to station. However, in case of P2P, we need to change the interface
type to P2P_CLI.
Add p2p field to the authentication params, and consider it for choosing
the correct interface type.
Signed-off-by: Eliad Peller <eliad@wizery.com>
Commit 9f51b11395 added support for P2P
interfaces when adding a new interface. However, it didn't handle the
case in which the same interface is being used and its type is being
changed. Add support for this case.
Consequently, when doing "ap_scan_as_station" we now need to save the
actual AP interface type (AP/P2P GO) in order to restore it properly.
For that, change ap_scan_as_station type from int to nl80211_iftype, and
set it to NL80211_IFTYPE_UNSPECIFED when not used.
Signed-off-by: Eliad Peller <eliad@wizery.com>
When driver_nl80211.c has to set the netdev down to change iftype, an
RTM_NEWLINK event is generated. Do not generate
EVENT_INTERFACE_DISABLED event based on that.
Reassociation Request/Response frame validation need to count all IEs in
the RIC. In addition, TIE is not protected, so it should not be included
in the count.
Signed-off-by: Hong Wu <hong.wu@dspg.com>
Only accept Probe Request frames that have a Wildcard BSSID and a
destination address that matches with our P2P Device Address or is the
broadcast address per P2P specification 3.1.2.1.1.
When a new interface is created and already has a separate MAC address
assigned by the kernel, then we need to use that address, not just when
we've created a locally administered address.
This fixes use_p2p_group_interface=1 for iwlagn as it already makes
mac80211 assign an address for a second interface since the hardware has
two addresses assigned.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Add support to wpa_supplicant for device-based GTK rekeying. In order to
support that, pass the KEK, KCK, and replay counter to the driver, and
handle rekey events that update the latter.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Send STA flags to kernel when adding a new station. This ensures
stations are added with up to date flags by kernel drivers.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
When adding a new station, set the STA flags as part of the sta_add()
command. This ensures the flags are up to date when the station is added
by lower level drivers.
Signed-off-by: Arik Nemtsov <arik@wizery.com>
A new hostapd configuration parameter, disable_pmksa_caching=1, can now
be used to disable PMKSA caching on the Authenticator. This forces the
stations to complete EAP authentication on every association when WPA2
is being used.
In general, this patch attemps to extend commit
00468b4650 with dbus support.
This can be used by dbus client to implement subject match text
entry with preset value probed from server. This preset value, if
user accepts it, is remembered and passed to subject_match config
for any future authentication.
Signed-off-by: Michael Chang <mchang@novell.com>
Data path is broken when hostapd reloads its configuration
disabling the security which was previously enabled (WEP/WPA),
using kill -1, as old keys were not cleared.
The patch clears the keys configured when hostapd reloads
its configuration.
Signed-off-by: Yogesh Ashok Powar <yogeshp@marvell.com>
This phase1 parameter for TLS-based EAP methods was already supported
with GnuTLS and this commit extends that support for OpenSSL and the
internal TLS implementation.
Fix a bug with the current dev_found()/dev_lost() usage. Previously
in p2p_device_free() dev_lost() was invoked for devices that had
not been dev_found(). This caused dbus related msgs to stderr like:
"Attempted to unregister path (path[0] = fi path[1] = w1) which isn't
registered"
Signed-off-by: Angie Chinchilla <angie.v.chinchilla@intel.com>
Some P2PDevice properties were not updated in p2p->cfg structure:
reg_class, channel, op_reg_class, and op_channel. Hence, update p2p->cfg
parameters through p2p core calls in wpas_p2p_update_config().
Signed-off-by: Jean-Michel.Bachot <jean-michelx.bachot@intel.com>
Signed-off-by: Jayant Sane <jayant.sane@intel.com>
Move level parameter from p2p_device to p2p_device_info in order to
expose this information and modify D-Bus P2P handler to return this new
parameter through the P2P device properties.
Signed-off-by: Fabien Marotte <fabienx.marotte@intel.com>
Signed-off-by: Jayant Sane <jayant.sane@intel.com>
Previously hostapd just masked the STAs HT caps with its own. However,
some HT caps are not symmetric and as such need to be handled
different.
hostapd shouldn't overwrite the STAs SMPS mode as otherwise the driver
cannot know it has to use RTS/CTS to wake the receiver from dynamic
SMPS for MCS rates > 7.
hostapd shouldn't mask the RX and TX STBC caps with it's own. They are
already handled in a special case below.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
This file is a part of the radiotap parser that Andy Green agreed to
relicense under the BSD license (per email, 11 Aug 2007 07:42:05
+0100). The copyright/license statement was updated in radiotap.c,
but this radiotap_iter.h file was forgotten at that point.
In set_key handler, the seq[8] is in little endian order defined by
WPA. BSD kernel uses a u_int64_t value ik_keyrsc to represent it
internally. The kernel expects the native endian order for the value.
Thus, we need to detect the endian order and swap bytes when
necessary.
This fixes an issue with SA Query Response frames not being processed
anymore after wpa_supplicant started registering a handler for those.
This handler registration is in generic driver_nl80211.c code, so
hostapd uses it, too.
Implement PMKSA cache operations add, remove, and flush using nl80211
commands NL80211_CMD_{SET,DEL,FLUSH}_PMKSA to support PMKSA caching
with drivers that select the AP and generate the RSN IE internally.
When provision discovery fails, this new callback will be called
so P2P users can react to the failure.
Signed-off-by: Jayant Sane <jayant.sane@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Since the peer may not be in Listen state when the provision discovery
request is sent, try to send the request again number of times when in
IDLE state. This was already done when p2p_find is in progress, but this
commit adds retries to the case where no other P2P operations are in
progress.
Signed-off-by: Jayant Sane <jayant.sane@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This can be used to avoid rejection of first two 4-way handshakes every
time hostapd (or wpa_supplicant in AP/IBSS mode) is restarted. A new
command line parameter, -e, can now be used to specify an entropy file
that will be used to maintain the needed state.
Some of these are required for proper functionality (like
get_seqnum); others may not be needed yet, but including them
allows some extra ifdef/endif blocks to be removed.
.set_rate_sets is not defined for non-hostapd, which prevents
configuring basic_rates when working as P2P GO.
Signed-off-by: Eliad Peller <eliad@wizery.com>
If the driver wrapper is setting up the interface up only at commit(),
UPnP initialization fails. Fix that by moving UPnP setup to happen after
the driver commit() call.
Windows 7 uses incorrect way of figuring out AP's WPS capabilities by
acting as a Registrar and using M1 from the AP. The config methods
attribute in that message is supposed to indicate only the configuration
method supported by the AP in Enrollee role, i.e., to add an external
Registrar. For that case, PBC shall not be used and as such, the
PushButton config method is removed from M1 by default. If pbc_in_m1=1
is included in the configuration file, the PushButton config method is
left in M1 (if included in config_methods parameter) to allow Windows 7
to use PBC instead of PIN (e.g., from a label in the AP).
cfg80211 maintains separate BSS table entries for APs if the same
BSSID,SSID pair is seen on multiple channels. wpa_supplicant does
not use frequency as a separate key in the BSS table, so filter out
duplicated entries. Prefer associated BSS entry in such a case in
order to get the correct frequency into the BSS table.
It may take some time for the TX status to be delivered for a
(Re)Association Response frame and if any Data frames are received
during that time, they may end up getting dropped as Class 3 frames in
not-associated state. This results in a Disassociation frame being sent
to the station and it assuming that the association has been lost.
Work around the issue by remembering that the (Re)Association Request
has already been accepted and skip the Deauth/Disassoc sending because
of the possible Class 3 frames before the TX status callback is
received.
In theory, these calls could fail, but it is not really likely to
happen in practice in the use case here. Anyway, check that they do
not return an error before accepting the length of the file.
This function does not get called with in_data == NULL in practice, but
it seems to be at least partly prepared for that case, so better make it
consistent by handling the NULL value throughout the function.
Both the SA and IEs from the received Probe Request frames must be
included and the Probe Request RX callback functions may assume that
these are not NULL.
The change to use wpa_dbg() in wpa_sm_parse_own_wpa_ie() could result
in a NULL pointer dereference if the function were called when WPA
state machine has not been initialized. While this cannot really
happen in practice, it is better to be prepared for that since that
was the case before the wpa_dbg() change.
The size_t value here can be 64-bit and result in implicit sign
extension. In this particular case, that gets masked out by
host_to_be32(), so there is no practical difference, but it is better
to get rid of the 64-bit variable explicitly.
The supportedTypes parameter is a list of TNC_MessageType values
and the buffer to be copied should use size of TNC_MessageType, not
TNC_MessageTypeList. In practice, these are of same length on most
platforms, so this is not a critical issue, but anyway, the correct
type should be used.
This avoids reading past the end of the IE buffer should the FTIE
be too short. In addition, one debug hexdump was using uninitialized
pointer to the FTIE buffer, so fixed it to use the pointer from the
parse data.
AF_INET6 is not always enabled by default, so use AF_INET instead. In
addition, use the old fixed length, 2048, as a failover value if the
sysctl fails for any reason.
Commit 03d3f28a69 broke initialization of
EAPOL authenticator state machines since an error value from
wpa_auth_sta_key_mgmt() (-1) was not handled properly and the fixed
wpa_key_mgmt_wpa_psk() identified the value as a PSK-based AKM because
of all bits being set to 1. The special error value needs to handled
separately to avoid skipping EAPOL state machine initialization.
These can be used in some cases with a bitfield having multiple
AKMs enabled (e.g., WPA-PSK and WPA-PSK-SHA256). Address those
cases by checking whether any of the matching AKM are included.
When the SME is in the driver or cfg80211, the automatic selection
of auth_alg is done by leaving out the NL80211_ATTR_AUTH_TYPE
attribute from the NL80211_CMD_CONNECT command.
This should fix EAPOL reauthentication and rekeying timeout issues
with Intel clients when using WMM (e.g., with IEEE 802.11n). These
stations do not seem to be able to handle EAPOL data frames as
non-QoS Data frames after the initial setup.
This adds STA flags to hapd_send_eapol() driver op to allow
driver_nl80211.c to mark the EAPOL frames as QoS Data frame
when injecting it through the monitor interface.
This provides a means for the supplicant to directly request signal
quality metrics from the driver. This is useful, for example for
background scan algorithms that might ask desire this information
out-of-band with CQM events.
Signed-off-by: Paul Stewart <pstew@google.com>
When building IE from an empty string and CONFIG_WPS_STRICT is
not defined, redundant 2 bytes are being added to the string.
We have to return right after building the "dummy" string.
Signed-off-by: Eliad Peller <eliad@wizery.com>
Some deployed supplicants update their SNonce for every receive
EAPOL-Key message 1/4 even when these messages happen during the
same 4-way handshake. Furthermore, some of these supplicants fail
to use the first SNonce that they sent and derive an incorrect PTK
using another SNonce that does not match with what the authenticator
is using from the first received message 2/4. This results in
failed 4-way handshake whenever the EAPOL-Key 1/4 retransmission
timeout is reached. The timeout for the first retry is fixed to
100 ms in the IEEE 802.11 standard and that seems to be short
enough to make it difficult for some stations to get the response
out before retransmission.
Work around this issue by increasing the initial EAPOL-Key 1/4
timeout by 1000 ms (i.e., total timeout of 1100 ms) if the station
acknowledges reception of the EAPOL-Key frame. If the driver does
not indicate TX status for EAPOL frames, use longer initial
timeout (1000 ms) unconditionally.
Store Manufacturer, Model Name, Model Number, and Serial Number
attributes for P2P peers and expose these through the existing
peer information mechanisms.
Previously, a bug in GKeyDoneStations count would remain in effect
until the authenticator is restarted. Make this more robust by
clearing the station count before each rekeying setup. While this
is not really supposed to be needed, there has been bugs in this
area in the past and it is safer to make the implementation recover
from such an issue automatically.
If the STA to be freed is still included in GKeyDoneStations count,
decrement the count when the STA is freed. This does not happen in
AP mode since there is enough time to go through the authenticator
state machine to clear the STA. However, in the current RSN IBSS
implementation, the authenticator state for the STA is freed in a
way that does not allow the state machine to go through the clearing.
To address this, make sure that wpa_free_sta_sm() decrements the
GKeyDoneStations count if the STA happened to be in the process of
GTK rekeying.
CFLAGS += -DCONFIG_RSN_TESTING in .config and rsn_testing=1 in
hostapd.conf can now be used to enable a testing mode that adds
extensions to RSN element. This can be used to check whether
station implementations are incompatible with future extensions
to the RSN element.
The AP code might, currently only in the case of HT40, defer actual
enabling to after a scan. In this case, the wpa_s AP code gets confused.
Add a callback for it to use and make it use it to finish only when the
setup has actually completed.
With appropriate hacks in place this allows using HT40 in P2P mode.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Make the P2P code keep track of WPS vendor extensions received from
peers so they can be exposed via DBus later.
Signed-off-by: Jean-Michel Bachot <jean-michelx.bachot@linux.intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This adds the ability to add WPS vendor extension attributes in P2P
frames, like GO Negotiation and Probe Response frames.
Signed-off-by: Jean-Michel Bachot <jean-michelx.bachot@linux.intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This adds the ability to add WPS vendor extensions to an AP (or GO).
They will be added to the WSC IE(s) in Beacon and Probe Response frames.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Instead of converting back and forth from the string representation,
always use the binary representation internally.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Use only the UUID-E to remove active PBC session(s) at the completion of
successful PBC protocol run. This fixes potential issues with Enrollees
that use multiple MAC addresses and as such, can get multiple entries in
the PBC session list.
Add Secondary Device Type List attribute in WSC IE for P2P
Probe Response frames if one or more secondary device types
are configured.
Signed-off-by: Jean-Michel Bachot <jean-michelx.bachot@linux.intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The secondary device type list is an optional attribute in the WSC IE.
Signed-off-by: Jean-Michel Bachot <jean-michelx.bachot@linux.intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Currently hostapd will force HT Mixed Mode if at least one non-GF STA is
associated. This will force _all_ HT transmissions to be protected.
802.11n-2009 doesn't require HT Mixed Mode to be used in case of non-GF
STAs but instead the HT information element contains a flag if non-GF
STAs are present. All STAs are required to protect GF transmissions in
that case. Hence, setting HT Mixed mode if non-GF STAs are present is
superfluous.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Use a specific return value, WPS-PBC-OVERLAP, to indicate a reason
for rejecting a wps_pbc command in wpa_supplicant AP mode if the
PBC mode cannot be started due to PBC session overlap having been
detected during monitor time.
This allows the user to complete WPS provisioning using PBC by
selected a specific Enrollee even if there are other Enrollees
in active PBC mode at the same time. The other Enrollees will be
rejected should they try to connect at the same time.
Ignore possible mismatches in the source address of the frame and only
use UUID-E to check whether a Probe Request or M1 is from the same
Enrollee when figuring out whether there is PBC session overlap. This
is needed to avoid potential issues with Enrollee devices that may have
multiple interfaces indicating active PBC state.
If there is already a link the requested peer, request start of
renegotiation instead of completely new link. This seems to be needed
to allow some driver to accept the trigger for a new negotiation.
tdls_testing 0x80 can now be used to request wpa_supplicant to send
a TDLS Setup Request frame (i.e., try to initiate TDLS setup)
whenever a TDLS Setup Request frame has been received from a peer.
Blindly clearing all struct wpa_tdls_peer members is a risky
operation since it could easily clear pointers to allocated
memory, etc. information that really should not be removed.
Instead of hoping that new code gets added here to restore
the important variables, reverse the approach and only clear
structure members one by one when needed.
This allows the same code path to be used for both protected and
unprotected configurations to limit need for duplicated code in
both the driver and wpa_supplicant.
This adds code for validating that the same Key Lifetime is used
throughout TPK handshake and enables TPK lifetime expiration and
renewal/teardown. These part seem to be working, but the actual
handling of TDLS Setup Confirm and renewal are not fully
functional yet.
tdls_testing bits:
bit 3 = use short TPK lifetime (301 seconds)
bit 4 = use wrong TPK lifetime in TDLS Setup Response
bit 5 = use wrong TPK lifetime in TDLS Setup Confirm
These special test cases can be configured at run time with "wpa_cli
tdls_testing <value>" where <value> is an integer (either as a decimal
or as a hex value with 0x prefix) bitmap of special features with
following bits available at this point:
bit 0 = long frame (add dummy subelement to make FTIE very long)
bit 1 = use alternative RSN IE (different RSN capab value and no extra
replay counters)
bit 2 = send incorrect BSSID in Link Identifier of TDLS Setup Request
(e.g., 1 = long FTIE, 2 = different RSN IE, 3 = both of those)
This is disabled by default and can be enabled for the build by
adding the following line to .config:
CFLAGS += -DCONFIG_TDLS_TESTING
This allows driver wrappers to indicate whether the association was
done using Association Request/Response or with Reassociation
Request/Response frames.
CFLAGS += -DCONFIG_VALGRIND can now be used to enable support for
Valgrind client requests to help in removing some non-issues from
Valgrind reports. This is mainly aimed at allowing unknown ioctl
results to be marked as defined data.
If the os_time_t variable used for the expiration time (seconds)
overflows when the registered timeout value is being added,
assume that the event would happen after an infinite time, i.e.,
would not really happen in practice. This fixes issues with
long key timeouts getting converted to immediate expiration due
to the overflow.
os_snprintf() can be a preprocessor macro and according to
C99 (6.10.3 clause 11) the results of having preprocessor directives
inside the macro arguments is undefined.
Build options can now be used to replace the location of client
sockets for UNIX domain socket control interface:
CFLAGS += -DCONFIG_CTRL_IFACE_CLIENT_DIR=\"/tmp\"
CFLAGS += -DCONFIG_CTRL_IFACE_CLIENT_PREFIX=\"wpa_ctrl_\"
The buffer size for routing socket is fixed to 2048.
This patch fix it to obtain the size from OS.
This patch worked on x86 platform with NetBSD 5.0.2.
This allows keystore:// prefix to be used with client_cert and
private_key configuration parameters.
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
This will only retrieve information about peers that have been fully
discovered, not peers that are only half-discovered based on their Probe
Request frames.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The DBus code will want to have perfect matching of dev_found and the
dev_lost it adds so it doesn't need to keep track internally. Enable
that with a new flag in the core that tracks whether we have already
notified about this -- the existing users can ignore it.
The part where this is always set to 1 if the new device is discovered
by a driver that has P2P in the driver is buggy -- the driver should
feed the P2P peer database and then that should feed the notification
here instead.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This embeds some information about each P2P peer that will be publically
visible in a struct that is shared.
The dev_found notification function is also passed the new struct, which
requires some work for the driver-based P2P management.
Signed-off-by: Konguraj(Raj) Kulanthaivel <konguraj.kulanthaivel@intel.com>
Signed-off-by: Fabien Marotte <fabienx.marotte@linux.intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
When using multiple vifs and dynamic vlan tagging is enabled on any
interface, the created AP VLAN interfaces get the BSSID of the first
AP mode interface instead of the BSSID of the corresponding AP mode
interface.
Example:
wlan0 - xx:xx:xx:xx:x0
wlan1 - xx:xx:xx:xx:x1
Assume a STA connects to the AP interface wlan1 and gets a dynamic
VLAN tag 100 assigned by the RADIUS server. Hostapd will create an AP
VLAN interface wlan1.100 but doesn't set an address for this interface
which results in wlan1.100 getting the same address as wlan0:
wlan1.100 - xx:xx:xx:xx:x0
As a result the STA that was moved to wlan1.100 isn't able to finish its
4-way handshake since mac80211 won't pass its frames to wlan1.100 due to
the different address.
To fix this issue make use of the address of the AP interface when
creating an AP VLAN interface.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Add MAC addresses for stations and use wpa_msg instead of printf
methods to make it easier to grep logs and find messages for the
station in question.
Signed-off-by: Ben Greear <greearb@candelatech.com>
If parsing of the certificate or private key succeeds using any of
the tried encoding types, clear the OpenSSL error queue without
showing the pending errors in debug log since they do not really
provide any useful output and can be confusing.
This patch fixes a problem I had when I tried to connect
an embedded system [wpa_supplicant, CONFIG_TLS=internal]
to my TLS secured network.
TLSv1: Send CertificateVerify
TLSv1: CertificateVerify hash - hexdump(len=36): ha .. ha
PKCS #1: pkcs1_generate_encryption_block - Invalid buffer lengths \
(modlen=512 outlen=454 inlen=36)
It turned out that a fixed 1000 byte message buffer was just
a little bit too small for the 4096 bit RSA certificates
I'm using.
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
In multi BSS setups it wasn't possible to set up an HT BSS in
conjunction with a WEP/TKIP BSS. HT needed to be turned off entirely
to allow WEP/TKIP BSSes to be used.
In order to allow HT BSSes to coexist with non-HT WEP/TKIP BSSes add a
new BSS conf attribute "disable_11n" which disables HT capabilities on a
single BSS by suppressing HT IEs in the beacon and probe response
frames. Furthermore, mark all STAs associated to a WEP/TKIP BSS as
non-HT STAs. The disable_11n parameter is used internally; no new entry
is parsed from hostapd.conf.
This allows a non-HT WEP/TKIP BSS to coexist with a HT BSS without
having to disable HT mode entirely. Nevertheless, all STAs associated to
the WEP/TKIP BSS will only be served as if they were non-HT STAs.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Use l2_packet with Ethernet header included so that the source
address for RRB packets can be forced to be the local BSSID. This
fixes problems where unexpected bridge interface address may end
up getting used and the recipient of the frame dropping it as
unknown R0KH/R1KH.
This is needed to keep kernel and hostapd in sync. In addition,
the obsolete entry in hostapd prevented use of Deauthentication
or Disassociation frame with reason code 6/7 mechanism to indicate
to the STA that there is no association if the STA missed the
broadcast Deauthentication frame for any reason.
The internal pointer to RADIUS client configuration needs to be
updated whenever a new hostapd configuration is loaded. Without
this, freed memory may be dereferenced and this can result in
segmentation faults.
This converts number of debugging messages to use wpa_msg() in order
to allow the interface name to be shown with the messages.
A new function, wpa_dbg(), is introduced to allow
CONFIG_NO_STDOUT_DEBUG=y builds to remove the debug strings. This is
otherwise identical with wpa_msg(), but it gets compiled out if stdout
debugging is disabled.
This is needed to allows WPS PBC session overlap detection to work
with drivers that process Probe Request frames internally. This
code is is run in hostapd, but the wpa_supplicant AP mode did not
have call to the hostapd_probe_req_rx() function even though it
registered handlers for hostapd Probe Request RX callbacks.
An optional parameter, p2p_dev_addr, can now be given to WPS_PBC
command on P2P GO to indicate that only the P2P device with the
specified P2P Device Address is allowed to connect using PBC. If
any other device tries to use PBC, a session overlap is indicated
and the negotiation is rejected with M2D. The command format for
specifying the address is "WPS_PBC p2p_dev_addr=<address>", e.g.,
WPS_PBC p2p_dev_addr=02:03:04:05:06:07
In addition, show the PBC session overlap indication as a WPS failure
event on an AP/GO interface. This particular new case shows up as
"WPS-FAIL msg=4 config_error=12".
This makes log files much more readable if multiple interfaces
are being controlled by the same process. The interface name is
added to stdout/file/syslog entries, but not to the messages
sent to control interface monitors to avoid issues with parsing
in external programs.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Add a callback for station's authorized transitions
to allow wpa_supplicant to emit events in DBus.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
To enable making state change notifications on the WLAN_STA_AUTHORIZED
flag, introduce ap_sta_set_authorized(), and to reduce use of the flag
itself also add a wrapper for testing the flag: ap_sta_is_authorized().
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
If the driver advertises max_remain_on_chan data, use it instead of
the hardcoded value of 5000. Keep the default at 5000 since that is the
value used by earlier versions of cfg80211/mac80211 and not advertised
in nl80211 for those.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
When authenticating, and the interface type is not already
NL80211_IFTYPE_STATION, we need to call wpa_driver_nl80211_set_mode()
only once. Remove the excessive call.
Signed-off-by: Eliad Peller <eliad@wizery.com>
We need to call this function after having completed the neighboring
channel scan to figure out whether HT40 can be used and if so, which
channel is the secondary channel.
ieee80211n_supported_ht_capab is called after ieee80211n_check_40mhz in
function hostapd_check_ht_capab. ieee80211n_check_40mhz can return 1 in
a valid scenario where the initialization is completed in a callback. In
this case ieee80211n_supported_ht_capab is skipped and hostapd does not
check the ht capabilities reported by the driver. Fix this issue making
sure ieee80211n_supported_ht_capab gets called.
Signed-off-by: Nishant Sarmukadam <nishants@marvell.com>
If WMM is not disabled explicitly (wmm_enabled=0 in hostapd.conf),
enable WMM automatically whenever HT (ieee80211n) is enabled. Use
the default WMM parameters for AP TX queues and the EDCA parameters
advertised for stations in WMM IE if no overriding values are
included in the configuration.
We can use the P2P interface types to check if the driver supports P2P
and to tell the kernel that a given interface is going to be used for
P2P (when it is created).
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
There is no real reason to maintain these in the current development
branch anymore. If someone really needs support for the obsolete
driver interfaces, these can be found in older wpa_supplicant
branches.
driver_atmel.c
- vendor-specific interface for ATMEL AT76C5XXx cards
- for some old out-of-tree driver; not for the upstream atmel*
drivers
driver_ndiswrapper.c
- vendor-specific interface for an out-of-tree driver
- ndiswrapper should work with driver_wext.c, too
driver_ipw.c
- vendor-specific interface for old ipw2100/2200 driver
- the upstream driver works with driver_wext.c (and does not work
with the old interface)
driver_hermes.c
- vendor driver that was not even included in the main wpa_supplicant
releases
In order to enable protection mechanisms for different HT opmodes the
driver needs to be aware of the current HT opmode that is calculated by
hostapd. Hence, pass the current opmode to the nl80211 driver via
the bss attribute NL80211_ATTR_BSS_HT_OPMODE.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
tdls_prohibit=1 and tdls_prohibit_chan_switch=1 and now be used to
disable use of TDLS or TDLS channel switching in the BSS using
extended cabilities IE as defined in IEEE 802.11z.
Make sure that received management frames are long enough before
processing them. This avoids a potential segmentation fault if a
driver delivers an invalid frame all the way to hostapd.
The changes are:
1. the word "and" in the hunting-and-pecking string passed to the KDF
should be capitalized.
2. the primebitlen used in the KDF should be a short not an int.
3. the computation of MK in hostap is based on an older version of the
draft and is not the way it's specified in the RFC.
4. the group being passed into computation of the Commit was not in
network order.
Previously, only the Configuration Error values were indicated in
WPS-FAIL events. Since those values are defined in the specification
it is not feasible to extend them for indicating other errors. Add
a new error indication value that is internal to wpa_supplicant and
hostapd to allow other errors to be indicated.
Use the new mechanism to indicate if negotiation fails because of
WEP or TKIP-only configurations being disallows by WPS 2.0.
This needs to be done both in the more normal location in
p2p_timeout_connect_listen() (internal timeout after driver event) and
in p2p_listen_end() as a workaround for the case where the driver event
is delayed to happen after the internal timeout.
Previously, both NULL and ff:ff:ff:ff:ff:ff addr were used in various
places to indicate default/broadcast keys. Make this more consistent
and useful by defining NULL to mean default key (i.e., used both for
unicast and broadcast) and ff:ff:ff:ff:ff:ff to indicate broadcast
key (i.e., used only with broadcast).
When hostapd is removing a virtual BSS interface, the loop here was
incorrectly not updating the iterator during list traversal and
ended up in an infinite loop in some cases.
Use NULL instead of (u8 *) "" as the seq value and make sure the
driver wrapper implementations can handle NULL value. This was
previously already done in number of places, but not everywhere.
wpa_supplicant seems to crash from time to time on a NetBSD 4.0 MIPS
platform. The root cause turned out to be a MIPS alignment issue.
In my wpa_supplicant crash case, in function
wpa_driver_bsd_event_receive (from driver_bsd.c), the buf[2048] address
is started from i.e. 0x7fffd546, which is not 4 bytes aligned. Later
when it is casted to (struct if_msghdr *), and rtm->rtm_flags is used.
rtm->rtm_flags is "int" type, but its address is not 4 bytes aligned.
This is because the start address of rtm is not 4 bytes aligned.
Unfortunately in NetBSD MIPS kernel (unlike Linux MIPS kernel emulates
unaligned access in its exception handler), the default behavior is to
generate a memory fault to the application that accesses unaligned
memory address. Thus comes the early mentioned wpa_supplicant crash. An
interesting note is when I'm using the wpa_supplicant version 0.4.9, I
never saw this problem. Maybe the stack layout is different. But I
didn't look into details.
I used below patch to resolve this problem. Now it runs correctly for at
least several hours. But you might have a better fix (maybe we can use
malloc/free so that it is at least cache line aligned?). I'm also not
sure if other drivers should have the same problem.
This adds partial callbacks and events to allow P2P management to be
implemented in a driver/firmware. This is not yet complete and is
very much subject to change in the future.
CONFIG_WPS_REG_DISABLE_OPEN=y can be used to configure wpa_supplicant
to disable open networks by default when wps_reg command is used to
learn the current AP settings. When this is enabled, there will be a
WPS-OPEN-NETWORK ctrl_iface event and the user will need to explicitly
enable the network (e.g., with "select_network <id>") to connect to
the open network.
If the underlying driver supports off-channel TX, it will now be used by
the nl80211 driver wrapper, setting WPA_DRIVER_FLAGS_OFFCHANNEL_TX
accordingly.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
With the new kernel functionality coming to Linux to allow off-channel
TX, we can take advantage of that in the P2P code that currently uses
remain-on-channel. If a driver advertises support for it, it will be
asked to handle off-channel TX by itself.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The nl80211 driver can report low ACK condition (in fact it reports
complete loss right now only). Use that, along with a config option, to
disconnect stations when the data connection is not working properly,
e.g., due to the STA having went outside the range of the AP. This is
disabled by default and can be enabled with disassoc_low_ack=1 in
hostapd or wpa_supplicant configuration file.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
ClientTimeout changes from EAP peer methods were not supposed to
change behavior for other EAP peer methods or even other sessions
of the same method. Re-initialize ClientTimeout whenever an EAP
peer method is initialized to avoid this. This addresses problems
where WPS (EAP-WSC) reduces the timeout and consecutive EAP runs
may fail due to too small timeout.
If the peer you want to connect to is no longer available (does not
acknowledge frames) when wpa_supplicant sends GO Negotition Request
frames, retransmission of this frame is done until the associated
p2p_device structure is removed on timeout. In that case, no signal
is emitted to inform the GO Negotiation has failed.
When sending an Invitation Request frame, the same retransmission
mechanism is in place but limit the transmission to 100 and hitting
the limit generates an event.
This patch adds the same mechanism as the one in place for Invitation
Request, but with limit of 120 to match the existing wait_count for
for GO Negotiation.
If a station received unprotected Deauthentication or Disassociation
frame with reason code 6 or 7 from the current AP, there may be a
mismatch in association state between the AP and STA. Verify whether
this is the case by using SA Query procedure. If not response is
received from the AP, deauthenticate.
This implementation is only for user space SME with
driver_nl80211.c.
Some new code will require access to P2P group members, so add API to
retrieve the number of members and iterate the members themselves.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
If there is a pending GO Negotiation when p2p_cancel is used,
unauthorize the peer to avoid immediate reconnection from being
accepted without a new p2p_connect command.
This can be used by external programs (e.g., wlantest_cli) to inject
raw frames (hex dump of the frame header and body). The data can be
requested to be sent as-is or protected with the current key.
Previously, pairwise and group cipher suites were configured only
when kernel SME (nl80211 connect API) was used. However, mac80211
needs this information even in the user space SME case for one
thing: to disable HT when TKIP/WEP is used. Add
NL80211_ATTR_CIPHER_SUITES_PAIRWISE to fix this special case with
user space SME. This allows mac80211 to disable HT properly when
the AP is configured with configuration that is not allowed.
This add preliminary code for setting the per-STA RX GTK for
RSN IBSS when nl80211 drivers. For some reason, this does not
seem to fully work, but at least driver_nl80211.c is now aware of
what kind of key is being set and the whatever is missing from
making this key configuration go through should be specific to
nl80211/cfg80211.
The frame needs to be sent from an individual (non-group) address,
so drop invalid frames before sending Deauth/Disassoc frames to
not associated STAs.
One of the pointers to the PAC buffer was not updated after realloc
and if the realloc ended up returning new pointer, the *pos pointer
was still pointing at the old location (i.e., freed memory at
this point).
When controlling multiple virtual interfaces on the same physical
radio, share the scan results events with sibling interfaces. This
decreases the time it takes to connect many virtual interfaces.
This is currently only supported on Linux with cfg80211-based
drivers when using nl80211 or wext driver interface.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Some drivers are not providing exactly reliable error codes (e.g.,
with WEXT), but others may actually indicate reliable information.
Allow driver wrappers to indicate if that is the case and use
optimizations if so. For now, this improves nl80211 with
NL80211_CMD_CONNECT for a case where connection request fails.
driver_param=use_p2p_group_interface=1 can now be used to test
nl80211-drivers with separate P2P group interface. In other words,
the main interface (e.g., wlan0) is reserved for P2P management
operations and non-P2P connections and a new group interface (e.g.,
p2p-wlan0-0) is created for the P2P group.
This implementation is very minimal, i.e., it only support address
allocation for a single P2P group interface (if the driver does not
handle this internally). In addition, not all functionality has yet
been tested, so for now, this is disabled by default and needs that
special driver_param to enable.
WPA_DRIVER_FLAGS_P2P_MGMT_AND_NON_P2P flag can now be used to
indicate that the initial interface (e.g., wlan0) is used for
P2P management operations and potentially non-P2P connections.
This is otherwise identical to
WPA_DRIVER_FLAGS_P2P_DEDICATED_INTERFACE, but the possibility of
non-P2P connections makes some operations differ.
Getting rid of these inline functions seems to reduce the code size
quite a bit, so convert the most commonly used hostapd driver ops to
function calls.
This is not needed anymore and just makes things more difficult
to understand, so move the remaining function pointers to direct
function calls and get rid of the struct hostapd_driver_ops.
send_eapol, set_key, read_sta_data, sta_clear_stats,
set_radius_acl_auth, set_radius_acl_expire, and set_beacon
to use inline functions instead of extra abstraction.
Commit bf65bc638f started the path to
add this new abstraction for driver operations in AP mode to allow
wpa_supplicant to control AP mode operations. At that point, the
extra abstraction was needed, but it is not needed anymore since
hostapd and wpa_supplicant share the same struct wpa_driver_ops.
Start removing the unneeded abstraction by converting
send_mgmt_frame() to an inline function, hostapd_drv_send_mlme().
This is similar to the design that is used in wpa_supplicant and
that was used in hostapd in the past (hostapd_send_mgmt_frame()
inline function).
driver.h defines these functions to return 0 on success, not
number of bytes transmitted. Most callers are checking "< 0" for
error condition, but not all. Address this by following the driver
API specification on 0 meaning success.
On Linux, verify that the kernel entropy pool is capable of providing
strong random data before allowing WPA/WPA2 connection to be
established. If 20 bytes of data cannot be read from /dev/random,
force first two 4-way handshakes to fail while collecting entropy
into the internal pool in hostapd. After that, give up on /dev/random
and allow the AP to function based on the combination of /dev/urandom
and whatever data has been collected into the internal entropy pool.
wlan0: RADIUS No authentication server configured
MEMLEAK[0x999feb8]: len 1040
WPA_TRACE: memleak - START
[3]: ./hostapd(radius_msg_new+0x33) [0x8074f43]
radius_msg_new() ../src/radius/radius.c:117
[4]: ./hostapd() [0x806095e]
ieee802_1x_encapsulate_radius() ../src/ap/ieee802_1x.c:439
ieee802_1x_aaa_send() ../src/ap/ieee802_1x.c:1496
For example, this error occured when I used WPS hostapd without
"eap_server=1" definition in configuration file.
By default, make hostapd and wpa_supplicant maintain an internal
entropy pool that is fed with following information:
hostapd:
- Probe Request frames (timing, RSSI)
- Association events (timing)
- SNonce from Supplicants
wpa_supplicant:
- Scan results (timing, signal/noise)
- Association events (timing)
The internal pool is used to augment the random numbers generated
with the OS mechanism (os_get_random()). While the internal
implementation is not expected to be very strong due to limited
amount of generic (non-platform specific) information to feed the
pool, this may strengthen key derivation on some devices that are
not configured to provide strong random numbers through
os_get_random() (e.g., /dev/urandom on Linux/BSD).
This new mechanism is not supposed to replace proper OS provided
random number generation mechanism. The OS mechanism needs to be
initialized properly (e.g., hw random number generator,
maintaining entropy pool over reboots, etc.) for any of the
security assumptions to hold.
If the os_get_random() is known to provide strong ramdom data (e.g., on
Linux/BSD, the board in question is known to have reliable source of
random data from /dev/urandom), the internal hostapd random pool can be
disabled. This will save some in binary size and CPU use. However, this
should only be considered for builds that are known to be used on
devices that meet the requirements described above. The internal pool
is disabled by adding CONFIG_NO_RANDOM_POOL=y to the .config file.
This commit adds a new wrapper, random_get_bytes(), that is currently
defined to use os_get_random() as is. The places using
random_get_bytes() depend on the returned value being strong random
number, i.e., something that is infeasible for external device to
figure out. These values are used either directly as a key or as
nonces/challenges that are used as input for key derivation or
authentication.
The remaining direct uses of os_get_random() do not need as strong
random numbers to function correctly.
This adds more time for the system entropy pool to be filled before
requesting random data for generating the WPA/WPA2 encryption keys.
This can be helpful especially on embedded devices that do not have
hardware random number generator and may lack good sources of
randomness especially early in the bootup sequence when hostapd is
likely to be started.
GMK and Key Counter are still initialized once in the beginning to
match the RSN Authenticator state machine behavior and to make sure
that the driver does not transmit broadcast frames unencrypted.
However, both GMK (and GTK derived from it) and Key Counter will be
re-initialized when the first station connects and is about to
enter 4-way handshake.
The example GMK-to-GTK derivation described in the IEEE 802.11 standard
is marked informative and there is no protocol reason for following it
since this derivation is done only on the AP/Authenticator and does not
need to match with the Supplicant. Mix in more data into the derivation
process to get more separation from GMK.
The previous commit broke completion in various places. The proper
way of handling the completion of full word is to verify whether
there are more than one possible match at that point.
Completion needs to be done even if the full word has been entered.
In addition, fix the space-after-full-word to properly allocate room
for the extra character when completion is used in the middle of the
string.
This adds new commands for wlantest_cli to request wlantest to
inject frames. This version can only send out Authentication
frames and unprotected SA Query Request frames, but there is
now place to add more frames and encryption with future commits.
ap_setup_locked=2 can now be used to enable a special mode where
WPS ER can learn the current AP settings, but cannot change then.
In other words, the protocol is allowed to continue past M2, but
is stopped at M7 when AP is in this mode. WPS IE does not
advertise AP Setup Locked in this case to avoid interoperability
issues.
In wpa_supplicant, use ap_setup_locked=2 by default. Since the AP PIN
is disabled by default, this does not enable any new functionality
automatically. To allow the read-only ER to go through the protocol,
wps_ap_pin command needs to be used to enable the AP PIN.
There is no point in maintaining two almost identical versions
of this parser. Move WPA IE parser into wpa_common.c similarly
to what was already the case with RSN IE parse.
When multiple wireless interfaces are used with WPS, the UPnP
subscriptions need to be removed whenever a matching Registrar
instance gets removed. This avoids a segfault due to access to
freed memory during hostapd shutdown.
In addition, the UPnP interface instance structure needs to be
freed to avoid memory leak.
WPS external Registrars can get confused about multiple UPnP
instances (one per radio) on a dual-concurrent APs. Simplify the
design by sharing a single UPnP state machine for all wireless
interfaces controlled by hostapd. This matches with the previous
changes that made a single command enable WPS functionality on
all interfaces.
This is relatively minimal change to address the sharing of the
state among multiple struct hostapd_data instances. More cleanup
can be done separately to remove unnecessary copies of information.
When the peer does not include our initial preference in the Channel
List attribute during GO Negotiation, try to use the best channel of
the other band as the new preference instead of falling back to the
first channel in the intersection.
The driver wrapper may now indicate the preferred channel (e.g., based
on scan results) on both 2.4 GHz and 5 GHz bands (and an overall best
frequency). When setting up a GO, this preference information is used
to select the operating channel if configuration does not include
hardcoded channel. Similarly, this information can be used during
GO Negotiation to indicate preference for a specific channel based
on current channel conditions.
p2p_group_add command can now use special values (freq=2 and freq=5)
to indicate that the GO is to be started on the specified band.
Commit d8d940b746 introduced a regression
that prevented TSN APs from being used with WEP since the AP was
rejected if it advertised WPA or RSN IE when we were configured to use
WEP. Resolve this by checking whether the AP is advertising a TSN, i.e.,
whether the AP allows WEP to be used as a group cipher. If so, allow
the AP to be selected if we are configured to use static WEP or
IEEE 802.1X (non-WPA).
It should be noted that this is still somewhat more restricted in AP
selection than earlier wpa_supplicant branches (0.7.x or older) that
ignore the WPA/RSN IE completely when configured for non-WPA.
This tool can be used to capture IEEE 802.11 frames either from a
monitor interface for realtime capturing or from pcap files for
offline analysis. This version is only adding basic infrastructure for
going through the frames and parsing their headers.
Commit c3fc47ea8e fixed EAP passthrough
server to allow Logoff/Re-authentication to be used. However, it
broke EAP standalone server while doing that. Fix this by reverting
the earlier fix and by clearing the EAP Identity information in the
EAP server code whenever an EAPOL-Start or EAPOL-Logoff packet is
received.
We need to drop the pending frame to avoid issues with the new GO
Negotiation, e.g., when the pending frame was from a previous attempt at
starting a GO Negotiation.
These are not used by any driver wrapper, i.e., only the four
data queues (BK, BE, VI, VO) are configurable. Better remove these
so that there is no confusion about being able to configure
something additional.
Pass data frames from unknown STAs to hostapd in order to reply with
a Deauthentication or Disassociation frame. This fixes compliance
with IEEE Std 802.11-2007, 11.3.
Furthermore, this does not cause a lot of overhead (at least with
mac80211 drivers) since mac80211 does not pass all data frames (but
at least from unauthenticated and unassociated STAs) to cooked monitor
interfaces.
Tested with rt2800pci on a MIPS board.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
drivers.mak is shared between hostapd and wpa_supplicant and
compiling them together may create "wrong" flags.
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
We are not actually interested in the PD Response in join-a-group
case, so there is no point in trying to send PD Request until the
response is received. This avoids an extra PD getting started after
a join-a-group operation in some cases.
This is needed to avoid problems with other applications setting and
leaving the interface to IF_OPER_DORMANT state. In AP mode, the interface
is ready immediately after the keys are set, so we better make sure the
DORMANT state does not prevent normal operations after that.
This is needed to avoid issues with APs that restart their UPnP,
e.g., when ER reconfigures them. The previously known settings are
now cached and taken into use if an AP is detected to leave
(ssdp:byebye) and then return.
Verify that the UUID given to wps_er_pbc command is known and only
activate PBC mode on the matching AP. The UUID can be that of the
AP or the station/Enrollee.
Add a new driver event, EVENT_INTERFACE_UNAVAILABLE, for indicating
that the driver is not able to continue operating the virtual
interface in its current mode anymore, e.g., due to operating
channel for GO interface forced to a DFS channel by another virtual
interface.
When this happens for a P2P group interface, the P2P group will
be terminated and P2P-GROUP-REMOVED event shows the reason for
this as follows:
P2P-GROUP-REMOVED wlan0 GO reason=UNAVAILABLE
A new configuration parameter, p2p_group_idle, can now be used to set
idle timeout value for P2P groups in seconds (0 = no timeout). If set,
this values is used to remove P2P group (both GO and P2P client)
interfaces after the group has been idle (no clients/GO seen) for the
configuration duration.
The P2P-GROUP-REMOVED event is now indicating the reason for group
removal when known. For example:
P2P-GROUP-REMOVED wlan0 GO reason=REQUESTED
P2P-GROUP-REMOVED wlan1 client reason=IDLE
Some driver wrappers may end up indicating a disassociation or
deauthentication event without the address of the station, e.g.,
based on a previous non-AP mode event. Avoid crashing hostapd
by verifying that the require address parameter is available in
the event before processing it.
In order to avoid confusing the driver with a new remain-on-channel
request, delay sending of a new Action frame if the driver indicates
Action frame RX diromg a scan.
This is needed to optimize response to GO Negotiation Request frames.
The extra remain-on-channel cancel followed by new remain-on-channel for
the same channel takes too much time with some driver/firmware
designs for the response to go out quickly enough to avoid peer
timing out while waiting for our response.
This command can be used to enroll a network based on a local
network configuration block instead of having to (re-)learn the
current AP settings with wps_er_learn.
This test command is supposed to change the WPS version number in all
places immediately, so make sure that the IEs used in management
frames get updated immediately.
The P2P group component is now tracking of associated stations
in the group and the Group Limit bit in the Group Capabilities
is updated based on whether there is room for new clients in
the group.
The workaround to ignore no ctrl::ack received for GO Negotiation
Confirmation frame was only supposed to be used when the frame was
actually transmitted and just the ack was not received. However, due
to the way the driver failure on transmitting the frame were reported,
this ended up getting applied for all failures in sending the GO
Negotiation Confirmation frame.
Improve this by providing a mechanism to indicate whether send_action
operations fail locally before the frame was actually transmitted or
because of not receiving ack frame after having transmitted the frame.
UPnP event subscriptions are not supposed to be dropped based on
whether events can be delivered quickly enough. Leave dropping to
happen only based on failed deliveries to avoid issues with a burst
of events kicking out still active subscribers.
Do not send more than five Probe Request WLANEvent notifications
per second. Even though the limit should in theory apply to all
WLANEvents, it is better not to drop EAP notifications because
of Probe Request frames and really, the only real reason for
event bursts is Probe Request frames.
Instead of queuing all events for a subscriber, only queue more
important events if delivery of event notifications starts failing.
This allows more time for the subscriber to recover since the maximum
queue length if not reached because of Probe Request frames only.
