Since the 6 GHz band has no DFS channels, enable 6 GHz 160 MHz bandwidth
as the default configuration for IEEE 802.11s mesh.
example:
network={
ssid="6gmesh160"
key_mgmt=SAE
mode=5
frequency=6275
psk="1234567890"
}
Signed-off-by: P Praneesh <ppranees@codeaurora.org>
Signed-off-by: Pradeep Kumar Chitrapu <pradeepc@codeaurora.org>
Channel numbers of the 6 GHz band overlap those of the 2.4 GHz and 5 GHz
bands. Thus converting to frequency based mesh channel selection helps
accommodate 6 GHz mesh.
Signed-off-by: Pradeep Kumar Chitrapu <pradeepc@codeaurora.org>
Some distros carry patches to specify driver fallback, but only in
specific conditions (e.g. the systemd service definition[1]). This leaves
other wpa_supplicant instances needing to define fallback themselves,
which leads to places where wpa_supplicant thinks it can't find a
driver[2]. Instead, when -D is not specified, have wpa_supplicant try
all the drivers it was built with in an attempt to find a working one
instead of just giving up if the first doesn't work.
[1] https://salsa.debian.org/debian/wpa/-/blob/debian/unstable/debian/patches/networkd-driver-fallback.patch
[2] https://bugs.launchpad.net/netplan/+bug/1814012
Signed-off-by: Kees Cook <kees@ubuntu.com>
After roaming from WPA2-AP (group=CCMP) to WPA-AP (group=TKIP) using
driver-based SME and roaming trigger, GTK renewal failures are observed
for the currently associated WPA-AP because of group cipher mismatch,
resulting in deauthentication with the AP.
Update the group cipher and pairwise cipher values in wpa_sm from
association event received from the driver in case of SME offload to the
driver to address GTK renewal failures (and similar issues) that could
happen when the driver/firmware roams between APs with different
security profiles.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Add an explicit check for modes != NULL instead of depending on
num_modes > 0 implying that. This is to silence invalid static analyzer
reports.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Indicate DSCP Policy capability by including a WFA Capabilities element
containing the relevant bit set to 1 in the (Re)Association Request
frames when enabled by user.
Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
Add support to parse received DSCP Policy Request frames and send the
request details as control interface events.
Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
This function is not specific to GAS, so make it available throughout
wpa_supplicant without requiring CONFIG_GAS.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
"SET disable_mscs_support 1" can be used to disable indication of MSCS
support in the Extended Capabilities element for testing purposes. This
is also disabling addition of the MSCS element even if valid
configuration parameters had been configured.
Signed-off-by: Vinita S. Maloo <vmaloo@codeaurora.org>
Add support to receive and process SCS Response frames from the AP and
indicate the status to upper layers.
Signed-off-by: Vinita S. Maloo <vmaloo@codeaurora.org>
"SET disable_scs_support 1" can be used to disable indication of SCS
support in the Extended Capabilities element for testing purposes.
Signed-off-by: Vinita S. Maloo <vmaloo@codeaurora.org>
Add support to parse SCS control interface command and form the SCS
Request frame to be sent to SCS enabled AP.
Signed-off-by: Vinita S. Maloo <vmaloo@codeaurora.org>
P2P connections in the 6 GHz band use SAE authentication algorithm after
getting credentials with WPS connection. During WPS connection as it
doesn't use SAE, SAE PT is not derived. After getting SAE credentials,
the STA connects to the same SSID using SAE auth algorithm. Earlier, SAE
H2E PT was not derived while connecting to the same SSID to which the
STA is connected last time. Due to this, the P2P group formation fails
for 6 GHz channels when H2E is enabled as the PT will not be setup by
the P2P client before proceeding to the SAE authentication. Same could
happen with infrastructure WPS when wps_cred_add_sae=1 is used.
Set up the SAE H2E PT while connecting to the same SSID again also to
make sure that the H2E PT is set up in the STA to derive the PWE for
successful SAE authentication. The PT derivation will be skipped in
wpa_s_setup_sae_pt() if PT is already available for that SSID.
Signed-off-by: Sreeramya Soratkal <ssramya@codeaurora.org>
External applications can store PMKSA entries persistently and
reconfigure them to wpa_supplicant after restart. This can result in
wpa_supplicant having a PMKSA for FILS authentication without having
matching ERP keys for it which would prevent the previously added
mechanism for dropping FILS PMKSA entries to recover from rejected
association attempts.
Fix this by clearing PMKSA entries configured by external applications
upon FILS connection failure even when ERP keys are not available.
Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
wpa_supplicant generates both a PMKSA cache entry and ERP keys upon
successful FILS connection and uses FILS authentication algorithm for
subsequent connections when either ERP keys or a PMKSA cache entry is
available.
In some cases, like AP/RADIUS server restart, both ERP keys and PMKSA
becomes invalid. But currently when an AP rejects an association,
wpa_supplicant marks only ERP keys as failed but not clearing PMKSA.
Since PMKSA is not cleared, consecutive connection attempts are still
happening with FILS authentication algorithm and connection attempts are
failing with the same association rejection again instead of trying to
recover from the state mismatch by deriving a new ERP key hierarchy.
Clear PMKSA entries as well on association rejection from an AP to allow
the following connection attempt to go with open authentication to
re-establish a valid ERP key hierarchy. Also, since clearing PMKSA
entries on unprotected (Re)Association Response frames could allow DoS
attack (reduce usability of PMKSA caching), clear PMKSA entries only
when ERP keys exists.
Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
It is safer to remove and free these entries with a shared helper
function to avoid issues with potentially forgetting to unregister or
free something if this structure is extended in the future.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
When a STA makes an association request that is rejected by an OCE AP
due to the RSSI being insufficient, the AP is added to the driver
disallow list by wpa_set_driver_tmp_disallow_list().
Once the AP increases TX power which makes the AP RSSI higher than
Association Rejection RSSI threshold, the AP is supposed to be removed
from the driver disallow list but that was not the case.
wpa_is_bss_tmp_disallowed() is called in the scan result handler, so it
is the best place to put the logic of removing the AP from the driver
disallow list with sufficient AP RSSI.
This is needed with drivers that use the temporarily disallowed BSS list
(which is currently supported only with a QCA vendor command). The
wpa_supplicant internal functionality was already taking care of this
with the wpa_is_bss_tmp_disallowed() return value even for cases where
the entry remaining in the list.
Signed-off-by: Hu Wang <huw@codeaurora.org>
Add checks for features supported by the specific hardware mode of the
local device that has the channel for which the throughput is being
estimated instead of assuming the local device supports all optional
features. This is more accurate for cases where the local capabilities
might differ based on the band. In addition, this is in preparation for
extending rate estimates to cover optional VHT and HE features.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
wpa_supplicant_ctrl_iface_deinit() was executed only if the
per-interface control interface initialization had been completed. This
is not the case if driver initialization fails and that could result in
leaving behind references to the freed wpa_s instance in a corner case
where control interface messages ended up getting queued.
Fix this by calling wpa_supplicant_ctrl_iface_deinit() in all cases to
cancel the potential eloop timeout for wpas_ctrl_msg_queue_timeout with
the reference to the wpa_s pointer. In addition, flush any pending
message from the global queue for this interface since such a message
cannot be of use after this and there is no need to leave them in the
queue until the global control interface gets deinitialized.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
An EAPOL frame may be pending when wpa_supplicant requests to
deauthenticate. At this stage the EAP SM cache is already cleaned by
calling eapol_sm_invalidate_cached_session(). Since at this stage the
wpa_supplicant's state is still set to associated, the EAPOL frame is
processed and results in a crash due to NULL dereference.
This wasn't seen previously as nl80211 wouldn't process the
NL80211_CMD_CONTROL_PORT_FRAME, since wpa_driver_nl80211_mlme() would
set the valid_handler to NULL. This behavior was changed in commit
ab89291928 exposing this race.
Fix it by ignoring EAPOL frames while the deauthentication is in
progress.
Fixes: ab89291928 ("nl80211: Use process_bss_event() for the nl_connect handler")
Signed-off-by: Andrei Otcheretianski <andrei.otcheretianski@intel.com>
The wpa_ie buffer is now allocated here and needs to be freed before
returning from the function.
Fixes: d2ba0d719e ("Move assoc param setting into a helper function")
Signed-off-by: Jouni Malinen <j@w1.fi>
Use a shared code path for freeing the wpa_ie buffer to avoid
unnecessary complexity with a separate return for the non-FILS case.
Signed-off-by: Jouni Malinen <j@w1.fi>
When the driver SME is used, offloaded RSN handshakes like SA Query, GTK
rekeying, FT authentication, etc. would fail if wpa_supplicant enables
OCV in initial connection based on configuration but the driver doesn't
support OCV. To avoid such failures check the driver's capability for
enabling OCV when the driver SME used.
This commit also adds a capability flag for indicating OCV support
by the driver.
Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
There's a chance that prior to config reload being requested a scan work
was started. As such forcing wpa_supplicant to WPA_DISCONNECTED was
removing any hints that the actual driver is busy with work. That led to
wpa_supplicant reporting "Failed to initialize AP scan" over and over
again for a few seconds (depending on driver/capabilities) until the
untracked scan finished.
Cancelling a scan isn't really a solution because there's a bunch of
scanning state bits sprinkled across wpa_supplicant structure and they
get updated as driver events actually flow in in async manner.
As far as I can tell this is only preventing unnecessary warning
messages. This doesn't seem like it was crippling any logic per se.
Signed-off-by: Michal Kazior <michal@plume.com>
Previously, the channel number was set in hostapd_freq_params only with
the presence of HT capabilities. Set the channel number before the check
for HT mode to accommodate the 6 GHz band cases.
Signed-off-by: Pradeep Kumar Chitrapu <pradeepc@codeaurora.org>
Currently, the external_scan_running flag is not reset when an interface
is removed. Thus, if a connection attempt is made on another iface, it
will fail due to wpa_supplicant incorrectly assuming the radio is still
busy due to the ongoing scan.
To fix this, convert external_scan_running to a pointer to the interface
that started the scan. If this interface is removed, also reset the
pointer to NULL so that other operations may continue on this radio.
Test:
1. Start scan on wlan0
2. Remove wlan0
3. Can connect to a network on wlan1
Signed-off-by: David Su <dysu@google.com>
Add PASN implementation to wpa_supplicant
1. Add functions to initialize and clear PASN data.
2. Add functions to construct PASN Authentication frames.
3. Add function to process PASN Authentication frame.
4. Add function to handle PASN frame TX status.
5. Implement the station side flow processing for PASN.
The implementation is missing support for wrapped data and PMKSA
establishment for base AKMs, and only supports PASN authentication or
base AKM with PMKSA caching.
The missing parts will be added in later patches.
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
PASN requires to store the PTK derived during PASN authentication
so it can later be used for secure LTF etc. This is also true
for a PTK derived during regular connection.
Add an instance of a PTKSA cache for each wpa_supplicant
interface when PASN is enabled in build configuration.
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Commit 02c21c02d0 ("wpa_supplicant: Do not disconnect on deinit if
WoWLAN is enabled") prevents the disconnection on deinit if the driver
indicates that WoWLAN is enabled. This is not the expected behavior in
some earlier use cases where the wpa_supplicant process is left running
when going to sleep and killing of the wpa_supplicant process is used
only when there is an expectation of Wi-Fi connection being disabled.
To support the use cases which require the WLAN to disconnect on deinit
even if WoWLAN is enabled, introduce a configuration parameter
wowlan_disconnect_on_deinit. This is set to 0 by default thereby not
impacting the functionality in the above mentioned commit. Setting it to
1 restores the old behavior before the commit identified above.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
When STA is performing roam from WPA3 AP to WPA2 AP, the STA was
including key mgmt FT-SAE instead of FT-PSK in FT Authentication request
RSNE when using driver-based SME. This is because the RSNE/MDE/FTE were
updated and forwarded to the driver using the NL80211_CMD_UPDATE_FT_IES
command before updating key mgmt properly. Because of this, the AP is
rejecting FT Authentication request with WLAN_REASON_UNSPECIFIED reason
code which is due to the invalid keymgmt in RSNE.
Fix this by reordering IE population to happen earlier in the sequence
so that the updated key mgmt information can be provided when using
NL80211_CMD_UPDATE_FT_IES.
Signed-off-by: Shiva Sankar Gajula <sgajula@codeaurora.org>
Replace the implicit boolean checks that used int variables with use of
a more explicit bool variable type.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Support possible band combinations of 2.4 GHz, 5 GHz, and 6 GHz with
QCA_WLAN_VENDOR_ATTR_SETBAND_MASK attribute. Ensure backwards
compatibility with old drivers that are using
QCA_WLAN_VENDOR_ATTR_SETBAND_VALUE attribute and supporting only 2.4 GHz
and 5 GHz bands.
Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
Send mesh group started notification after join completion
callback is called.
Implement outstanding TODO, to leave the mesh network on deinit.
Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
This makes it easier to change the internal struct wpa_bss design for
storing the variable length IE buffers.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This removes need from the callers to know the struct wpa_bss details
for the location of the memory area for storing the IEs.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Use NL80211_ATTR_SAE_PWE attribute to indicate the sae_pwe value
to the driver during the NL80211_CMD_START_AP and NL80211_CMD_CONNECT
in WPA3-Personal networks which are using SAE authentication.
Signed-off-by: Rohan Dutta <drohan@codeaurora.org>
Commit e8b85c078e ("iface match: Unspecified matched interfaces should
not log driver fails") removed the only use of the added interface wpa_s
pointer, but left that pointer setting in place. Remove it to keep
static analyzers happy.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The D-Bus implementation of RemoveAllNetworks differs wildly from the
CLI implementation. Let's share the implementations.
This resolves use-after-free bugs I noticed, where we continue to use
the 'wpa_s->current_ssid' wpa_ssid object after freeing it, because we
didn't bother to disconnect from (and set to NULL) current_ssid before
freeing it.
Signed-off-by: Brian Norris <briannorris@chromium.org>
If the stored configurations for an SSID have changed, we can no longer
trust the current blacklist state of that SSID, since the updated
configs could change the behavior of the network. E.g., the BSS could be
blacklisted due to a bad password, and the config could be updated to
store the correct password. In this case, keeping the BSS in the
blacklist will prevent the user from connecting to the BSS after the
correct password has been updated.
Add the value was_changed_recently to the wpa_ssid struct. Update this
value every time a config is changed through wpa_set_config(). Check
this value in wpa_blacklist_get() to clear the blacklist whenever the
configs of current_ssid have changed.
This solution was chosen over simply clearing the blacklist whenever
configs change because the user should be able to change configs on an
inactive SSID without affecting the blacklist for the currently active
SSID. This way, the blacklist won't be cleared until the user attempts
to connect to the inactive network again. Furthermore, the blacklist is
stored per-BSSID while configs are stored per-SSID, so we don't have the
option to just clear out certain blacklist entries that would be
affected by the configs.
Finally, the function wpa_supplicant_reload_configuration() causes the
configs to be reloaded from scratch, so after a call to this function
all bets are off as to the relevance of our current blacklist state.
Thus, we clear the entire blacklist within this function.
Signed-off-by: Kevin Lund <kglund@google.com>
Signed-off-by: Brian Norris <briannorris@chromium.org>
wpa_supplicant keeps a blacklist of BSSs in order to prevent repeated
associations to problematic APs*. Currently, this blacklist is
completely cleared whenever we successfully connect to any AP. This
causes problematic behavior when in the presence of both a bad AP and
a good AP. The device can repeatedly attempt to roam to the bad AP
because it is clearing the blacklist every time it connects to the good
AP. This results in the connection constantly ping-ponging between the
APs, leaving the user stuck without connection.
Instead of clearing the blacklist, implement timeout functionality which
allows association attempts to blacklisted APs after some time has
passed. Each time a BSS would be added to the blacklist, increase the
duration of this timeout exponentially, up to a cap of 1800 seconds.
This means that the device will no longer be able to immediately attempt
to roam back to a bad AP whenever it successfully connects to any other
AP.
Other details:
The algorithm for building up the blacklist count and timeout duration
on a given AP has been designed to be minimally obtrusive. Starting with
a fresh blacklist, the device may attempt to connect to a problematic AP
no more than 6 times in any ~45 minute period. Once an AP has reached a
blacklist count >= 6, the device may attempt to connect to it no more
than once every 30 minutes. The goal of these limits is to find an
ideal balance between minimizing connection attempts to bad APs while
still trying them out occasionally to see if the problems have stopped.
The only exception to the above limits is that the blacklist is still
completely cleared whenever there are no APs available in a scan. This
means that if all nearby APs have been blacklisted, all APs will be
completely exonerated regardless of their blacklist counts or how close
their blacklist entries are to expiring. When all nearby APs have been
blacklisted we know that every nearby AP is in some way problematic.
Once we know that every AP is causing problems, it doesn't really make
sense to sort them beyond that because the blacklist count and timeout
duration don't necessarily reflect the degree to which an AP is
problematic (i.e. they can be manipulated by external factors such as
the user physically moving around). Instead, its best to restart the
blacklist and let the normal roaming algorithm take over to maximize
our chance of getting the best possible connection quality.
As stated above, the time-based blacklisting algorithm is designed to
be minimally obtrusive to user experience, so occasionally restarting
the process is not too impactful on the user.
*problematic AP: rejects new clients, frequently de-auths clients, very
poor connection quality, etc.
Signed-off-by: Kevin Lund <kglund@google.com>
Signed-off-by: Brian Norris <briannorris@chromium.org>
Within wpas_connection_failed(), the 'count' value of wpa_blacklist is
erroneously used as a tally of the number times the device has failed
to associate to a given BSSID without making a successful connection.
This is not accurate because there are a variety of ways a BSS can be
added to the blacklist beyond failed association such as interference
or deauthentication. This 'count' is lost whenever the blacklist is
cleared, so the wpa_supplicant stores an additional value
'extra_blacklist_count' which helps persist the 'count' through clears.
These count values are used to determine how long to wait to rescan
after a failed connection attempt.
While this logic was already slightly wrong, it would have been
completely broken by the upcoming change which adds time-based
blacklisting functionality. With the upcoming change, 'count' values
are not cleared on association, and thus do not necessarily even
approximate the "consecutive connection failures" which they were being
used for.
This change seeks to remove this unnecessary overloading of the
blacklist 'count' by directly tracking consecutive connection failures
within the wpa_supplicant struct, independent of the blacklist. This new
'consecutive_conn_failures' is iterated with every connection failure
and cleared when any successful connection is made. This change also
removes the now unused 'extra_blacklist_count' value.
Signed-off-by: Kevin Lund <kglund@google.com>
Signed-off-by: Brian Norris <briannorris@chromium.org>
If there is no matching interface given, but interface matching is
enabled, all interfaces on the system will try to be initialized. Non
wireless interfaces will fail and the loopback device will be one of
these, so just log a diagnostic rather than an error.
Signed-off-by: Roy Marples <roy@marples.name>
Teach wpa_supplicant to {de,}initialize bgscans when bgscan parameters
are set after initial connection.
Signed-off-by: Matthew Wang <matthewmwang@chromium.org>
