Relocate the declaration of the hostapd_find_by_sta() function to top
The hostapd_find_by_sta() function may be utilized by multiple driver
events. Therefore, it should be declared at the outset to ensure
accessibility by all event handlers.
hostapd: Add Max Channel Switch Time element support
The Max Channel Switch Time element indicates the time delta between the
time the last beacon is transmitted by the AP in the current channel and
the expected time of the first Beacon frame transmitted by the AP in the
new channel.
IEEE P802.11be/D7.0, 35.3.11 (ML procedures for (extended) channel
switching and channel quieting) indicates that if an AP affiliated with
an AP MLD is switching channel, Max Channel Switch Time element shall be
included in every Beacon and Probe Response frame it transmits.
Add support to include the Max Channel Switch Time element in Beacon and
Probe Response frames during channel switch announcement for an AP
affiliated with an AP MLD.
Share usec-to/from-TU conversion macros in a general location
These can be used outside the context of FST and hostapd.c, so move the
macro definitions into a shared header file. In addition, fix the
currently unused US_TO_TU() definition and rename "US" to "USEC".
AP MLD: Ensure successful addition of link item into list
Currently, hapd->link is added to the MLD links list during driver
initialization and setup BSS operation. However, a call trace has been
observed where a BSS link item is not present in the list and an attempt
is made to delete it from the list. This scenario occurs during the
deinitialization operation, which calls hostapd_bss_link_deinit() and
tries to remove the hapd->link which is not present in the list.
Ensures that the link item is added to the list only after the
successful operation of link addition. Also ensure that mld->num_links
increments only when the addition is successful. Therefore, return from
hostapd_bss_link_deinit(), if mld->num_links is zero. Since the mld
object is shared among all the links, num_links has to be incremented
only when the addition is successful.
SAE: Add group 20 in default groups when AP started with SAE-EXT-KEY
hostapd used to always enable only the group 19 when SAE groups were not
configured explicitly in hostapd.conf. This may cause undesired
connection delay with STAs which use group 20 as the primary group with
SAE-EXT-KEY AKM during SAE authentication attempt.
To avoid this, enable group 20 in default groups when AP supports
SAE-EXT-KEY and SAE groups have not been configured explicitly in
hostapd.conf.
nl80211: Fix simultaneous scanning failure in case of MLO
Currently only one scan can be performed on per phy level at a time in
the driver. Due to this, if another scan request is sent via anoother
underlying phy, the kernel returns -EBUSY. This would result in hostapd
trying to set the interface into station mode if it was originally in AP
mode and retry sending a scan request. However, this behavior is
expected in case of multi link operation and hence there is no need to
switch the mode as such.
Hence, add logic to not change the NL mode if the kernel returns -EBUSY
during multi link AP operation. The caller can accordingly decide and,
if needed, it can re-schedule a scan request after some time.
nl80211: Fix scan request and its related events handling with MLO
Currently, whenever a scan is started, it uses drv's first BSS only
whether it is AP or STA interface. However, with AP MLD related changes,
the same drv could be used by other BSSs as well which needs scanning.
Hence, the current logic will not work since scan needs to be handled on
a non-first BSS as well.
Move the logic of always using drv's first BSS during scan events to
using BSS on which the event arrived.
Also, for AP MLD operation, even though the BSS is same, the link BSS
also needs to be identified. Hence, add a back pointer in the BSS struct
which would be used to point to the link BSS which requested the scan on
that BSS. This will help in routing the scan events to an appropriate
BSS ctx.
Gal Savion [Wed, 1 May 2024 11:11:39 +0000 (14:11 +0300)]
Avoid sending DEAUTH or DISASSOC packet when using flag tx=0
hostapd would send DISASSOC packet (after quiet DEAUTH) or DEAUTH packet
(after quiet DISASSOC) to the station after some inactivity timeout,
even though the command has tx=0 parameter. Fix this so that tx=0 cleans
the STA info without sending any DISASSOC or DEAUTH packets.
This can be used with a clangd server to get code completion and cross
references in editor. To simplify the generation, create .cmd files for
most object files while building that contains the base directory and
command that was used when compiling it.
A very simple gen_compile_commands.py is provided which will read one or
more build directories and generate the compile_commands.json file for
it.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Matthew Wang [Tue, 4 Jun 2024 21:02:16 +0000 (21:02 +0000)]
Trigger a 6 GHz scan if RNR contains matching short SSID
If a scan triggers a regdom update into a 6 GHz-allowed regdom, and an
RNR element in one of the legacy band scan results points to a 6 GHz
scan result with a short SSID matching the current_ssid, delay
connection in favor of a 6 GHz-only scan. This will optimize the case in
which we first connect to a 5 GHz AP, then later roam to a 6 GHz one by
directly connecting to the 6 GHz one.
Signed-off-by: Matthew Wang <matthewmwang@chromium.org>
Matthew Wang [Tue, 4 Jun 2024 21:02:16 +0000 (21:02 +0000)]
Fix success check for triggering 6 GHz scan
wpas_trigger_6ghz_scan() returns 1 on success and 0 on failure to
trigger the scan. Checking the return value to be less than zero is not
correct. This was supposed to check for success, i.e., greater than
zero.
Fixes: 42add3c27ba0 ("Scan 6 GHz channels after change to 6 GHz-allowed regdom") Signed-off-by: Matthew Wang <matthewmwang@chromium.org>
Jintao Lin [Wed, 31 Jul 2024 18:50:01 +0000 (18:50 +0000)]
P2P: Use specified frequency for group client in P2P_GROUP_ADD command
In P2P_GROUP_ADD command and GroupAdd dbus method, frequency is passed
in as a parameter when restarting a persistent group. This is the group
operating frequency determined out of band. Use this pre-determined
frequency in P2P client as well to expedite the P2P scan.
Signed-off-by: Jintao Lin <jintaolin@chromium.org>
Ruth Mekonnen [Wed, 12 Jun 2024 20:09:43 +0000 (20:09 +0000)]
dbus: Emit ScanInProgress6GHz property
Expose whether a 6 GHz scan is in progress with the ScanInProgress6GHz
property and flush properties as soon as the property is updated, so
that platforms can choose not to disconnect while a 6 GHz scan is in
progress. Once the 6 GHz scan has completed and scan results have been
received, the ScanInProgress6GHz property is reset to false.
When non_coloc_6ghz = false, the STA is expected to scan for colocated
APs. However, if the colocated AP is on a non-PSC channel, it will not
be detected during the 6 GHz-only scan because the frequency list is
limited to PSC channels. Even when the NL80211_SCAN_FLAG_COLOCATED_6GHZ
is set, the cfg80211 only scans a subset of the channels in the original
6 GHz scan request. Therefore, this patch adds non-PSC channels to the
original 6 GHz scan request.
Benjamin Berg [Mon, 29 Apr 2024 11:51:57 +0000 (13:51 +0200)]
WNM: Scan for BSSID if there are forbidden neighbors
The test to scan for a single BSSID assumed that there is only a single
neighbor in the candidate list. Also do this optimization if there are
multiple neighbors but only one of them is valid.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:55 +0000 (13:51 +0200)]
WNM: Always parse candidate list
The preferred candidate list included bit just makes parsing the list
mandatory. The AP may still include a candidate list which we should use
to be able to optimize scanning. As such, always parse out the list but
still verify that if the list is not empty if the bit is set.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:54 +0000 (13:51 +0200)]
WNM: Consolidate the scanning paths for BTM requests
There was an ancient code path to trigger a scan that was apparently
forgotten when the code was extended over time. It does not make any
sense to trigger a scan twice, so remove the earlier scan.
The earlier scan call was avoiding to trigger a new scan if a fixed
BSSID is configured. This seems like a reasonable restriction to do, so
add this check before starting a scan.
Consolidate everything so that scanning happens at the end of the
functions unless we bail out before. Add a "reset" label for all other
cases to ensure that we don't leave things in the a bad state.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
MBO: Always accept BTM request with disassociation imminent bit set
According to Multiband Operation specification (r17, section 3.5.2),
a BSS Transition Management Request with the disassociation imminent
bit set should always be accepted.
This is enforced in case the request did not include a candidate list.
However, in case a candidate list was included but none of the APs in
the candidate list was found in the scan results, the request is
rejected.
Fix that by always accepting a request with the disassociation imminent
bit set even if no roaming candidate was found.
Signed-off-by: Avraham Stern <avraham.stern@intel.com> Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:50 +0000 (13:51 +0200)]
tests: Set the abridged bit in BTM tests
Many of the WNM tests assume that the supplicant will only select a
neighbor that was included in the candidate list. However, without the
abridged bit being set, the supplicant is allowed to also select any BSS
that is not explicitly included in the list.
Prepare for the supplicant handling the abridged bit by setting it in
the tests where relevant.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:48 +0000 (13:51 +0200)]
WNM: Move neighbor report test into wnm_is_bss_excluded()
Having it in wnm_is_bss_excluded() is more generic as it works for other
locations (e.g., MLD link selection). So move the test and add a check
for the abridged bit while at it. Note that without the abridged bit
check another check would be needed (e.g., checking wnm_dialog_token) to
ensure that there isn't a rejection unless a BTM is in progress.
compare_scan_neighbor_results() calls wpa_scan_res_match() which calls
wnm_is_bss_excluded() so the previous behavior is maintained for WNM
scan result processing.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:47 +0000 (13:51 +0200)]
WNM: Swap logic in wnm_is_bss_excluded() to allow more checks
Following commits will move more checks into wnm_is_bss_excluded().
Prepare for that by changing the logical flow so that further checks can
be inserted.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:46 +0000 (13:51 +0200)]
WNM: Remove unused age parameter for neighbor comparison
compare_scan_neighbor_results() was taking an age parameter to check
whether the BSS has been seen recently. This was used historically in a
codepath when no new scan was done. However, the logic was changed in
commit 20ed289a785c ("WNM: Clean up old scan data processing") and the
parameter is not used at all anymore as it is replaced by a different
logic. Remove it.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:45 +0000 (13:51 +0200)]
WNM: Split candidate list parsing into a separate function
ieee802_11_rx_bss_trans_mgmt_req() is already dealing with a lot of
things including the decisions on how to act on the frame. Split out
candidate list parsing to make it easier to work with the function.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:43 +0000 (13:51 +0200)]
WNM: Store whether disassociation address is an MLD MAC address
Commit 17a2aa822c26 ("WNM: Follow BTM procedure if the last link is
dropped") added code to store either the MLD MAC address or BSSID when
being disassociated. However, it did not save which one was stored
making the tests later on awkward.
Store whether it was an MLD MAC address or not and then do the test
accordingly.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Benjamin Berg [Mon, 29 Apr 2024 11:51:42 +0000 (13:51 +0200)]
WNM: Only trigger selection logic for own scans
Commit e508c070c40e ("WNM: Keep BTM information until connection
completes") changed the logic so that much of the information about a
transition management request will be kept around for longer. However,
doing this also implies that the scan logic can be called multiple times
with wnm_dialog_token being set.
Add a guard to bail out if the scan was not done for a BTM request. But,
add it after the transition candidate validity check so that we reset
the state when a new scan invalidated it.
However, invalidation does not make sense for any scan, primarily an ML
probe request during a connection attempt should not trigger
invalidation. So move the call to wnm_scan_process() further down in the
list to avoid issues.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Add QCA vendor attribute to skip recently scanned channels
Add a new vendor attribute
QCA_WLAN_VENDOR_ATTR_SCAN_SKIP_CHANNEL_RECENCY_PERIOD for sub command
QCA_NL80211_VENDOR_SUBCMD_TRIGGER_SCAN to skip the recently scanned
channels within specified time (in ms), thus reducing the number of
channels for requested scan.
For MLO association, specify the destination address as the AP MLD MAC
address for sending EAPOL frames. Previously, this was set to the BSSID
in all cases (and hoped for the driver to map it to MLD MAC address when
needed).
David Bauer [Sun, 28 Apr 2024 13:13:43 +0000 (15:13 +0200)]
bgscan: Add OWE transition mode SSID to network scan
Add transition mode SSIDs to the SSID scan-list for bgscan. This is
currently missing and bgscan fails to scan for SSIDs on the OWE
transition network if one is currently used.
David Bauer [Sun, 28 Apr 2024 13:13:41 +0000 (15:13 +0200)]
OWE: Enable roaming between OWE APs
This allows to use the ROAM control interface command to force roaming
on a transition network. Previously, this was not possible, as the open
SSID is stored for the connection profile. Add a new function to also
return OWE transition networks if the profile SSID is set as the
transition-ssid for the OWE RSN network.
Johannes Berg [Wed, 24 Jul 2024 17:37:07 +0000 (19:37 +0200)]
hostapd: Fix BSS parameters change count on link removal
The value currently is simply incremented, but it must not
reach 255 and rather wrap around from 254 to 0, since in
the Reduced Neighbor Report 255 means unknown. Fix that.
Fixes: d95838b7932b ("AP: Add support for testing ML link removal") Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Add QCA vendor subcommand to request audio transport switch
Add a new sub command QCA_NL80211_VENDOR_SUBCMD_AUDIO_TRANSPORT_SWITCH
to request the audio data transport switch in both the command and event
path. This is used when two or more audio data transports are available
between peers.
SAE: Send external auth status after sending Authentication frame
The driver might delete the STA node details on receiving external auth
status. In SAE authentication failure case on the AP, an Authentication
frame with a failure status is sent to the driver after sending the
external auth status. Authentication frame transmission didn't have a
dependency with STA node details earlier, but with MLO, Authentication
frames should be transmitted using the link address for a given MLD MAC
address by the supplicant. Since MLD to link address translation is done
by the driver, the STA node details are required to transmit the
Authentication frame. Hence, send external auth status to the driver
after sending the Authentication frame.
SAE external auth was introduced in commit 4ffb0fefe4e7 ("hostapd:
Support external authentication offload in AP mode"), which requires the
change in order of sending Authentication frame and auth status in case
of failure when using MLO.
RSNO: Verify all RSNE/RSNXE variants in multi-link cases
Use the RSN Override Link KDE to include the override variants of the
RSNE/RSNXE for each link so that all variants are verifies when
processing the protected EAPOL-Key message 3/4.
RSNO: Use SNonce cookie to indicate support for RSN overriding
This provides an implicitly protected (SNonce is used as an input to PTK
derivation) mechanism for a STA to indicate support for RSN overriding
in a manner that does not cause interopability issues with deployed APs.
In addition, update sm->SNonce on the Authenticator only based on
message 2/4 since that is the only EAPOL-Key message that is defined to
provide the actual SNonce value. While clearing of this internal buffer
on message 4/4 might not cause issues, it is better to keep the actual
SNonce value here since the SNonce cookie can be used at a later point
in the sequence.
RSNO: Use the RSN Selection element to indicate which variant was used
This replaces the use of the RSNE Override and RSNE Override 2 elements
with empty payload to indicate which RSNE variant was used.
In addition, this adds stricter validation of the RSNE in
(Re)Association Request frame to allow only the pairwise cipher suites
and AKMs listed in the indicated RSNE variant to be used.
hostapd_eid_assoc_fils_session() might return NULL if memory allocation
fails. This NULL value then will be used for invalid calculation and
cause unspecified behavior or be dereferenced unexpectedly. Avoid this
with an explicit check of the returned pointer.
Testing functionality to allow EAPOL-Key Reserved field to be set
The new hostapd configuration parameter eapol_key_reserved_random=1 can
be used for testing STA/Supplicant functionality to accept a random
value in the Reserved field within EAPOL-Key frames.
Add an explicit rsn_override_omit_rsnxe=1 configuration parameter to
allow the RSNXE to be omitted when using the RSNXOE and wanting to
minimize interoperability issues with STAs that might recognize the
RSNXE, but not handle it correctly, e.g., when multiple octets of
payload is included.
Hu Wang [Wed, 26 Jun 2024 03:05:12 +0000 (20:05 -0700)]
Enhance select_network() to trigger new scans in some cases
wpa_supplicant select_network() relies on fast_associate to reuse old
scan results. However, this approach does not apply in some cases in
Android:
1 - If the selected network is hidden, and the SSID is in Chinese,
Android switches between fallback SSIDs, necessitating a new scan for
switching between different hidden SSIDs.
2 - Similarly, if the selected SSID is OWE (Opportunistic Wireless
Encryption), and the OWE SSID bands have been changed, select_network()
requires a fresh scan to discover hidden OWE SSIDs.
To address these, enhance select_network() to trigger new scans instead
of relying on fast_associate. This improves network selection behavior
in Android.
Fallback to RSNXE when AP is not using valid RSN Overrding
wpa_supplicant was ignoring RSNXE also if the AP is not using valid
RSN overriding combination when the STA supports RSN overriding. Fix
this fallback to the RSNXE when AP is not using valid RSN overriding.
Vinay Gannevaram [Tue, 12 Dec 2023 07:38:40 +0000 (13:08 +0530)]
Add QCA vendor interface to support Unsynchronized Service Discovery
Add a new QCA command QCA_NL80211_VENDOR_SUBCMD_USD and related
attributes to support Unsynchronized Service Discovery (USD).
Device in active subscriber role sends SDF frames with service
information to find devices that advertise or publish the required
services. Device address, operation type, instance id, service_id,
channel info, service specific info, element container attribute, and
timeout are sent to the driver to initiate USD publish or USD subscribe
operation. The driver sends QCA_NL80211_VENDOR_SUBCMD_USD event with
instance_id and the USD operation status. QCA_NL80211_VENDOR_SUBCMD_USD
is also used to update publish operation and to cancel the publish or
subscribe operation for a service. It also allows all the operations to
be terminated using the USD flush command.
Jouni Malinen [Wed, 11 Oct 2023 09:48:05 +0000 (12:48 +0300)]
RSNE/RSNXE overriding for AP
Allow hostapd to be configured to advertised two separate sets of
RSNE/RSNXE parameters so that RSNE/RSNXE can use a reduced set of
capabilities (e.g., WPA2-Personal only) for supporting deployed STAs
that have issues with transition modes while the new override elements
can use a newer security option (e.g., WPA3-Personal only) for STAs that
support the new mechanism.
Update the version number for the build and also add the ChangeLog
entries for both hostapd and wpa_supplicant to describe main changes
between v2.10 and v2.11. Update the copyright years for the main
programs.
dbus: Fix memory leak with Bonjour params for a P2P UPnP service
Using D-Bus, it is possible to add a valid UPnP service where 'query'
and 'response' are specified. In this case, memory for 'query' and
'response' is allocated but not used nor freed. Valgrind complains as
follows:
42 bytes in 1 blocks are definitely lost in loss record 32 of 75
at 0x484C214: calloc (vg_replace_malloc.c:1675)
by 0x41C673: wpabuf_alloc (wpabuf.c:124)
by 0x41C673: wpabuf_alloc_copy (wpabuf.c:162)
by 0x54F41A: wpas_dbus_handler_p2p_add_service (dbus_new_handlers_p2p.c:2762)
by 0x53B9A2: msg_method_handler (dbus_new_helpers.c:356)
by 0x53B9A2: message_handler (dbus_new_helpers.c:412)
by 0x4EAB4B8: dbus_connection_dispatch (in /usr/lib64/libdbus-1.so.3.19.13)
by 0x5495DF: dispatch_data (dbus_common.c:37)
by 0x5495DF: process_watch (dbus_common.c:73)
by 0x5495DF: process_watch_read (dbus_common.c:89)
by 0x41EE8E: eloop_sock_table_dispatch.part.0 (eloop.c:603)
by 0x41FA46: eloop_sock_table_dispatch (eloop.c:597)
by 0x41FA46: eloop_run (eloop.c:1233)
by 0x56A3CE: wpa_supplicant_run (wpa_supplicant.c:8074)
by 0x40DB06: main (main.c:393)
49 bytes in 1 blocks are definitely lost in loss record 37 of 75
at 0x484C214: calloc (vg_replace_malloc.c:1675)
by 0x41C673: wpabuf_alloc (wpabuf.c:124)
by 0x41C673: wpabuf_alloc_copy (wpabuf.c:162)
by 0x54F348: wpas_dbus_handler_p2p_add_service (dbus_new_handlers_p2p.c:2755)
by 0x53B9A2: msg_method_handler (dbus_new_helpers.c:356)
by 0x53B9A2: message_handler (dbus_new_helpers.c:412)
by 0x4EAB4B8: dbus_connection_dispatch (in /usr/lib64/libdbus-1.so.3.19.13)
by 0x5495DF: dispatch_data (dbus_common.c:37)
by 0x5495DF: process_watch (dbus_common.c:73)
by 0x5495DF: process_watch_read (dbus_common.c:89)
by 0x41EE8E: eloop_sock_table_dispatch.part.0 (eloop.c:603)
by 0x41FA46: eloop_sock_table_dispatch (eloop.c:597)
by 0x41FA46: eloop_run (eloop.c:1233)
by 0x56A3CE: wpa_supplicant_run (wpa_supplicant.c:8074)
by 0x40DB06: main (main.c:393)
Fix this ensuring that query and resp are freed both in the error and
non-error path of wpas_dbus_handler_p2p_add_service(). Also, add a test
in test_dbus.py to verify the correct behavior.
Sunil Ravi [Thu, 18 Jul 2024 23:10:46 +0000 (23:10 +0000)]
Move NULL check for driver private data (drv_priv)
In the hostapd_cleanup_driver() function, the NULL check for the driver
private data (drv_priv) is located within the CONFIG_IEEE80211BE feature
flag. As a result, on platforms where CONFIG_IEEE80211BE is not defined,
driver->hapd_deinit(drv_priv) is called with drv_priv set to NULL. This
leads to a null pointer exception in wpa_driver_nl80211_deinit().
To address this issue, move the NULL check for drv_priv outside the
CONFIG_IEEE80211BE build flag. This ensures that the check is performed
regardless of the value of CONFIG_IEEE80211BE, preventing the NULL
pointer exception.
Fixes: df34c2ced34f ("AP MLD: De-initialize/disable link BSS properly") Signed-off-by: Sunil Ravi <sunilravi@google.com>
Currently wpa_ft_parse_ies() is setting FTE information to wpa_buf
pointer data which is generated after defragmentation. This data will
not contain FTE and Fragment element(s) headers.
IEEE P802.11be/D5.0 describes the MIC to be calculated on the
concatenation of FTE and corresponding Fragment element(s) which
implies the element headers are included for each element in the
fragmented case.
Fix this by correctly populating FTE information when FTE is
fragmented.
Fixes: 43b5f11d9 ("Defragmentation of FTE") Signed-off-by: Veerendranath Jakkam <quic_vjakkam@quicinc.com>
Michael-CY Lee [Tue, 9 Apr 2024 06:50:23 +0000 (14:50 +0800)]
MLD STA: Find partner links by BSSID and SSID
Non-AP MLD finds AP MLD's partner links by BSSID from the scan results.
However, if the scan results contain BSSs with the same BSSID but
different BSS information, the non-AP MLD might assign a wrong BSS to
one of the AP MLD's partner links.
Avoids the problem by using both BSSID and SSID to find the AP MLD's
partner links.
Signed-off-by: Michael-CY Lee <michael-cy.lee@mediatek.com>
During CSA with DFS channels, disable, enable interface is a part of the
algorithm. When interface was enabled old operating class before switch
and new channel were used causing mismatch in
configured_fixed_chan_to_freq() function.
Example of log when switch from channel 157 to 108 was triggered:
"Could not convert op_class 124 channel 108 to operating frequency"
Fixes: bb781c763 ("AP: Populate iface->freq before starting AP") Signed-off-by: Marek Kwaczynski <marek@mazekey.pl>
SAE: Drop default dot11RSNASAESync value from 5 to 3
Going through five extra rounds of SAE commit messages in cases where
SAE peers are somehow unsynchronized feels unnecessary much to do by
default, so drop the default value to 3.
SAE: Disable protocol instance temporarily on sync error in mesh
Reduce the number of unwanted SAE commit retries in synchronization
error cases when Sync > dot11RSNASAESync in mesh cases by discarding
received SAE commit messages for 10 seconds after a sync error has been
detected.
Nicolas Escande [Fri, 26 Apr 2024 12:35:53 +0000 (14:35 +0200)]
nl80211: Fix AP scan with STA fallback error path
I've run into a case where I can bring a 2.4 GHz AP up without it being
in a bridge while I used the 'bridge=' config option. This happens when
the HT scan needed for 40 MHz operation failed to start from the get go
because another dev on the same phy was already in a scan. At the end
the AP is up and running but not added into any bridge.
Upon looking at the code, it seems that some hardware fails to issue a
scan while in AP and thus we have a fallback that switches the interface
from AP to STA before retrying another scan (it will change it back to
AP later on when finished). As we cannot have a (non-WDS) STA in a
bridge, during that procedure, we also remove/add the AP/STA from/to the
bridge as needed.
However, in wpa_driver_nl80211_scan() we do not set
drv->ap_scan_as_station until the end of the switch-to-sta &
retry-scan-start block. This means that when the recursive call to
wpa_driver_nl80211_scan() fails (the hardware is busy in my case) we
restore the interface to AP with wpa_driver_nl80211_set_mode() but that
will not add it back to the bridge. Problem.
To fix this lets always set drv->ap_scan_as_station before calling
wpa_driver_nl80211_set_mode(). In case wpa_driver_nl80211_set_mode() or
wpa_driver_nl80211_scan() fails lets call nl80211_restore_ap_mode() that
will set the mode back to AP but also handle the bridge thing as needed.
Signed-off-by: Nicolas Escande <nico.escande@gmail.com>
Chien Wong [Sat, 27 Apr 2024 08:19:46 +0000 (16:19 +0800)]
nl80211: NAN: Register multicast action frames if possible
The USD passive subscriber and solicited transmission only publisher
require receiving multicast NAN action frames in order to work.
Currently, we are not requesting to receive multicast when
registering NAN action frames. As a result, USD passive subscribe or
solicited only publish may not work.
The NL80211_ATTR_RECEIVE_MULTICAST attribute corresponds to wiphy
ext feature NL80211_EXT_FEATURE_MULTICAST_REGISTRATIONS, which
requires driver support and only a few drivers in the kernel tree
support it. Namely ath9k, ath9k_htc, ath10k and hwsim. We should
fall back to register non multicast action frames if the driver
has no support.
It was confirmed that ath9k_htc starts to work after the changes.
Note that even without requesting to receive multicast Action
frames, some drivers would still upload them. For example, rtl8192cu
and hwsim. This is why test cases like test_nan_usd_match would not
fail.
Tested-on: TP-LINK TL-WN821N v3(AR7010+AR9287, ath9k_htc) Signed-off-by: Chien Wong <m@xv97.com>
Chien Wong [Sat, 27 Apr 2024 08:19:44 +0000 (16:19 +0800)]
NAN: Reject undefined publish type
The Wi-Fi Aware Specification v4.0 only defines unsolicited
transmissions only, solicited transmissions only and both unsolicited
and solicited transmissions publish. The other possibility is undefined
so we should reject it.
Michael-CY Lee [Tue, 14 May 2024 01:20:35 +0000 (09:20 +0800)]
nl80211: AP MLD: Reassign drv->ctx correctly to prevent hostapd crash
When the first link is deleted and there are still remaining links,
drv->ctx should be updated to the new default link on the bss.
Otherwise, drv->ctx points to the address that has already been freed
and makes hostapd crash.
Fixes: d2b62b3fe500 ("AP MLD: Support link removal before removing interface") Signed-off-by: Michael-CY Lee <michael-cy.lee@mediatek.com>
tests: Make SSID verification through beacon protection more robust
Clear scan results at the beginning of the test case to avoid incorrect
behavior if there are multiple entries for the same BSS. In addition,
use a bit longer wait for receiving an updated Beacon frame in scan
results.
Make Beacon frame checks less frequent for SSID verification
Instead of checking the latest scan results every second indefinitely,
add more latency between the checks in case the driver does not update
the time stamp value (i.e., does not report new Beacon frames during an
association).
Ilan Peer [Wed, 1 May 2024 07:19:53 +0000 (10:19 +0300)]
tests: Fix run_eht_mld_sae_two_links()
In the case that the AP MLD is disabled and enabled again, flush
the wpa_supplicant BSS table before reconnecting as otherwise
the previous AP MLD BSSs would be in the BSS table and the wpa_supplicant
would try to connect to them.