Currently, total_weight is not initialized before calculating
default ML BSS weightage with INI configuration. This may result in
unexpected behavior.
To resolve this, calculate ML bss weight after calculating non-ML
BSS score weight.
Change-Id: I4b9455e0ce69dec5c837f92acad60ac83cdd2502
CRs-Fixed: 3127454
Currently, the code does not handle an MLO connect when in
non-init state. Add logic to clean up any existing connection
before going ahead with the latest request.
Change-Id: Ia5a10bcc1e0d2c8e82d6ea0d82d1a5bf0604a1e5
CRs-fixed: 3116687
Firmware doesn't expect any vdev commands from host while RSO stop
is happening. It sends a response to the RSO_STOP command once
it's done with cleanup. Host needs to run a timer and wait for a
maximum of 6 seconds for the response. Host can stop the timer and
allow the commands to firmware in the below cases,
1. RSO_STOP response with success status
2. RSO_STOP response with HO_FAIL status followed by
HO_FAIL event: Host needs to wait till HO_FAIL event is
received
If firmware doesn't send any response in the 6 seconds wait, issue
a recovery to help to check the firmware state.
Change-Id: I4577f9b0aac71c0c86bd32b59e69d9527bf107b9
CRs-Fixed: 3106032
Host driver creates two vdev as part of initial connection.
If the connected AP supports MLO, both(sta and link) of the vdevs
would be moved to UP state. Otherwise, link vdev would be in INIT
state and not moved to UP state.
Firmware sends roam sync indication to host when it roams to a
new AP. The event carries new AP info and the new AP could be
MLO AP. If the current AP is non-MLO and new(roamed) AP is MLO AP,
move the link vdev state to UP as it's in INIT state.
Change-Id: Ib5dd216074d68b0001755024153263ba9654edbe
CRs-Fixed: 3106059
EHT operation IE contains disabled subchannel bitmap, which providing
a list of subchannels that are punctured within the BSS bandwidth.
Throughput is affected if subchannels are punctured, refer to system
document to get new bandwidth score for different puncture pattern.
Change-Id: I91b9a7ec4331e78373440d4c97eb9f1ba6c45b4c
CRs-Fixed: 3098009
When do SAE-5.7.3 certification test case, STA will choose non SAE-PK
AP to connect which is wrong expected result, non SAE-PK AP beacon
will carry with QBSS IE that contributes more score than SAE-PK AP
score config.
Fix is to give SAE-PK AP more score to pass SAE-5.7.3 certification
case.
Change-Id: Ib3e516ccfa673e19f9bf0869aeffe4f5b1862a71
CRs-Fixed: 3101317
If the MLO module is sending link disconnect from the north, retain its
disconnect source. If it is a southbound disconnect, update the source
to be from MLO manager. This is needed so some legacy modules handle the
disconnect properly and send vdev stop when appropriate. Also make sure
that the SB disconnect will be processed if the link is in connecting
state.
Change-Id: I6644751934697496147b9177ee7beab4d7355e10
CRs-fixed: 3086012
With existing STA code changes, The vdev crypto params are filled
in cm_connect_active() but when STA tries to roam from WPA2 AP to
WPA3 AP or vice versa, the new keymgmt of AP2(AP2 keymgmt is FT-SAE
if STA roams from WPA2 to WPA3 and AP2 keymgmt is FT-PSK if STA
roams from WPA3 to WPA2 AP's) is not updating in the vdev crypto
params instead it holds the AP1 keymgmt.
So, Adding changes to update the vdev crypto params from the connect
request structure during roam.
Change-Id: Ia0ddd6192a45f5d9e37b44665aded5a1fc573243
Clear MLO cap only when it is the last disconnect req as osif would not
have informed userspace for other disconnect req because of
cm_id mismatch
Change-Id: Ifcfde3e0efd6fb28e73679cbbccfa6f64ed35902
When connect from wpa_supplicant and roam from F/W happens at same
time, supplicant hold cm cm_req_lock and wait rsp from scheduler
thread, sheduler thread is blocked to wait for cm_req_lock.
Change-Id: I6b9010972d29be9a32d3b85ee72891ed62b50801
CRs-Fixed: 3087193
In case the vdev is already disconnected, the indication to
the upper layer, would have been sent as part of previous
disconnect/connect failure.
If the upper layer is in process of connecting, sending
the disconnect indication back again may cause it to incorrectly
think it as connect failure. So sending a disconnect indication
again is not advisable.
So if a new disconnect is received in INIT state, drop the
disconnect and return failure.
Also remove osif_cm_reset_id_and_src() from osif disconnect
to avoid race between disconnect complete of old disconnect
and new disconnect request. With osif_cm_reset_id_and_src()
old disconnect might also get dropped in osif and with this
fix new disconnect will also get dropped, so make sure that
last/old disconnect indication is sent to upper layer.
Change-Id: Icf7352d8904473329edff9ec124c6197f214f88b
CRs-Fixed: 3074093
Modify MLO disconnect handling to issue disconnect on all links in
parallel
Also modify connect resp to accommodate ml info in connect resp
Change-Id: I57575c6ce7c8ebff7770f862dc81e7dfd20aa42d
Add flag to indicate whether MLO is supported by the candidate or not,
which is helpful when checking concurrency.
Change-Id: I166ad2963163ae4dc7989516b654dce4cb635cec
CRs-Fixed: 3013418
The MLO cap is getting reset as part of getting valid candidate, this is
leading to reset of MLO cap for supplicant selected ML candidat as well.
Add fix to Clear vdev MLO Cap as part of updating vdev mlme address
Change-Id: I19ce9711011173505a829dc5b308e8890b8ffe6c
Clear the MLO capability bit in the following scenarios:
- On a successful disconnect complete
- On a failing connect complete
- Every time we fetch a new candidate
This will ensure subsequent connections are not effected by this bit.
Change-Id: I4cbda78715450b5db5bc25cb7f6a141d636b712b
CRs-fixed: 3054192
Scenario: In case of STA + STA, NB/SB issues disconnection for
vdev 0 when vdev 1 is in connecting state.
As connection is blocking serialization command and it can try
multiple candidates and can take upto 30+ sec, there is a
chance that disconnect may get timeout and is followed by
vdev delete from osif. This can lead to vdev delete sent
without vdev down/stop and bss peer delete for vdev 0.
So abort connection when trying for candidate if any of the
vdev is waiting for disconnect, to avoid disconnect timeout.
Change-Id: I7ef50b2e870317d2b095dd0385fcbe1c5b86fd20
CRs-Fixed: 3051567
When processing a peer or southbound disconnect on the non assoc link,
there is a vdev stop issue due to the correct state not being set. For
this link, set the disconnect source as MLO instead of passing it down
from the caller.
Change-Id: I8db444e7c80659a1a1026a53326e99c96ef631f0
CRs-fixed: 3040658
Add the following fixes:
-Don't update the SA in mgmt frames when vdev is not operating as
STA
-Set the MLO vdev capability when setting the link address. In
case of non ML connection, this bit should not be set.
Change-Id: I23f8fb8d43afbed687f619c7029893a7d09696d2
CRs-fixed: 3038626
STA connection with OWE transition mode SSID fails due to join
failure timeout. Since, it is OWE transition mode, STA needs probe
response for join success.
The scan table entry of the candidate BSSID is sent in the join request.
This entry is used to create session and to send probe request to AP.
In case of OWE transition mode, the beacon carries wildcard SSID.
Therefore, the scan table entry also has wildcard ssid. Also,
during the candidate selection, scm_ignore_ssid_check_for_owe check
allows even entries with wildcard SSIDs to be selected as candiates
for OWE transition mode.
The join request with Wildcard SSID in the scan entry is used by LIM
to create session, as well as to send probe request.Therefore,
the AP doesn't respond to the probe request with wildcard ssid
leading to join failure timeout.
This change adds ssid from connect request to the cm_vdev_join_req,
which can be used by LIM to create session and to send probe request
whenever the SSID in scan entry is wildcard.
Change-Id: If43d7e8a65bc3e37d67e22609932cf18f804f0ab
CRs-Fixed: 3041142
During p2p AGO and GC connection, beacon interval of GC doesn't
match with AGO which causes connection failure.
As part of fix, update beacon interval of p2p_cli from scan entry
Change-Id: Icb38d129809dea0ef0645df29723a6315e32b0f5
CRs-Fixed: 3039930
When getting rnr IE information, instead of getting rnr pointer through
input parameter, we can just return rnr pointer.
Change-Id: I57f683eb86bb813018a38fa3a6286e9b6e40037f
CRs-Fixed: 3025194
ESS (Extended Service Set) operating mode allows wireless
device to roam anywhere within the area covered by multiple
APs.
IBSS (Independent Basic Service Set) allows wireless device
to connect in peer-to-Peer mode only.
Fix is to allow connection manager module to filter only
those AP(s) which is/are part of ESS network.
Change-Id: I9536557daa624b9e44505efa0e0720e47b645517
CRs-Fixed: 3015220
It is mandatory request to check security mode in 6Ghz band,
so enable check_6ghz_security by default. Keep default disable
for WIN as requested.
Change-Id: Idcc4cb9add98ad4bda8e4357e83460ca25221692
CRs-Fixed: 3006930
Currently driver has only one name for an INI, with
this change add second name as well for few INIs based on
new requirement.
Change-Id: I67bd50652c3be7912c516b81d40f6a1ea7a7b90a
CRs-Fixed: 3007172
Currently driver reset self OCV capability from RSN based
on AP capability which is not required as it may lead to
failure of wifi-cert test case 15.1 that is sta should
advertise self OCV capability irrespective of AP capability.
Fix is not to reset OCV from rsn based on AP
capability to support OCV.
Change-Id: I24cda586ae0f364bb4eb2936985194e01f74875b
CRs-Fixed: 3002963
MLO sta needs mld mac and assoc link information to create bss peer
Get mld address and assoc link info from connection manager and pass
it to bss peer creation.
Change-Id: I147ac8365f3c99c6231df589ac17d39122d57362
CRs-Fixed: 2987496
Management frames from STA should carry SA as link MAC address for ML
connection and net device address for non-ML connection.
To achieve this, update vdev mlme MAC address to link MAC address for
ML connection and to net dev MAC address for non-ML connection.
Change-Id: I2285a43882039dfc706fbf73a494cc09bcba19c7
CRs-Fixed: 2993031
Use seperate event for HO/roaming disconnect for old AP, as
if a disconnect is received during roaming, the state is moved
to disconnect and thus WLAN_CM_SM_EV_DISCONNECT_DONE is handled
as a normal disconnect and does not clean up the roaming
contexts in VDEV and peer.
Change-Id: I10e616be435b16e1cff6666940171fef6b2b4065
CRs-Fixed: 2994558
In the case of Roam Start sub-state, if a new connect/disconnect
is received, an internal disconnect/actual disconnect is queued
in serialization and then the state is changed to connecting
or disconnecting, but if the disconnect command becomes active
directly the EV_DISCONNECT_ACTIVE is received in roam substates
but is not handled.
As EV_DISCONNECT_ACTIVE is not handled the disconnect is not
processed and VDEV can remain in UP state while CM assuming that
the disconnect is completed will allow a new connection which
cause 2nd BSS peer creation.
Fix is to handle EV_DISCONNECT_ACTIVE event in Roam sub-states.
Change-Id: I30180db2e15af9b793e52e891f86d54c55b7b5e0
CRs-Fixed: 2986468
Refactor connection manager command timeouts to handled unexpected
timeouts in VDEV/PEER SM.
Also flush the req list before deinit to avoid memleaks due to these
timout, if SM is not handled properly.
Change-Id: I6cd1857464691838c75b1ceda32f94ad4cda46a2
CRs-Fixed: 2971466
Define the following functions for MLO operation:
Peer create and delete
STA/SAP up and down
VDEV create and delete
Is MLD AP or STA
Change-Id: I5c9b3276b8dbb4b2a3f055480cab4a865b8af946
CRs-fixed: 2935769
Refine QDF_STATUS_E_NOSUPPORT to QDF_STATUS_E_ALREADY in
cm_check_for_reassoc_hw_mode_change() as QDF_STATUS_E_NOSUPPORT shows
that no need to do action for hardware mode change, it's not an
error or failure.
Change-Id: I6e977729f3c75cc23b921ea38728ff4c5aae7d75
CRs-Fixed: 2961362
Add missing change I4ab3c956c3169b50a69c05d9ed1c36e055aeb586
in connection manager to disable OCV in STA RSN capability
if AP doesnt support it.
Change-Id: I84b666a489b8f2e410a4901b908f57edae86b5d4
CRs-Fixed: 2960828
