With MLO AP removal procedure, AP MLD can remove a link that happens to be
the assoc link for a given Non-AP MLD and the Non-AP MLD can continue to
function after that. In the existing implementation, Non-AP MLD is
returning failure in the MLO disconnect path if the assoc link is NULL, fix
the same.
Change-Id: Ib14c87da083a99f5350c9b73df2b31d08749ff0a
CRs-Fixed: 3604869
1.Enabling CFR capture for pebble.
2.Clean up redundant check for Pebble while fetching rtt info
3.Clean up redundant check for Miami while fetching rtt info
4.Add 2023 to copyright year
Change-Id: I2ac845a1c5914004a0af4a007264d0cd5f431d3f
CRs-Fixed: 3596397
Fix dec and test bit of shmem arena deinit count, as
qdf_dec_and_test returns true if the decremented resultant
value is 0,or false for all other cases.
Change-Id: I59523078bec409badbe253243443a12c508c036c
CRs-Fixed: 3593556
On disconnect ML peer state is set to disconnect initiated,
this helps to avoid sending multiple deauth/disassoc OTA
frames to AP on each VDEV.
If link switch fails the ML peer will be in this state so
for next userspace disconnect request deauth/disassoc frame
is not sent OTA and AP might not be aware of STA leaving the
network.
Do not change ML peer's state on VDEV link switch disconnect
to ML_PEER_DISCONN_INITIATED.
Change-Id: Icbbb46bbf78cea796a3a1e07754be3a626390ad0
CRs-Fixed: 3601739
EIRP Power value in master channel list for power type VLP
is filled with garbage values even though VLP rules are not present.
Due to this, best power mode selection algorithm selects VLP as
best power mode and fails while setting frequency.
To fix this, memset the master channel list 6 GHz AP.
Change-Id: I21bfd69445aeb4756474ad452be0099498e9d5a6
CRs-Fixed: 3574999
Currently, driver does not use "WLAN_TWT_CONV_SUPPORTED" flag in
TWT APIs declaration. So, for some targets, it causes
compilation failure.
So, this change fixes the compilation issue.
Change-Id: Ie1b92119ef37b85f4c7eb1d0e23e7910eeab3415
CRs-Fixed: 3603668
1. Upper layer sends QCA_WLAN_VENDOR_ATTR_LINK_STATE_CONTROL_MODE
vendor command to configure control mode in host/fw.
2. Host receives GET_ML_LINK_STATE driver command to return
control mode (configured at step 1) to upper layer.
Host should reply as a response of GET_ML_LINK_STATE with same
value of control mode configured at step 1 to upper layer.
Change-Id: I3b5cb3bec31957ee8479d7f366d1e360421a7573
CRs-Fixed: 3601735
Currently the T2LM timer is maintained by host.
Code change made to add support to run the T2LM timer in FW.
CRs-Fixed: 3594141
Change-Id: I7eee4800f68b07d5492a0336e3fe8e0db17d2020
Currently if fcc_constraint 0/2 is set and there are
active 6 GHz STA/P2P client connections, host keeps the
bonded channels enabled, but in super channel list
incorrect set of frequencies remain active. This
happens because the active frequency start and end
indexes are not according to the super channel list
indexes.
To address this issue, convert the active 6 GHz
range start and end indexes according to the super
channel list indexes.
CRs-Fixed: 3595197
Change-Id: I8bb3a5093eb3b8bd970fcd371a4939570207385b
Currently, if FW advertises more than three HW modes and more
than five caps, host returns error as the current max is below.
Increase max hw mode to 4 and mac phy cap to 8 to support cases
where we advertise more than hw 3 modes and more than 5 phy caps.
eg. SMM, DBS_or_SBS, AUX_EMLSR_SPLIT, AUX_EMLSR_SINGLE.
Change-Id: Idbc978a1a259d66e7d893d82f8becd6325019976
CRs-Fixed: 3599091
When updating the AP link address and channel info after
roaming, update the self link address also to the link switch
context on each vdev. This information will be fetched while
sending roam + auth event to userspace after roaming.
Change-Id: I363a5ee8d65bb21d84c15879996369f4e397f2a6
CRs-Fixed: 3589401
Hopping radar happens on whole band, hope all DFS sub channels are put to
NOL, but F/W only reports single radar event on special freq to host.
When DFS puncture enabled, sub channel marking is enabled, host only put
single affected 20M sub channel to NOL instead of all current bonded sub
channel, AP still stays on current freq for valid puncture bitmap
instead of leave DFS freq.
If the radar is a frequency hopping radar that truly hops over entire DFS
band, then the AP and clients are best served by moving out of the DFS
band. But it could well be that only part of the band is used. It is
difficult to find product specifications with info on the total bandwidth.
Mostly military systems use frequency hopping to avoid jamming. So a middle
ground would be to put full bandwidth of current AP to NOL for FCC type 6
frequency hopping radar.
When frequency hopping radar happens, F/W set flag that full bandwidth need
put to NOL in new added tlv radar_flags of event
WMI_PDEV_DFS_RADAR_DETECTION_EVENT, When find the flag is set, host will
put full bandwidth of current AP to NOL even sub channel marking is
enabled.
To avoid compatibility issue between old/new F/W and new/old host, add a
F/W service bit wmi_service_radar_flags_support and host service flag
is_full_bw_nol_supported.
If the wmi_service WMI_SERVICE_RADAR_FLAGS_SUPPORT is enabled by FW, host
shall set/unset a bit in the host service flag is_full_bw_nol_supported
based on host capabilities. This host service flag is then sent to FW via
WMI_INIT_CMDID.
Change-Id: I6e280ca4ab00dce97c5aec90c7a4b4fc7c4a440a
CRs-Fixed: 3595476
Link ID for the bridge vap is set during creation.
However during mlo ap vdev attach, it gets overwritten by 0.
Due to this, during 2.4 GHz link removal using ml-reconfig,
2.4 GHz linkid and bridge vap as primary vaps linkid matches
which leads to full disconnection of the peer.
To fix this, retain previously set linkid for bridge vap.
Change-Id: I28a3da110aa6427ae7024fe173ae04c5a6bebe41
CRs-Fixed: 3592545
In util_parse_rv_info_from_linkinfo, accessing the array of data
without validating the index of the array will cause OOB issue.
Add check with the maximum size of the array before accessing the
data.
Change-Id: Ie8b33c83fa1cd18cfbfcb6c7a82b79ddd86e8808
CRs-Fixed: 3580821
Currently kernel-doc script doesn't support qdf_bitmap
struct member. So, all structure which has qdf_bitmap
member are excluded.
Support for qdf_bitmap has been added to kernel-doc script.
So, include all structures which have qdf_bitmap.
Change-Id: I95095d458152a0285a24fba398b46effd94525db
CRs-Fixed: 3598814
The definition of the function reg_get_pdev_from_phy_id() is placed
inside the 6 GHz specific macro CONFIG_BAND_6GHZ.
The function is a generic function applicable for non-6 GHz platforms.
Move the definition outside the macro.
Change-Id: I899813268daec85f1393d83b593a7664ae4aaabd
CRs-Fixed: 3595223
Added logs to make sure ml_peer_count has decremented every time mlo
peer is detached and catch rare instance when ml_peer_count is not
decremented properly.
Change-Id: Ic174a0bce22a8a85e9df4ccb840de023b45a7f9d
CRs-Fixed: 3580583
Add support to send TWT IE in assoc/probe rsp to fulfill WFA requirement
FW sends TWT IE in mgmt rx event for probe req and assoc req. Send this
IE in probe rsp and assoc rsp. This change takes care of extracting TWT
IE in mgmt rx event.
Change-Id: I1a97eb58c03f76c7ced31349dc0e3b4a908ceda0
CRs-Fixed: 3585625
When setting the country from user space, the thread sleeps and waits for
the init_cc event to awake. Right now the thread is woken up only in 2
case:
(i) when the country configuration is success from FW and regulatory and,
(ii) when country configuration failed in FW.
The thread is not woken up if there is failure in regulatory. This is
leading to timeout which will assert the kernel. In order to handle all the
return errors from the regulatory init_cc event handler should execute and
the status should be returned to thread waiting for the init_cc event
irrespective of the status of the init_cc handler. In order to achieve
this, the following actions are taken:
(i) Introduce a new regulatory tx_ops set_wait_for_init_cc_response_event()
that will wake up the thread waiting for init_cc response and return the
status of the init_cc response handler.
(ii) Remove the tx_ops set_country_failed() which just sends the failure
status to the thread waiting for init_cc event. This will be handled by the
new tx_ops.
(iii) Send the status of init_cc handler using the new regulatory tx_ops.
Change-Id: I1b29651682e9b7219d428c13d6d0ea299d6f81ba
CRs-Fixed: 3552823
VDEV is moved to INIT state as part of link switch disconnect,
before set MAC address response is received, any disconnect
request on this VDEV will not be handled as VDEV is in INIT
state, if link switch is in progress then it will abort link
switch and starts link switch dequeue process.
If the new disconnect request is from userspace it will
increment the OSIF ops, but if link switch is on assoc VDEV
OSIF is notified to restore the adapter deflink as part of
link switch complete where it wait for all OSIF ops to
complete. This is a deadlock case where driver is waiting
for ops completion on same thread where ops is initiated.
To fix this issue, do not handle link switch dequeue on the
same thread, instead move the link switch state to abort
and when actual link switch thread comes it will flush from
serialization.
If userspace disconnect is not queued as VDEV is in INIT
state due to link switch, kernel won't be notified about
the disconnect as this notification is only done on assoc
VDEV and any further connect requests from supplicant gets
dropped in kernel saying already connected and supplicant
will immediately try disconnect which driver will again
drop as VDEV is in INIT state. To avoid this kernel-driver
out of sync, forcefully move VDEV to disconnecting state
and queue the disconnect request.
Change-Id: I116859601ebba21d44797e74e160b56532ef833c
CRs-Fixed: 3588936
For multi link connection there might be the case where
rsnxe of the AP is not present on the link vdev.
This causes the link vdev alone to downgrade to 11ax
mode and assoc vdev is in 11be dot11 mode.
So for vdev with ML peer, the dot11mode is 11ax,
which causes abnormal firmware behavior.
Reject the partner link that doesn’t pass the security
check and validate the next available partner link.
if none of the partner link passes the security check,
proceed connection with single link.
Change-Id: I080557027180c0566a1c284a93fcc4b69c61a9c8
CRs-Fixed: 3581189
Move the frame control values used for link specific rsp
generation to a header file to access in other source files.
Change-Id: Ib531b9ae88604e98b9119efa610757dc5f399596
CRs-Fixed: 3586052
Currently, If FW sends both mac and counter stats data then
some data may get overwritten as union of structures can
handle only one structure data at a time.
Change-Id: I30f82ac1b3965f504d268e39a2516f57edc6ce38
CRs-Fixed: 3592413
Add support to parse the Max ML Peer ID's parameter received from target
capabilities and update the parameter in global MLO context.
Change-Id: I4118d288a312dc4a18182caf765b0ded014b985a
CRs-Fixed: 3556542
Handle case where device topology needs bridge peer.
So if the device have connection on two opposite links
create bridge peer on the central vap.
Sample topology:
AP
2 GHz<-->6 GHz MLO Links
STA topology
(5 GHz Low VAP)
chip0
/ \
(2 GHz VAP)chip1 chip2 (6 GHz VAP)
\ /
chip3
(5 GHz High VAP)
In the above case we will have Bridge peer on either
5 GHz Low or 5 GHz High VAP
CRs-Fixed: 3575939
Change-Id: I923cc01b3c6e23099436a25565cbabab5a08d93c
In some scenario, mbssid_info->prof_residue could be set to
true, hence mbssid_info->split_prof_continue will also be
set to true. Then for the next loop if buffer split_prof_start
is freed but split_prof_end does not reinitialize to NULL,
then use-after-free happens.
To address this issue, reinitialize split_prof_end properly
when split_prof_start is freed.
Change-Id: Iad7448868cfa4c2dd7922f6c1b2622cf20a6a28c
CRs-Fixed: 3583521
Currently PM_LL_LT_SAP_MODE is not handled in policy manager.
With this change add support for PM_LL_LT_SAP_MODE in the policy
manager.
Change-Id: I7b893b04498957eb4a7e9f74c4b4395d4a823bf9
CRs-Fixed: 3536612
Add enter and exit debug logs for mgmt_txrx module to assist
in debugging timeout issues.
Change-Id: Iad28dca3fa749f803829cd7a2e052578a5c579e8
CRs-Fixed: 3584331
Currently if host receives invalid link switch params from FW,
the request is dropped in host and reject status is not sent to FW.
Update link switch status to FW in all error cases.
Change-Id: If1ce5e30517d98c38e869baa1cfc620302b839b2
CRs-Fixed: 3588204
IEEE802.11be D3.0 has renamed the 'Delete Timer' subfield to 'AP
Removal Timer' in the STA Info field in per-STA profile in
Reconfiguration Multi-Link element. The earlier name did not correctly
reflect the AP removal operation and was a source of confusion.
Update documentation related to the erstwhile Delete Timer to indicate
the new name. Similarly, update a print related to this timer.
CRs-Fixed: 3589907
Change-Id: Ie5c9c4c8756057791095e87532e0e14bcf7ef133
The API reg_is_freq_txable() is not used by_REG_CLIENT devices and
should be moved out of converged code.
Move reg_is_freq_txable() and corresponding dispatcher API
wlan_reg_is_freq_txable() to unconverged code.
Change-Id: I1031a20d40ebcf30326c14901d413669206a8433
CRs-Fixed: 3568932
APs can broadcast support for 20 & 40 MHz in the
HT cap IE and operate in 20 MHz also.
Therefore, use the extension channel offset field
of the HT info IE to identify the bandwidth of the
BSS.
Change-Id: I9987d45ec34ba3327b43ef70eb891d346e9e78c9
CRs-Fixed: 3587792
Do mlo register only for mlo capable soc, since non mlo
chips also could be part of mlo capable target
Change-Id: Ibef1899257f60a00ba3933af44efdb736550b9e4
CRs-Fixed: 3582269