Add an ini configuration that configures the inactivity timeout for NDP
peer. The value is sent to the firmware via the VDEV param
WMI_VDEV_PARAM_NDP_INACTIVITY_TIMEOUT.
Change-Id: Ifffe6bb40df94761d739463cf7e54a0be7e2a375
CRs-Fixed: 2531756
Currently, operating channel is populated for hdd_config
In 6G channel numbers can be same with 2G/5G
which can lead to erroneous behaviour in case of 6G.
Fix is to add operating frequency list in structure hdd_config
as frequency will always be unique.
Change-Id: I3f21b75c6763054afd42df6ea13d5ce96b0ed0a8
CRs-Fixed: 2517114
For STA+SAP on SCC on DFS channel and the DFS master
capability is disabled on SAP by INI. The CAC is skipped
in protocol stack. Do not block tx for SAP in
such condition.
Change-Id: I62a130b246d37d10492d287776c5b51e48d63c5f
CRs-Fixed: 2525465
The nss value in the "iw <dev_name> link" command and
"iw <dev_name> dump station" is updated from the vdev mlme
object vdev_mlme->proto.generic.nss. This nss value is
updated only from lim_process_assoc_rsp_frame() and
lim_extract_ap_capability() only is the ap nss is less than
the session_nss. In roaming case the session_nss is updated
with the AP nss in lim_populate_peer_rate_set() so in assoc
response path the the global mlme vdev nss is not updated.
Update the vdev mlme object nss in lim_process_assoc_rsp_frame()
from the pe_session.
Change-Id: I2575a591f73ee172a83d4681e5f3134d5c9c60b0
CRs-Fixed: 2525710
This is a propagation of qcacld-2.0 commit 98821c7dcf49d ("qcacld-2.0:
Add CONFIG_MULTI_IF_LOG to support multi if log"). Add CONFIG_MULTI_IF_LOG
to support multi if log by assigning the specific netlink protocol
when dual wlan cards are loaded.
Change-Id: I3ea9574899c23a8d2ebbdea5b958cc3ad7947fe5
CRs-Fixed: 2525483
The bss_idx represent vdev id in pe session. pe session already
have smeSessionId/vdev_id to represent this.
Thus remove duplicate bss_idx from struct pe_session.
Change-Id: I18c59d49233dd547f38d7f8a23be5ebbe7d72c30
CRs-Fixed: 2531014
The bssId in struct sDphHashNode represent vdev_id and same can
be used from pe session which already has this info so remove
bssId param.
Change-Id: I47a7f27e95bfaa9ecd617cb531895601a27dfbf5
CRs-Fixed: 2530965
Currently the SAP memory is not freed as part of stop
adapter as the sap ctx is mem zero in deinit ctx, hence
the addr to free in NULL
Fix is to extract the the sap config from adapter and
then free the sap ch list memory from it.
Change-Id: I8c0bf66765c34f0936d694d260ce1544791beecc
CRs-Fixed: 2530985
Many legacy params of tpAddStaParams are unused, thus
cleanup the unused params from tpAddStaParams.
Change-Id: I1be957d2f34d2202bcc1d44e69ad5cba0812c60b
CRs-Fixed: 2530961
Refactor the start bss request to use channel frequency instead
of channel number.
Change-Id: Ic936fe060a5b2bc8d058709950417f3939a3c3d4
CRs-Fixed: 2518026
Currently get valid channel vandor command is defined
under EXTSCAN feature and also depends on gExtScanEnable
ini. when Extscan feature is disabled by ini or by feature
flag this command doesn't work.
As this command is to get the valid channel list, this
should not depend on extscan feature.
To address above issue, decouple get valid channel command
from extscan feature.
Change-Id: I6496cb94e6330f071f8027607e7d55a8f5d3db8a
CRs-Fixed: 2518313
Some target chipset support fw stats version 2, but we didn't enable
it in some configure file. Enable it in these configure files which
depend on chipset.
Change-Id: Ief8b87bdb466a1838e6a61928f0154868f0ff220
CRs-Fixed: 2524266
Currently sta id validation is not there in del
station command for tdls. There could be a case where peer
is there but sta id is not valid.
Consider a case where there is tdls connection between 2
p2p clients and one of the client decided to connect to SAE
AP, this brings up concurrency scenario on that client for
p2p client and station, so tdls connection should be tear
down. device tear down the tdls link and inform the supplicant
regarding the same. Now in the driver scheduler, connect
command is there which is queued when device tries to connect
to SAE AP, this command gets a chance to execute.
When SAE connection is executed, driver gets a command from
supplicant to disable tdls links for the previous connection
which is already teardown, this command is queued into
scheduler which blocks the supplicant thread.
For SAE connection to get complete driver waits for auth
request from supplicant.
Here supplicant waits for the tdls disable link command's
response which will never come as driver waits for SAE
authentication which leads to deadlock.
To address above issue, add a check to validate sta id
before queing tdls down link command to scheduler, if the sta
id is invalid, return response to supplicant as tdls peer is
already deleted.
Change-Id: Ie983aa9f217c0d7b4508639bb33ad1a44b3675c7
CRs-Fixed: 2521529
A check has been introduced while implementing SAE roaming to send
the SAE status to SME only when the STA is not in joined state.
But this doesn't handle the SAP case where the status is expected
to be updated always.
Update the status to SME in below cases
1. SAP mode: Always
2. STA mode: When the device is not in joined state
If the device is in joined state, send the status to WMA which
is meant for roaming.
Change-Id: I20c134a3eea7c65db1fe1950c02b8b9e1ee3a34d
CRs-Fixed: 2524973
Currently as part of tdls enable link, If the peer is buff
capable or tdls uapsd is enabled in INI. UAPSD is enabled in
FW. Avoid sending this command to FW as it effects the behavior
with AP.
Change-Id: Idc023cb9de3ab63e7643815d473fe0efa98069b9
CRs-Fixed: 2528208
struct bss_params has many legacy params which are not used. So
remove the unused params from struct bss_params.
Change-Id: I7d930e83ff0ce56bf8fb07789f9fca9f94fb6f74
CRs-Fixed: 2530215
The add_bss_req struct is used while handling add bss resp,
and is stored in vdev priv mlme.
The resp path require the info which can be filled from vdev resp
and thus no need to store the add_bss_req. From response fill and
use the add_bss_resp in add bss response handling path
Change-Id: I7ec1189c7d399e181f451a442a2578964eb4e47d
CRs-Fixed: 2529289
Add 6GHZ operation information element structure and HE extended
capabilities for 6GHZ information element structure to dot11f frame
specification and parsing files. Also add them as optional IE in
required managment frames.
Change-Id: I5da20543c245e3e02806a20fbfe3ff4326db61f4
CRs-fixed: 2530409
Update SAP data structures to use channel frequency values instead
of using the channel id values to support 6GHz channels in SAP.
Change-Id: I9ef5857e8dcf3f7d879495d3f3c3ead083fe0bf0
CRs-Fixed: 2513083
Module depends on msm headers which are not yet available. Thus, disable
the module under QMAA.
Change-Id: I91000cb3140f7d47e418c6a97c9dcf8b464ecdd2
CRs-Fixed: 2529636
Platform driver is still doing MHI cleanup at the current location of
clearing the recovery flag. Clear it after platform driver callback.
Change-Id: Ifa81950790fa3d2704ed1aed85301a99c6383470
CRs-Fixed: 2526873
As a part of 802.11ax amendment, 6GHz band operation is added.
Since the 6 GHz channel numbers are overlapping with existing 2.4GHz
and 5GHz channel numbers, use frequency to identify unique channel
operation instead of channel number. Channel frequency is unique across
bands.
As a part of above requirement, in the existing WMI interface handlers
and related functions, wherever missing, add frequency attributes to
identify unique channel operation.
Change-Id: I757f60cef414d206b275b4fbd192d5d22bbc93d7
CRs-Fixed: 2517901
In function wma_process_update_opmode, interface phy mode
is used to update peer phy mode to FW, which will cause
a mis-behavior in FW during channel switch.
Use peer phy mode, instead of interface phy mode when
driver is sending WMI_PEER_PHYMODE cmd to FW.
Also to make peer phy mode population to FW possible,
add the following APIs to do host to FW phy mode
conversion:
wma_host_to_fw_phymode_160
wma_host_to_fw_phymode_11ax
wma_host_to_fw_phymode
Fix some kernel-doc discrepancies for the following
existing functions:
wma_fw_to_host_phymode
wma_fw_to_host_phymode_160
wma_fw_to_host_phymode_11ax
Change-Id: I6184a1735d37f648763b149795ac7930bce45346
CRs-Fixed: 2497684
When wma_shutdown_notifier_cb post wma_cleanup_vdev_resp_and_hold_req
the msg.bodyptr is wma. Now during flush the drive will try to free
the msg.bodyptr and may lead to assert.
Fix this by adding dummy flush callback so that driver will not try
to free msg.bodyptr for this msg.
Change-Id: I0734a2ee80dff1d9d4437a009c4c0061485b4de5
CRs-Fixed: 2529950
Currently the driver does undo acs which clears away
the acs cfg ch list, master ch list, and sap_ctx->ch_list
before the ACS is complete (race condition), which can
lead to pointer access after free.
Fix is to wait for ACS complete event, and then clear away
the above mentioned ch lists.
Change-Id: I55de1e94d1fc3ebb99891465131de11ea3204778
CRs-Fixed: 2519650
