Move SR changes from WMA to Target_if and remove
WMA api's which was previously using to update
SR threshold.
Change-Id: I219cd56195471df561a5789fb235e375a14e756c
CRs-Fixed: 3374674
If precac interface is in the middle of CAC, do not
entertain any CSA request coming as part of userspace/
country change/unsafe channel handling. Processing the
Channel change request in middle of CAC leads to teardown
of the SAP as SAP is in SAP_STOPPED state.
This leads to pre-cac failure.
To fix this, prevent CSA while CAC and once CAC is finished,
check the sanity of the operating channel based on
unsafe/disable channel list and move it to a valid channel.
Change-Id: I2757753a2a13bdaef60f9b9a46fcd569ddb7306c
CRs-Fixed: 3505593
The issue has been addressed in the supplicant.
So revert change: I89b501fc4da8615ddb8973e5edce73294793558f
Change-Id: I02b74a512c97803c597942c1ea00e3a02fbf4599
CRs-Fixed: 3544617
TDLS off channel will be supported by FR 80104, so remove the
checking when setting flag ENA_TDLS_OFFCHAN.
Change-Id: Iae135e835bd74e32a3f712083263192ae7cb635f
CRs-Fixed: 3545087
Add vendor command API to set EPCS (Emergency Preparedness
Communications Service) capability and trigger EPCS
function enable or disable.
Change-Id: I2ffe4d268731c57dd6e3ddfe0586d4dcb77dc611
CRs-Fixed: 3511170
In NAN+eMLSR concurrency, when NAN is already present
and if eMLSR comes. Host disables(by sending
MLO_LINK_FORCE_MODE_INACTIVE_NUM) one link only after
both vdev up. But there is a race condition where
firmware enters to emlsr mode just after link vdev up
and host sends the disable command after
WMI_VDEV_INSTALL_KEY_CMDID. This causes inconsistency
in firmware eMLSR state machine. Firmware asserts as
other mac is occupied by NAN and eMLSR module tries to
change the hw mode to eMLSR mode. Both mac's should be
free/available for eMLSR STA operation when firmware
tries to switch hw mode to eMLSR.
As part of fix, Host to send mlo_force_link_inactive=1
in peer assoc command of link vdev to indicate firmware
not to enter in eMLSR mode. Host shall disable one link
as part of concurrency evaluation post connection anyway.
Once the concurrency is gone, host to enable both links
to let firmware enter into eMLSR hw_mode.
Change-Id: I0ee141b51868a3c5c5d9ff6cf1bd3765bc450fa4
CRs-Fixed: 3527911
The issue is host always uses assoc link channel width
to calculate the list of the channel(s) to trigger
wide band scan for both links.
In the case of N link MLO connection, the associated
channel and associated channel width may be different
for assoc link and non-assoc link(s).
Fix is to choose the proper associated channel width
and prepare the channel list for each link(s).
Change-Id: I2fce5d1153d396f1bfc9575572d90d2cfec0ac8a
CRs-Fixed: 3539444
Convert link id bitmap to vdev bitmap and update policy mgr
connection table.
Record the current force link command state to mlo context.
Change-Id: I0b28222de10a9493a153ee00189b87e4275a4231
CRs-Fixed: 3515467
Change the cfg enablement check APIs (currently FISA),
to static inline, in order to avoid function call
every time to check if the feature is enabled or not.
Change-Id: Ie541313c1b0a7512f71a6189a706871e57b3145e
CRs-Fixed: 3540739
Currently, RSO is stopped and PCL is updated to firmware when
SAP stop happens. But this causes inconsistency in roam sync
handling if roaming is in progress. As host driver RSO state moves
to STOPPED, host doesn't send ROAM_SYNC_COMPLETE though
ROAM_SYNC_IND is processed successfully. Firmware waits
for 6 seconds and asserts as it hasn't received either
ROAM_SYNC_COMPLETE or ROAM_ABORT(RSO_STOP).
So, don't stop RSO and avoid PCL updation when roaming is
in progress in SAP stop case. PCL gets updated to firmware
post roaming anyway. Same is applicable for SAP start also.
Change-Id: I9a6c4579d9cfab9041ee76739705d4db5a40192d
CRs-Fixed: 3518876
CPU min frequency is boosted to maximum once roam start is received.
Add API to reset CPU frequency in case of HO failure.
Change-Id: Ib59d4d69a68f311dbb01609e7f5f64c307fcedc2
CRs-Fixed: 3522942
Currently FISA accesses the dp_soc (converged DP soc handle)
for various misc info (like hal_soc handle etc). Since these
information are now stored in the non-converged DP component,
there is no need to access dp_soc.
Cleanup the access to dp_soc (converged DP soc handle) in
the FISA path.
Change-Id: I6f373eb1ee57933d631237f7281c99bfceb100c3
CRs-Fixed: 3512048
Add CDP APIs to send FISA related HTT msgs from the
non-converged DP component.
Change-Id: Ie8dcb5d68497ac0de4c0781c981ede2239df0a78
CRs-Fixed: 3512045
Remove only 6 GHz using policy_mgr_is_6G_chan_valid_for_ll_sap()
if WLAN_FEATURE_LL_LT_SAP_6G_SUPPORT is not defined.
Change-Id: I828c61adf3c88c2a5100fff603a007d8a6c85a88
CRs-Fixed: 3532992
In api mlme_connected_chan_stats_request(), dereferencing
of freed pointer is observed by static analysis tool.
Remove dereferencing of the freed pointer in the api
mlme_connected_chan_stats_request().
Change-Id: Ic09b68395a9c3c67aab4a704a915b0d92a00e16d
CRs-Fixed: 3530020
MLO sta + 2nd sta case, 2nd sta got pcl type: PM_NONE, total pcl
freq num 0, but dual sta follow pcl strictly, so PCL 0 is sent to F/W,
failed to roam to any freq.
To fix it, for PM_NONE and PCL freq num 0 case, don't follow pcl strictly,
send all valid freq as PCL to F/W.
Change-Id: Ie3529fd85dcfcfd3cc44ed6c25fc8fefe7a6c769
CRs-Fixed: 3534201
Currently host driver does not allow NAN+SAP SCC on unsafe channel.
With this change add support to allow NAN+SAP SCC on unsafe channel
if g_nan_sap_scc_on_lte_coex_chan ini is enabled.
Change-Id: Iccc37f28d8523cd335cce5fcfb0293f3f8cc8959
CRs-Fixed: 3531489
During FTM mode start modules there is possibility from recovery path
bus bw timer stop is being called before dp_ops gets initialized.
This will lead to accessing uninitialized dp_ops access, to avoid this
check for conn mode before accessing the dp_ops and skip timer stop.
Change-Id: I3aef52f70bc0ea6294c87e7cd889ec63a4a6e3a4
CRs-Fixed: 3531437
The roam candidate frame event gives the beacon of
the roam candidate. The SSID IE check expects minimum
size of the SSID to be 2 bytes. Because of this, the
candidate with ssid size 2 or less is not cached to
the scan table.
Fix this by not considering the size of the SSID IE.
Change-Id: Ifbe36ac8cf0f9c75a3830a2826f5330fa300df5b
CRs-Fixed: 3512310
Currently, PCL returns all 5 GHz channel and SCC channel for SAP
when an ML-STA is present. But this may lead to a case
where the CSA request might be for a DFS channel which needs CAC.
This needs a dedicated mac and ML-STA can't move to MLMR in such
cases. When firmware attempts to move STA to MLMR from MLSR,
firmware asserts as the other mac is occupied by DFS SAP for
CAC operation.
So, allow CSA to a DFS channel only if the channel is present in
PCL list as it covers the supported channel list anyway. Fix
policy mgr table entries also by allowing SCC channels only.
Change-Id: I079fa6beea97fcd7c926ad4bd2bf928a0442d146
CRs-Fixed: 3484572
Driver doesn't enable RSO currently for SL + Open auth cases.
Enable the RSO from connect complete path for open auth. Add changes
to enable the RSO only after all the ML links are up in case of
N-link MLO.
Change-Id: Ie3540493660e3605fe6739dc6d7706782b8a78ea
CRs-Fixed: 3509724
1.Host receives QCA_NL80211_VENDOR_SUBCMD_MLO_LINK_STATE
vendor cmd.
2.Parse new attribute
QCA_WLAN_VENDOR_ATTR_LINK_STATE_CONTROL_MODE.
3.Restore mlo link(s) state as per driver.
Change-Id: I79327618a007cd1f50ca1eafcb3633bcbac2e13f
CRs-Fixed: 3529507
1. Host receives QCA_NL80211_VENDOR_SUBCMD_MLO_LINK_STATE
vendor command.
2. parse the new attribute
QCA_WLAN_VENDOR_ATTR_LINK_STATE_MIXED_MODE_ACTIVE_NUM_LINKS
and get the number of MLO links to operate in the active state.
3. If any of the links is in an inactive state due to concurrency,
we need to reject the request.
4. Host sends to FW via WMI command WMI_MLO_LINK_SET_ACTIVE_CMDID
via the below WMI attributes with mode
WMI_MLO_LINK_FORCE_ACTIVE_LINK_NUM and set the wmi flag
control_flags to indicate if FW need to use force link number
instead of force link bitmap.
Change-Id: I7596c4815630563cbc419e0e9df28f1330775334
CRs-Fixed: 3529488
Configure PCIE Gen Speed and Lane width INI based on the
MCS, BW and TXRX throughput using traffic monitor.
Change-Id: Ie899033dfe54dea1533b47b895bda52deb12d0e9
CRs-Fixed: 3532028
1.Host receives QCA_NL80211_VENDOR_SUBCMD_MLO_LINK_STATE
vendor command.
2.Parse the new attribute
QCA_WLAN_VENDOR_ATTR_LINK_STATE_CONFIG and get the
link and corresponding link state.
3.Send it to FW via WMI command
WMI_MLO_LINK_SET_ACTIVE_CMDID
Change-Id: I78a055f4a9fb60ccce106f9e4761fbf6d4d2f262
CRs-Fixed: 3529456
When roamed from mlo to single link, link vdev1 disconnected, assoc vdev0
RSO is disabled, after vdev1 disconnected, even vdev0 RSO is enabled again,
roam scan channel list is trimmed.
To fix it, when link vdev disconnect, just stop RSO, don't disable RSO,
roam scan channel list won't be trimmed.
Disable RSO only when assoc vdev disconnect.
Change-Id: Ie624cffa3acaebcd73abb7be54ce1016e88447f6
CRs-Fixed: 3522793
In current scenario during OWE roaming when HB failure occurs, it
should clean copied reassoc response and roaming info. But it is not
getting freed which leads to host driver crash while on idle shutdown.
In mlo_roam_copy_reassoc_rsp() memory is allocated for
sta_ctx->cpied_reassoc_rsp, sta_ctx->cpied_reassoc_rsp->roaming_info
and connect_ies->assoc_rsp.ptr but if memory allocation fails for
these it directly retuning without feeing previously allocated memory.
Add change to free the prevoiusly allocated memory if memory
sta_ctx->cpied_reassoc_rsp->roaming_info and connect_ies->assoc_rsp.ptr
fails to allocate memory which avoid the memory leak issue.
Change-Id: I7ca6ee526bf08c614ff25e4c3457db0dd58fc9aa
CRs-Fixed: 3507296
Currently policy_mgr_modify_sap_pcl_for_6G_channels() only find the
first STA or P2P-CLI, then check its channel is 6 GHz or not,
It will lead to different behavior if there is MLO STA which has one
link connected on 6 GHz, sometime PCL modify for 6 GHz channels,
sometime not.
Change to find first 6 GHz STA or P2P-CLI, only if has, continue to
do 6 GHz channels modify.
Change-Id: I96d01ded297fb900f0ea715b75cbdadb9faade3d
CRs-Fixed: 3526650
Add pld_pcie_set_tsf_sync_period and pld_reset_tsf_sync_period to
update time sync period vote from wlan driver based on use-case
e.g. tsf sync start / stop. Extend gtsf_ptp_options ini config
to control this functionality.
Change-Id: Ie4a6c80beccd6d315c9f3b31263de7f3b8e0ac2a
CRs-Fixed: 3501976
Currently there is no feature capability support for ll_lt_sap
and host driver and user space assumes that support for ll_lt_sap
is always present which not always true.
This may result in unexpected behavior if firmware does not
support ll_lt_sap.
To address this issue, add a feature capability to indicate
to userspace if host driver and firmware supports ll_lt_sap.
Change-Id: I7fbdb1996cab473f7a89761a35b6859fd2729c0a
CRs-Fixed: 3521091
Currently, driver gets peer multicast address list from kernel for
all interfaces and it send this list to firmware. For NDI, on NDP
confirmation indication, driver appends the multicast list with
new NDP peer added and then send it to firmware. But this is done
for first NDP peer only, kernel does not provide multicast list
again as kernel already configures the multicast addresses, unless
there is some reset happens in kernel.
This can cause issue as firmware is expecting the updated multicast
filter on each NDP confirmation (new peer is added) to enable the
NS frame exchange. As new peer is not added in the filter, firmware
will reject the NS frame.
So, to fix this, update the multicast list on NDP peer addition and
deletion.
Change-Id: I6371199ae5c3a8f4088987f1e6f10c39cbed1685
CRs-Fixed: 3518887
Currently, host driver does not allow NDP formation if two STA is
in connected state. But driver should not allow NAN enable in first
if two STA is in connected state as STA + STA + NAN concurrency is
not allowed.
So, to fix this issue, add check for STA + STA concurrency in NAN
pre-enable in which NAN will drop if two STA + STA is present.
Change-Id: I6e6baa386d50b2903118660f10cc98ffcba60705
CRs-Fixed: 3481148
In MLO STA case, even though twt_responder ini is disable but
still this bit is getting populated in he_cap for non assoc link.
Reason:
1. twt_res is getting set from populate_dot11f_twt_he_cap()
for assoc link. For non assoc link, it's getting set from
mac context.
2. populate_dot11f_twt_he_cap() api doesn't have the interseted
value of ini and firmware caps according to twt component api.
As part of fix,
1. twt_req and twt_res should have intersected value in
populate_dot11f_twt_he_cap()
2. Update twt_res from populate_dot11f_twt_he_cap() for non
assoc link also as it will update the intersection of ini and
fw caps
Change-Id: I6ef34723f879aead4517c28cf6cfad2fa11165fe
CRs-Fixed: 3524903
If MLO STA connect with one 5 GHz low link and another 6 GHz (non-PSC)
link, and start 2 SAP on 5 GHz band, one SAP will start on 5 GHz low
channel doing SCC, another SAP will not start on 6 GHz (non-PSC)
channel and start on 5 GHz high doing MCC.
Add one PCL type which add SCC on 5 GHz low channel to
PM_SCC_ON_5G_HIGH_5G_HIGH_PLUS_SHARED_2G for this case, so that last
SAP can start on 5 GHz low channel doing SCC.
Change-Id: I4312873c159ac14652990c7c1079c13476dcd383
CRs-Fixed: 3508539
Currently, TDLS code sends setup request and then forwards
TDLS discovery response frame. For MLO case, since supplicant
gets the link id from the link identifier element of discovery
response frame, there is issue when supplicant tries to add
new station because link id has not been properly set.
To resolve this issue, tune the TDLS code to forward the TDLS
discovery response frame first and then send setup request event.
Change-Id: I89b501fc4da8615ddb8973e5edce73294793558f
CRs-Fixed: 3528071
With this change add two helper APIs to check if the input
concurrency mode is SAP mode and to check if the input mode
concurrency mode represents beaconing mode.
Change-Id: Iccc82fccf20e43f600efbb4a83c00d51f1d2cd14
CRs-Fixed: 3521665
Based on requirement a new concurrency mode PM_LL_LT_SAP is
added, add corresponding pcl table entries for LL_LT_SAP.
Change-Id: I12bf51547984c714499f7f0f35581251306e1a84
CRs-Fixed: 3514184
