Currently Refill buffer pool enqueue is using POOL_SIZE
macro, but there is possibility of dynamically configuring
max pool size less than that. So use max_bufq_len which will
be configured dynamically during refill pool init instead of
POOL_SIZE macro.
Change-Id: I610b5dff71861254aa8b6d30e2c562bd7316e120
CRs-Fixed: 3575599
Add flag in event response struct to indicate event processed
or not.
Use scheduler thread to process set link active event same as
other link switch and vdev event to avoid race condition.
Change-Id: I963f9106e296a59cff5078c85a4accf63c7026dc
CRs-Fixed: 3572868
Mark nbuf cb if Rx packet is routed via offloads, this gives the
info whether particular packet rx ring routing info is correct or
whether packet has entered offloads layer and rerouted back.
Change-Id: Ib3ad71216c514381d0bce1fb4744550d91880254
CRs-Fixed: 3540537
FW will report current active/inactive link bitmap for
each set link command. Extract them and save to respone
data struct.
Change-Id: I8513391fbf3b181c6182d6568d0ce8e83c627bc6
CRs-Fixed: 3564785
Once the FW sends the link switch request to host handle
the request from scheduler thread and send confirmation
back on completion of link switch process with status of
link switch (success/failure).
Add new serialization command type for link switch.
Introduce flags to get the current state of link switch
request, set the state to idle when no link switch in
progress or once the current link switch is completed.
Access to state is protected with MLO dev context lock.
Implement various helper API to:
a) Transition link switch to next state.
b) Get current state of link switch.
c) To check whether any link switch is in progress.
c) To check whether link switch is happening
on assoc VDEV or not.
Introduce a new VDEV flag to suggest the VDEV is in
link switch process and also implement helper APIs to
set/get/clear this VDEV flag.
a) The flag is set at start of link switch, once
the FW request params are validated and before
proceeding for link switch disconnect on VDEV.
b) Clear the flag once the Link switch confirmation
is sent to FW.
Validate the link switch request params:
a) IEEE link ID's received.
b) Check if new connection is part of MLO connection.
c) Check if VDEV is MLO STA VDEV or not.
d) Is VDEV in connected state or not, that means
VDEV is not in transitioning state due to disconnect.
e) Check if any link switch in progress on this MLD
f) Current link ID of VDEV equals the FW params.
If validation is successful, serialize the link switch
command and in the serialization activation start the
actual link switch process.
Change-Id: Ie582650541054c8cf39aaa8316e86a7a40256a15
CRs-Fixed: 3556422
A few RDPs can have 2 GHz + 6 GHz + Scan radio (2 GHz + 5 GHz + 6 GHz)
radios. In this case, the total number of channels are greater than
NUM_CHANNELS.
Hence, increase the channel array size by twice. And filter out the
duplicate channels before sending the channel list to LOWI application.
Change-Id: If9d1676dab24b8e46f5303d5992aaa59ba1e67ce
CRs-Fixed: 3554676
Hex dump in monitor descriptor processing is causing high
CPU utilization. Since it is not advisable to add prints/debug
in per packet processing, remove hex dump debug log.
Change-Id: Ic98c3608a468cf4665ea94b1ebee725ca7121545
CRs-Fixed: 3571491
Debug logs in per packet tlv processing is causing high CPU utilization.
Remove these debug logs.
Change-Id: I520b3a83c7a5cafefe6f2d25d14ea5c1b9816b11
CRs-Fixed: 3571491
Change few static crypto API's to global.
Add support to add new crypto entry based on passing the entire
entry structure itself.
Change-Id: Id2e0a46bb8b44a834d17d2a04b0dc28fc881b4e3
CRs-Fixed: 3571796
Fix tlv processing code for monitor mode in Big endian mode.
Changed all the 64bit tlv extraction apis and used 32 bit extraction apis.
This is needed in big endian mode since HW supports word swap option and
TLV has to be intepretted using 32bit extraction APIs.
TLV tag, userid and TLV len falls on the first 32 bit and hence using
both 32bit or 64 bit extraction APIs give same result.
Change-Id: I9b9ff78c79f21888964d405016c58c3b5988b254
CRs-Fixed: 3551002
If non-ML type connection fails, then the connection IEs
copied in sta ctx are not cleared as VDEV is not ML VDEV.
Clear the IEs if connection fails irrespective of VDEV type.
Change-Id: I2a2ca155aadeb408e9a4a425d3f1f03f4dbfb867
CRs-Fixed: 3572185
At the end of the link removal procedure, issue full disconnect of only
those ML peers that still have the removed link as primary TQM and there is
no ongoing PTQM migration of those peers.
Change-Id: I1a4c34313eb2cb49187c1d9b47e5918ada32ccf7
CRs-Fixed: 3550142
Introduce new twt statusconversion
HOST_ADD_TWT_STATUS_LINK_SWITCH_IN_PROGRESS and
HOST_ADD_TWT_STATUS_UNSUPPORTED_MODE_MLMR for
individual twt in 11be.
Change-Id: I76194ec0172d06c1b176e3afd3a5e66efdd12f4a
CRs-Fixed: 3568732
As the bss peer is closely tied with the vdev creation
and deletion, listeners might be uninitialized and result
race conditions. send an event only to the connected peers.
Change-Id: Ia0ae063c92bba38d2fb67fe4eae84f5562991fa2
CRs-Fixed: 3574176
Extract aux device capability from service ready ext2
event, and report such info to wma/mlme/hdd.
Change-Id: I888624fd3443118ba66fadd7a40fcce4f4b5d521
CRs-Fixed: 3549753
Add APIs to populate, destroy and get bridge vdev count.
Also send bridge VDEV details in partner link info via
VDEV START command to target.
Change-Id: Iea155b09051b6724d07cf2b6052a07c1d3beb7e8
CRs-Fixed: 3567098
1. APIs to check topology to see if the bridge vap/peer will be used.
2. API to check the max number of links supported.
Also check if we can have the primary and assoc links to be different
for WDS STATION. Based on the topology allow/block
connection to a particular AP.
CRs-Fixed: 3562149
Change-Id: Icf3b32036199f39b0d567ac88795dc07760ac1a5
When roam reason is periodic, host driver ignore the rssi
threshold in wmi event parsing.
Change-Id: I66c72576c7a2994d88e4c83bee726d8ec92c4189
CRs-Fixed: 3566441
Currently in htt tx completion for MLO connection
host is not converting peer id into ml peer id.
Due to wrong peer_id host is not able to find txrx_peer
due to which stats are not getting updated.
To fix the issue in case of MLO connection convert
received peer_id into MLO peer_id.
Change-Id: I2e90104053d59a319d46a2628c4e9cdcb32832a0
CRs-Fixed: 3574182
Compute the required pages properly, when num_descs is not
multiple of DP_CC_SPT_PAGE_MAX_ENTRIES(512).
For example, if num_descs less than DP_CC_SPT_PAGE_MAX_ENTRIES,
then 1 page needed rather than 0.
Change-Id: I1c4b12cfebef6a9b315baf9044355caab1b112de
CRs-Fixed: 3571562
Change name of reg_is_6ghz_chan_connected to reg_is_chan_connected
as this API can be used to check if channel is connected for all
bands.
Use reg_is_6ghz_band_set helper function to check if 6 GHz band
is set and reg_get_keep_6ghz_sta_cli_connection to check if
keep_6ghz_sta_cli_connection flag is set.
Change-Id: I106185bad72579e227f9042a40f77934a2cb81c1
CRs-Fixed: 3537696
Modify the DFS debug print level to reduce the prints during bootup
as per customer's requirement.
Change-Id: I9e2c07373ffe036d4dfa7599d49d589eefee5cff
CRs-Fixed: 3568927
With the current behavior, a NULL AFC response is treated to be invalid
and is not processed. Hence the AP continues to operate in its current
state. To resolve this, the NULL AFC response should be treated as if
there are no SP channels available in that geo-location and should be
processed as follows:
A) If there are no Tx channels available(in all power modes) then the AP
must be brought down.
B) Else AP must choose the best channel in the best power mode.
To implement above cases, remove all the checks in the driver that tries
to block the NULL response and finally call upper layer functions
(hostapd/acs) so that they can decide whether to bring down the AP or
choose a channel from the available power modes. Modify
'reg_set_ap_pwr_and_update_chan_list' to bypass setting power type when
ap_pwr_type is REG_CURRENT_MAX_AP_TYPE. This allows the current channel
list to be updated and the upper layers to be informed about this update.
Change-Id: Id048fa9ee402d537f352e3caa2b53729518ea37e
CRs-Fixed: 3565567
Currently, We are using WLAN_MLO_MAX_VDEVS to check
max number of links, However with n link mlo we need
WLAN_MAX_ML_BSS_LINKS to check max number link.
Change-Id: Idd8a96ae80030c813bbdcebcdc4c76d2f93bea5d
CRs-Fixed: 3567342
When CONFIG_QCOM_IOMMU_UTIL is enable in vendor config, we have to
use new set of API's to get iommu domain attributes, and while using
__qdf_iommu_attr_to_os seeing compilation issue as some identifiers
like, QCOM_IOMMU_MAPPING_CONF_S1_BYPASS are not declare.
even when CONFIG_QCOM_IOMMU_UTIL is not enable we have dependency
on qcom-iommu-util.h file as it has some identfiers which are in use,
with or without CONFIG_QOM_IOMMU_UTIL.
Fix is to include the header file qcom-iommu-util.h which contains
the declaration of all the identifiers used in __qdf_iommu_attr_to_os
and this file is included if MSM_PLATFORM or QCA_IPA_LL_TX_FLOW_CONTROL
is defined.
Change-Id: I82b1c13717d377a5ec4bcb74ff94f4d15fa1f638
CRs-Fixed: 3487912
When handle wmi_mlo_link_disable_request_event in tasklet, mlo mutex is
acquired, assert will happen.
To fix it, let wmi_mlo_link_disable_request_event be handled in scheduler
thread instead of tasklet.
Change-Id: I65b84d0b6dc92a6649925d5844657fa44df1fada
CRs-Fixed: 3563195
AFC request is sent for all the opclass supported in the global opclass
including 80p80. Not all the devices support 80p80 and so, it is
unnecessarily included in the AFC request.
Create a callback function in regulatory to check if 80p80 is supported by
the device. If the support is present then the AFC request can include
80p80 opclass (opclass number 135 in global operating class). If not, then
the AFC request excludes 80p80 opclass.
Include 80p80 opclass in the AFC request based on the devices' support.
Change-Id: I9e72c960fefe6fbdc106bb83fb240d84d8522b80
CRs-Fixed: 3538926
Add change to update monitor destination ring interrupt handling
thresholds from INI.
Change-Id: Ibfced2b809ac2101171e8fe9f58e021a5a1f8ccf
CRs-Fixed: 3569507
Add gActionOUISendSMPSFrameWithOMN to send SMPS frame along with OMN for
specified IoT APs
Change-Id: Ida35779d11f9c6d6340afdbebeb788520fa7afdc
CRs-Fixed: 3568462
Monitor destination ring interrupt processing is configurable.
For improved CPU utilization the batch counter threshold and
timer threshold is updated and also an option is provided to make
this values configurable through INI.
Change-Id: Ibe671927f53f3e8fac7f17540c21a2c7edae427b
CRs-Fixed: 3569505
Currently if OPT_DP filter reserve or add request
is received during IPA pipe disconnect, the request is
not rejected. This could cause a race condition when
WIFI is disconnected while OPT_DP transfer is initiated.
This change handles this scenario by rejecting the
OPT_DP request when pipes are going down.
Change-Id: I8ec9e2e50917ef6ec3a07f9e1833303d94134c6c
CRs-Fixed: 3520707
The dp_tx_tso_num_seg_pool_init_by_id() stub implementation has an
incorrect semicolon, so remove it.
Change-Id: Ie70aafdb83116e3e3fb7bd96223b0c770e80eb7f
CRs-Fixed: 3500205
Notify OSIF about each connected link once assoc VDEV connection
is completed, call the ops callback registered while filling
kernel data structure for each link.
Earlier post connection each VDEV's OSIF callback is called to notify
individually, but this leaves the OSIF unaware of standby link details
as the standby link doesn't have VDEV associated with it.
Change-Id: I027b91d02bdc4412400a7e66f11911da149e672b
CRs-Fixed: 3556378
The ops are called on BSS info to be updated to OSIF and
to call for MAC address update on particular VDEV
Change-Id: Ib7b07e6ea2927c365298555ed6f24377127ddd4a
CRs-Fixed: 3556338
The MLO dev context in VDEV is already made to NULL as
part of VDEV destroy and dereference of MLO dev ctx from
VDEV results in NULL pointer exception.
Change-Id: Idbbd625998aa09194ca533b0c8df2fd8e6d16d76
CRs-Fixed: 3570602
After mlo connection is completed driver send the connect resp event
to kernel. As of now, driver gets the ML param info from the assoc_rsp IE.
For stand_by link, driver don't have vdev. Due to this, driver can't send the
stand_by info to kernel.
Fix is, instead of assoc_rsp IE driver will use mlo_mgr to get the links
information.
Change-Id: I8da762bed94de4ad2b024183c09b3d0a0f1b7ca4
CRs-Fixed: 3530712
In Pebble (11BE qcn6432 chip), the radar-found frequency is
incorrectly computed in the current implementation as
dfs_translate_radar_params is executed.
The translation function (dfs_translate_radar_params) which
alters the frequency offset is needed for pre-11BE chipsets.
For 11BE chips where
wmi_service_radar_found_chan_freq_eq_center_freq
service is advertised by HALPHY, the frequency translation is not
needed as the offset sent by FW in wmi_dfs_radar_detection_event_id
can be directly used.
CRs-Fixed: 3562464
Change-Id: I824b94ca764d6bd62a67feb38ba3ccc966999e90
