For APsI pld_get_mhi_state() and pld_is_pci_ep_awake() return
proper value for buses other than IPCI where these APIs are not
supported.
Change-Id: I2dfe8118d29aa2bc75484ff4d708b21eba264350
Currently when host receives power stats from fw, before
processing these stats it check for the hdd callback and
processes these stats only if hdd callback is registered.
This callback is checked in
wma_unified_power_debug_stats_event_handler function before
processing of the fw event and outside of any lock. After
processing stats, this hdd cb is invoked, there is an issue
with this approach that while processing these stats if timeout
happens in hdd then hdd sets its callback as NULL and now after
processing stats wma handler invokes this hdd cb which is already
NULL, leading to null pointer dereference.
To avoid above issue, add sme callback and move null check for
hdd cb and call to hdd cb inside sme lock which is the same lock
used while setting hdd cb to null on timeout.
Change-Id: I2203083a17ca3404c5c5af98ddc48ba127e6aee2
CRs-Fixed: 2950633
Currently driver updates the WTC global parameters irrespective
of the WTC attributes present in the vendor command.
This causes the WTC parameters to be reset and WTC roam
trigger doesn't happen.
For example, with below sequence of commands:
wpa_cli DRIVER setwtcmode 0 1 -40 -95 -90 -85
wpa_cli DRIVER ADDROAMSCANFREQUENCIES_LEGACY 19 2412 2417 2422
The WTC parameters TLV is not included in the RSO
update command.
Update the WTC params only if the WTC attribute is present.
Change-Id: I53e5d934368169069dbf059335859936d320aa4e
CRs-Fixed: 2954458
Currently in WTC mode the candidate AP RSSI threshold
configuration is applied for candidates on all band.
Add support to configure the candidate AP RSSI threshold for
different bands.
Change-Id: I94442c0f1bd3ea20dc2f5ebe893a9ead7fb8d890
CRs-Fixed: 2951610
Race condition occurred as rx_thread did not get scheduled
before timeout and gro flush happened in vdev context. At
the same time, rx thread got scheduled and add skb to the
napi queue resulting in crash.
To fix this, Increase the timeout value so rx_thread gets
time to wake up, do gro flush and reset the vdev_del_event.
Change-Id: I2075eda91b1568cce3083869f510fc212dd721f8
CRs-Fixed: 2949803
Currently nbuf is being used to get gso segments after enqueue
to DP RX thread, this will lead to use after free issue because
RX thread may process and even free the buffer by the time nbuf
is accessed for getting gso segments in enqueue/softirq context.
Fix this by updating gso segments before nbuf enqueue to DP RX thread.
Change-Id: I2cc93bf9a44e2d487c1a6d474349cf5c0c5db76a
CRs-Fixed: 2958132
If 2.4G 40MHz is disabled by channel bond mode ini -
gChannelBondingMode24GHz, override 2.4G ACS result
to 20Mhz channel width. Otherwise hostapd will
start ap failed on 40Mhz channel because the HE cap
has no 40Mhz bw supported.
Change-Id: I943621bf847491983f0654f7ae495fc970114575
CRs-Fixed: 2951620
Currently, osif_request_alloc is done and waiting for response
after posting the message NDP_END_REQ. But this may lead to
invalid memory access if message gets processed in scheduler
thread and response also comes before the request is allocated.
Allocate the request before sending message to avoid this.
Change-Id: I88b0a8be229e779526f0f3cf184b191e00872ee8
CRs-Fixed: 2782045
Currently the driver does not inform userspace when
there is no channel available for the SAP to restart.
This case can happen if all the channels in the spectrum
are unsafe, in NOL list, SRD etc. and the driver cannot
find a safe channel.
Fix is to inform userspace about this scenario of
driver stopping the SAP.
Change-Id: I77f7a5156c8f7b1da29023675d5a7a448be0fb92
CRs-Fixed: 2815265
When ipa setup pipes fails in wlan_ipa_uc_ol_init API,
tx buff mapping is not done but during deinit path
tx buff unmapping is called without mapping.
Fix this by resetting uc_loaded flag to false when ipa pipes
setup failure is encountered.
Change-Id: Ic34320e56cee9e60381d9203dd95e268d0ef392b
CRs-Fixed: 2957612
Currently the KRK and BTK Keys are not copied from the supplicant
to the buffer in RSO command. Hence the NULL KRK and BTK keys are
generating invalid MIC key in Firmware for CCX roaming. And so
the CCX roam fails.
This fix copies the valid KRK and BTK keys from Session to the
ese structure in the RSO command.
Change-Id: I2f567f75dd74189becb40ec9a1f906ab0d870584
CRs-Fixed: 2958938
Currently channel switch command accepts only channel number as
parameter but with 6GHz changes channel switch can come for the
6GHz channel as well and to support 6GHz channel switch frequency
needs to be accepted as parameter.
To address above issue add support to accept frequency as
parameter for channel switch command.
Change-Id: I38978348c0d70e3e860d4644b4664f8775ecb777
CRs-Fixed: 2931485
Currently driver does not consider secondary channel list
for ACS which needs to be used for SAP related operations and
because of that in sap get freq list driver is selecting some
frequencies which are not valid for SAP operation and from
frequencies driver selects ACS frequency and gives to upper
layer resulting in an invalid ACS frequency to user space.
Now when user space tries to start ap on this frequency, SAP
doesn't come up on this frequecy as this is invalid frequency
for SAP.
To address above issue consider secondary channel list for ACS
in sap get freq list so that only valid frequencies for SAP
are selected for ACS.
Change-Id: Ic7f1cb8134615a9195884d55bfb8111898ff74cb
CRs-Fixed: 2955162
In lim_ft_prepare_add_bss_req due to change
Ic4a992acfce186f7248dcc99d0c7b44ce38f099b the action was removed
but if condition was not removed. This left over if condition is
causing the vht_caps to be NULL in VHT20 case.
Fix this by removing the left over if condition so that vht_caps
is filled all the time.
Change-Id: I196fd3cfb44fa3281b9843a61242520e0888d2c4
CRs-Fixed: 2958411
When flow is added in CMEM path init time stamp is not
handled, so initialize flow init time stamp in CMEM path.
Add Fisa error print during reo id mismatch, print flow
and nbuf details which helps to debug the issue even if
local stack variables are not available in dump.
Change-Id: Ibcc44bd8558cc687279bb34267d60b9882d33a24
CRs-Fixed: 2946825
Set intermediate EP vote access before host
wake up indication smp2p is sent to F.W,
this helps to avoid smp2p override.
Change-Id: I4b998ee5ea6b26921855bfbae434869fb93accf9
CRs-Fixed: 2953883
When roaming to 160M AP by LFR3, host will update BSS info after
received roam sync event from firmware, but 160M isn't considered
when calculate channel width.
Actually vht_op->chanWidth alone cannot decide the actual BW, we need to
check difference of center_freq0 and center_freq1 to decide 160/80+80 or
80 if vht_op->chanWidth is 1.
Fix it by copying ft_session->ch_width to BSS info which covered 160M
channel width.
Change-Id: Ic4a992acfce186f7248dcc99d0c7b44ce38f099b
CRs-Fixed: 2954900
Currently, the same twt disable api hdd_send_twt_disable_cmd()
sends to firmware for both requestor and responder role.
As part of this change, decouple disable api based on requestor
and responder role as per below:
for requestor: hdd_send_twt_requestor_disable_cmd()
for responder: hdd_send_twt_responder_disable_cmd()
Change-Id: Ib92b0fa46dde78dec0d5b7ca914d018c7c0970e5
CRs-Fixed: 2937591
1. Currently, the same twt enable api hdd_send_twt_enable_cmd()
sends to firmware for both requestor and responder role.
2. When firmware receives enable command, it will be busy to
handle that command. If any back to back enable command, it can
lead to some issues.
As part of this change, decouple enable api based on requestor
and responder role as per below:
for requestor: hdd_send_twt_requestor_enable_cmd()
for responder: hdd_send_twt_responder_enable_cmd()
Synchronize twt enable command to avoid back to back command
issue.
Change-Id: I3d3840740fee7413d6123ece919c42558e9b35d1
CRs-Fixed: 2937540
Add support to dynamically control rx aggregation based
on clsact ingress filter addition or deletion. Disable
GRO/FISA when the filter is configured and enable GRO/FISA
back when there is no filter configured.
Change-Id: I462f62edf0acd191b57f40800118567edb2ac82d
CRs-Fixed: 2952276
Roam offload init command should be sent to firmware only
for STA mode. For P2P vdev, if roam init command is sent,
it causes abnormal target behavior.
Don't send roam init command if the vdev opmode is not
Station.
Change-Id: I0bfc3c6b4bc2eef7459b092101bc4d72c68761e5
CRs-Fixed: 2924563
Currently there is no support to enable NAN on indoor channels,
as per the new requirement add an ini support to enable NAN on
indoor channels.
Change-Id: Ia3acbc133cab4865d8ad69992e893be91439f50f
CRs-Fixed: 2943668
In wlan_hdd_validate_operation_channel, "status" is override
to QDF_STATUS_SUCCESS after "ucfg_mlme_get_sap_allow_all_channels"
call. Reset "status" to QDF_STATUS_E_INVAL before checking
with supported channel list, otherwise QDF_STATUS_SUCCESS
will be returned for unsupported channel case.
Change-Id: Ibc3396eb89634b0efc77262e90427510d26e0f39
CRs-Fixed: 2946475
Currently, there is no panic when there is a net dev reference
leak. To catch the reference leak at the early stage, trigger
panic immediately after reference leak is detected.
Change-Id: Icbdd090dab50c8f695b5354c405f0506adce119d
CRs-Fixed: 2911563
