Currently WLAN host driver is triggering panic on fw ready time out
on all builds without checking if self recovery is enabled or not.
On end user builds WLAN self recovery is enabled which helps to avoid
rebooting the device in this case which is easily recoverable during
next WiFi turn on. Add self recovery enabled check on fw ready timeout
before calling qdf_bug.
Change-Id: I9c9ee3f4b87118679e79e646fb326cc4afdb1845
CRs-Fixed: 2141066
Rename pmo_ucfg_* APIs to ucfg_pmo_* to match existing
naming convention for component ucfg APIs.
Change-Id: I9fba71ee5d5cf6428f043c7d7782ef585951e059
CRs-Fixed: 2140474
Add max check for probe request length against max length of probe
request buffer to avoid buffer overflow.
Change-Id: Ie0fad7443b2c749c66bb9ad662625a16d3a840c3
CRs-Fixed: 2138555
In lim_roam_fill_bss_descr the nw type for 11b AP is not filled and
thus network type remain 11a for 11b AP. Now this entry is added in
scan cache and when driver try to connect to this AP as the network
type is 11a, the preferred dot11mode is set to 11a for an 11g channel
in csr_get_phy_mode_from_bss.
Now while calculating phymode as channel is 11g and dot11mode is 11a
the phymode is not calculated properly, leading to firmware crash.
Fix the logic to calculate nw type properly in lim_roam_fill_bss_descr.
Change-Id: Ie076d0315a7ee5ec123a3cb9dbd22668648e7af4
CRs-Fixed: 2138824
The Linux Coding Style enumerates a few special cases where typedefs
are useful, but stresses "NEVER EVER use a typedef unless you can
clearly match one of those rules." The tCsrRoamInfo typedef does not
meet any of those criteria, and the underlying struct is already being
used in some places, so replace all remaining tCsrRoamInfo references
with a reference to the underlying struct.
Change-Id: I1da4b04ec92e176fe28b33d30ec7823828c35c8d
CRs-Fixed: 2140349
Per the Linux coding style "mixed-case names are frowned upon" so
rename struct tagCsrRoamProfile.
Change-Id: If88fbb18cd521938105701d96173bae221be8de2
CRs-Fixed: 2140347
Per the Linux coding style "mixed-case names are frowned upon" so
rename struct tagCsrDelStaParams.
Change-Id: I60d6448f4bae15e604a8ea55e72a6070a6ebdfbe
CRs-Fixed: 2140346
Per the Linux coding style "mixed-case names are frowned upon" so
rename struct tagCsrRoamInfo.
Change-Id: I21e6c423408f74603e9b469cace00884274d2b61
CRs-Fixed: 2140345
Per the Linux coding style "mixed-case names are frowned upon" so
rename field enableSapSuspend in struct hdd_context.
Change-Id: I8f5321fa57c01a36df72dec17499969d0848e3dd
CRs-Fixed: 2140344
Change "qcacld-3.0: Add knobs to configure the scan paramters"
(Change-Id: I65a3734d8080d8adfe3e9ffa26905e6adc196197) introduced two
instances of "space prohibited before semicolon" flagged by
checkpatch, so fix these issues.
Change-Id: I5bce6d38d9ef0910f9dd6112971c90bbb3f2496f
CRs-Fixed: 2140342
Currently, host is using the wmi_resource_config instead
target_resource_config. There are two INIT command
method is exposed by WMI, hence to converge on one
use target_resource_config and perform cleanup
for other.
Change-Id: Ic50b1c644ce1e61568c2079627727cad766e02c8
CRs-Fixed: 2138506
Abstract CDS's use of CPU hotplug events by using the new QDF CPU
hotplug APIs.
Change-Id: Id37ef9028f710a55938539c8e00bb63ffeb86495
CRs-Fixed: 2139720
In function NDI peer is not considered leading to peer_mac address being
NULL and API failing for NDI peer. Fix the condition to include NDI
device mode and add handling of NDI vdev in hdd_objmgr_add_peer_object.
Change-Id: I34b467fdab4a92696d6377306a775f06fadfc1a9
CRs-Fixed: 2137423
When SAP stop and ref-STA reconnect event hit together, the reconnect
IPA client connect event add sap_num_connected_sta but SAP stop process
clear client STA without send IPA client disconnect event.
Fix it by moving hdd_ipa_wlan_evt into hdd_softap_deregister_sta which
will also been called when sap stopped.
Change-Id: I7abe53e8e395c01df5471697f3370d8b5ae792fb
CRs-Fixed: 2136199
Successful starting of SAP interface has logging
issue and even in case SAP is restarted successfully
it logs error message. Return after mutex unload
during success path and do not log the error message.
Change-Id: Iaf888a2de2bfcc5d52ce8f57e888699eb66ab8ac
CRs-fixed: 2137675
SME module propagates KeyRSC to MAC/PE module but MAC/PE doesn't
pass this counter to WMA and due to which WMA is not able pass to
next module.
Add a fix to propagate KeyRSC field from MAC to WMA module and further
down in stack.
CRs-Fixed: 2130761
Change-Id: I157a44610e184b5e10d838fbc5d6b810e3efd6db
Scan list memory is leaked when get
scan results failed.
Purge the scan list in failure scenario
as part of diag scan complete processing.
Change-Id: I3705dc3f8ffeb586ba745c46a5215bd3578f75e1
CRs-Fixed: 2137957
Memory leaks have been noted in protocol stack in few of the error
handling cases.
Fix following APIs through releasing the memory in error cases.
lim_process_mlm_auth_req()
lim_process_mlm_set_keys_req()
CRs-Fixed: 2139571
Change-Id: Ie28861dab49ab93d90fde5ebc5fd260d1c1c4eee
In hdd_wlan_startup, an object manager pdev is created, but never
released when startup fails. Release the pdev object as appropriate
under error conditions to close the leak.
Change-Id: I9bfd927e5614d7d8b5c8a2238084c3387db60fd8
CRs-Fixed: 2139597
When caller calls lim_process_mlm_deauth_req() API, it checks for
session's validity before passing it forward. If session is found
to be invalid then it is returning immediately without releasing the
memory. Caller is expecting this API to release the memory in any
error cases.
Release the memory before returning to caller in error case.
CRs-Fixed: 2138983
Change-Id: I800d717cd22a6e2731ef211743e4844852a132a3
Remove local variable str[] to print ssid as str[] is
used only for printing ssid.
Change-Id: I9a4af0637df0d24a5d52a43be5947d74f622801f
CRs-Fixed: 2139375
In function ol_txrx_vdev_handle wma_vdev_attach, along with
vdev_active check for is_vdev_valid to find if vdev already
present or not.
Change-Id: I39cc6d27037a8fec29cd6114067b2dab9a964ab0
CRs-Fixed: 2135737
qcacld-2.0 to qcacld-3.0 propagation
Reduce the beacon interval just before the channel switch.
This would help in reducing the downtime on the STA side which
is waiting for beacons from the AP to resume back transmission.
Switch back the beacon_interval to its original value after
the channel switch based on the timeout. This would ensure
there are atleast some beacons sent with increased frequency.
Change-Id: I37db52713d9e2c78abbb97a7cd0274d9da35bfd2
CRs-Fixed: 1094958
qcacld-2.0 to qcacld-3.0 propagation
Send one action frame along with each beacon tx to reduce failure
of channel switching on sta side.
Change-Id: Idbbe1a4bcb4976101680974214fe85d9886ba7d1
CRs-Fixed: 1094958
