| Commit message (Collapse) | Author | Age |
| ... | |
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This reverts commit e2f744f859654a1ec69d90a573ae3a90d36bd632
with change id I637cf0ee48700859512feb678119bd158b8e2c0f.
This change is already present in mainline.
Change-Id: I756821ee6e61da57f0c1aba18c51dcdedd6a50b3
CRs-Fixed: 873860
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Global configuration of channel bonding is currently
getting modified during the SAP bringup due to which
STA is falling back to SAP configured channel bond mode.
As part of this change a new config parameter is
created for STA mode and it will not influenced by
SAP.
Change-Id: Ib82f71aafc59b63861f40fb5acd79f195237fc54
CRs-Fixed: 872231
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The rssi_raw should not updated when the channel
in probe response does not match the channel in BD.
Change-Id: I517947c93bec32d9913b0ee85a1bc41e4d22d7d1
CRs-Fixed: 873549
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Add a new attribute to notify the driver whether to use the
current EPNO request as PNO or EPNO. Driver can take decision
based on its value.
Change-Id: Iecd57e1a0a794665691197cc9e10da14d4d858a9
CRs-fixed: 861059
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Release 4.0.10.141
Change-Id: I7b2c56c26618594301e075f5d3d66cada84d9728
CRs-Fixed: 688141
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Add WMI_SCAN_FLAG_PASSIVE macro.
Change-Id: Ia699c6fce54c56b6ca978c84843343c6034454f2
CRs-Fixed: 865207
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
In wma_vdev_start() pmac extracted from vos context is not checked
for NULL and is being dereferenced at a later stage.
Fix issue by NULL condition check and return failure if pmac
is NULL.
Change-Id: I77399354a988244f9703fc3cb84ba9d5fd6b1b9b
CRs-Fixed: 873939
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
In function lim_post_join_set_link_state_callback(), if state is not
eLIM_MLM_WT_JOIN_BEACON_STATE function should not do anything, but due
to a missing return statement it is being treated as failure and
LIM_MLM_JOIN_CNF is sent to upper layers and MLM state is also messed up.
This further causes Auth Rsp to be dropped at LIM and hence SME active
cmd list gets stuck followed by crash.
Change-Id: Ida92c43f7a6eedad8b99fd2d8a7b78ae08b91552
CRs-Fixed: 869308
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Use message bodyval to pass thermal level which is of type uint8. This
avoids any memory leak and pass the thermal level in an optimum way.
Change-Id: Ie4dfaac68e17ac2a0d2a3433d8595a928ca66ea8
CRs-fixed: 869091
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The function hdd_ipa_wlan_evt returns failure after inserting ipa
event into pending_event list during ipa microcontroller resource
loading/unloading. This causes an unexpected stopping of SoftAP.
Fix the return value to 0.
Change-Id: Ic5433086fa1d554dda275de001ec5a1e9eb9262f
CRs-Fixed: 874442
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Router advertisement max interval is 5 minutes which is not
very optimal for power save. Increase max interval value to
1 hour such that users can configure router advertisement
interval more than 5 minutes.
Change-Id: I90b7fb95aee43b8071c6752671fda0c719e3b34d
CRs-Fixed: 874385
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Waiting on VOS event variable without resetting results in stability
issues, if the same code path execute more than once.
Change done to reset VOS event variable before using.
Change-Id: I56b42a70524d8311e4ab3beba65380600cf90cf4
CRs-Fixed: 873543
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Add new option for world roaming that has channels
12/13 disabled. The following specific changes have been
added to add this feature:
1) Correct flags calculation for custom settings. Specifically
correct the flag calculation for DISABLED and DFS flag.
2) For kernel 3.10 and less; correct the wiphy channel
composition when "iw reg set 00" is used; restore the original
settings from wiphy. For kernel 3.18; this behavior is enforced
by the kernel itself.
3) Remove NULL check on wiphy->regd for CORE callback.
Wiphy->regd can be NULL during driver re-load.
Change-Id: Ib1db4c9a8d90e9ea854fbf078401201eebd74b67
CRs-Fixed: 873092
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Fix boundary values, power and naming for world regulatory rules.
Change-Id: I30e73a02f18fb0a721302a1bc410b83716a91340
CRs-Fixed: 873092
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Cafstaging Release 4.0.10.140
Change-Id: Ia5d5916266fc17cfc894b293580534b1b334873e
CRs-Fixed: 688141
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
For maximum RX performance, tlshim rx thread needs to be attached to
high performing core. The change dynamically finds out the highest core
in a multicluster system and attaches the tlshim rxthread to it.
Enable TCP checksum offload and IP checksum offload features in the ini
file. These features help with RX performance.
Change-Id: I3b6124b766751750d59c0864f0086b8cfd7deb22
CRs-Fixed: 874306
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This change contains a fix for a suspicious pointer dereference
before a NULL check.
Change-Id: I855f6c903680208cc715cef7db919b856206c376
CRs-Fixed: 873955
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Some pointers to TLVs can be NULL in link layer statistics.
Update the check on these pointers to return error only when
NULL value is unexpected.
Change-Id: Ibdc69b5034e1470830aa122d47b3456602f35905
CRs-Fixed: 874041
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Support is added for the flags parameter in the set config command
for OCB mode. Currently there is only one flag to set the absolute
expiry flag.
CRs-Fixed: 873181
Change-Id: I97409d1e24e5b4e9abe0c84264074e4d05a0ceac
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
In athdiag_procfs_remove, set proc_dir to NULL after removing the
directory. This is to avoid trying to remove the directory which is
already removed.
Change-Id: I01526f4b59ccc991d157c1028d4df5904338f182
CRs-Fixed: 868965
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Release 4.0.10.139A
Change-Id: I4db6f7711b3753856b5fc08128cc5c1508dd7ee3
CRs-Fixed: 688141
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Fix null check issue observed in static code analysis for
MCC to SCC switch update.
Change-Id: I87155dc5de7c256d2ab900361d45bce1d6ebb1ae
CRs-fixed: 870558
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Fix the possible NULL dereference issue observed in static
code analysis.
Change-Id: I1ecc8c4ed6456ebcfd85fe62534b9913ba31dcd9
CRs-Fixed: 869026
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Release 4.0.10.139
Change-Id: I2c72b80b78d4381da549b87e6bb4adfe4ef47c21
CRs-Fixed: 688141
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Add support for chain mask and NSS configurability
per vdev type.
- Configure the 2g and 5g chain mask to FW based on
INI and FW chain mask value.
- Configure the 2g and 5g Nss with the INI value of
each vdev type
- Program the HT and VHT IE to FW for 1x1 and 2x2 mode,
FW will included them in scan probe request frames.
Change-Id: I5cbf17a14ab6becad6cf5765ae5039fc284dc309
CRs-Fixed: 869026
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Fix a bug where SAP configured with ACS fails to start when
MCC to SCC overrides the ACS channel during ACS selection
procedure.
Change-Id: I9b008c17651e6e6affa37d1688eca179bb8e900a
CRs-fixed: 870558
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
P2P GO also follows start_bss cfg80211 call. But since P2P GO is not
configurable to setup VHT channel width in wpa_supplicant, override
wpa_supplicant configured vht channel width to driver INI configured
Change-Id: I51d25236c532a13e3d12fe89a7636e4c25830715
CRs-fixed: 872322
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
- Hostapd sends WPS configuration durign cfg80211 start bss
- Pass WPS configuration to WMA and in WMA check WPS config
before enabling WPS probe request wake up in FW.
Change-Id: Ia3119aed36ba96f7c859c487d508233099186eef
CRs-Fixed: 871318
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
WMI interface changes per band chain mask support
and per vdev spatial stream support.
Review comments of change Ia04f7f96ea7713be04758c7a5a4645c42ca02461
updated as part of this change.
Change-Id: Ic95efc31a8045fc14ec61e9ed1f759ae7d8bb81c
CRs-Fixed: 865207
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
HT20 protection mode is not reset when a HT20
STA disassociates from SAP operating in 11a mode.
Modified OBSS protection callback handle to
consider all the STA's capability before updating
the AP protection.
Change-Id: I52c0329ff6b24b7b7bf2301e2aa9f5c1e69438d3
CRs-Fixed: 871907
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Release 4.0.10.138
Change-Id: Ibc8fa7725ea9536e0aacf116133e70ce7e588a25
CRs-Fixed: 688141
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Pktlog proc entry creation failed as the entry is
not removed in the previous driver unload.
Add debug prints to find out where we returned without
removing the entry.
Change-Id: I25bc3af07141d93ea4bc18ac65acfb4440b59c75
CRs-Fixed: 868212
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
WMI interface changes per band chain mask support
and per vdev spatial stream support
Change-Id: Ia04f7f96ea7713be04758c7a5a4645c42ca02461
CRs-Fixed: 865207
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Currently while sending command to firmware and waiting for
response we will set gLimProcessDefdMsgs to false, during
this processing of lim request message will be deferred until
we get the response.
In noisy environment we might run out of message wrappers on
receiving lot of beacon frames in deferred state.
Change to drop received beacon during deferred state.
Change-Id: I709574d3a97f89a21500c3fbbb166bb257d26c7b
CRs-Fixed: 870159
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
LFR3 roaming common code sends cfg80211_roamed_bss() and
key mgmt+roam auth event to the supplicant. As per design, assuming
supplicant receives assoc/reassoc indication and roam+auth event
back to back, supplicant (for non-LFR3 roaming) is expected to ignore
roam+auth event and behave the pre-LFR3 way.
But this assumption is not always true, there may be race conditions
which can add delay between supplicant receiving assoc/reassoc
indication and roam+auth event.
In this particular case (non-lfr3 roaming case), assoc/reassoc indication
is received and roam+auth event is received when supplicant is in between
EAPOL exchange using previous KCK/KEK, then supplicant honors this event
and updates its KCK/KRK and replay counter. This makes supplicant to
cannot verify EAPOL Message_3 and drops the packet which triggers
full 802.1x authentication.
Hence the proper fix is, as this roam+auth event is intended for LFR3
roaming, send this event from driver to supplicant only if
roaming is LFR3.
Change-Id: If75de1bb2858f0a167dadf11277c251e2041cda8
CRs-Fixed: 871349
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Release 4.0.10.137
Change-Id: I6b63bb2028ac0bd355a478b54e862a7026225ca9
CRs-Fixed: 688141
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Runtime PM GET and PUT operations have logging but these APIs are
called as part of data path so remove the logs.
Change-Id: I7a493bddd5d49e8dc809cf2779a12b4337e555e9
CRs-fixed: 870374
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
While suspending, pci_disable_device is called to disable the
device which even disables bus master. During resume,
pci_enable_device is called but it doesn't enable bus master.
When bus master is disabled, PCIe is not entering L1ss after
Runtime Resume. Fix the L1ss issue by enabling bus master using
pci_set_master API.
Change-Id: I8f308b969d128631bb4a010f6072c71d06cb1c09
CRs-fixed: 867537
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
prima to qcacld-2.0 propagation
As a part of logging enhancement,to make logs more
interactive, changed pmcstate,scanType etc from
numerical values to human readable string in
Protocol stack.
Change-Id: I589088c2c5fd0afa106a1d0773a2ff525b0dd083
CRs-Fixed: 638916
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Validate arguments for TDLSSECONDARYCHANNELOFFSET command
and return failure on invalid arguments.
Change-Id: I3bda41e331fd9699f2fc33ccec14b5552ea3d733
CRs-Fixed: 856495
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
remove duplicate wmi_peer_type definition (unused)
Change-Id: I8a21000c19322345acbd8c7829524a553e45e6c4
CRs-Fixed: 865207
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Remove check for pending command because it was causing the host driver
to always fail the command if the firmware fails to send a response.
CRs-Fixed: 865864
Change-Id: I96086d6effaf871a9538a2f2696d7ddc6a51cc3d
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The firmware will adjust for the noise floor when sending the RSSI
in the HTT header on RX. The host driver shouldn't adjust for the noise
floor.
CRs-Fixed: 865578
Change-Id: If7af9360562866124e0914a602e7f54afa9185e6
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
When the driver in OCB mode is unloaded, the TX queues are not
properly flushed. This fluses the TX queues.
Change-Id: I1a553a17030c3473e4364ef6d21b39d837beaa7d
CRs-Fixed: 869323
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
When last BSS stopped, disable WDI pipes directly since IPA_OPCODE_
TX/RX_SUSPEND message will not be processed when unloding WLAN driver
is in progress
Change-Id: Ib2a7b82245af495378f90ec3e26cb2a150c9c290
CRs-Fixed: 867801
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
prima to qcacld-2.0 propagation
Currently CfgAgingTimer starts at the time of csrReady.
This timer removes aged out bss. Since aging out is only
meaningful once scan result is populated, this timer
shouldn't start before scan. As a part of fix, start
timer only when scan results are obtained.
Change-Id: I4c6c2fff654351724193667d9587129911d9da8d
CRs-Fixed: 831557
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
prima to qcacld-2.0 propagation
Few peer (GO) devices are toggling the capability bits in the
beacon and probe response frames more frequently. At present the
host will get the capabilities from scan results. But after scan
if the peer device changes the capabilities then host is not
updating its capabilities. This is leading to disconnection.
Change-Id: If986fe7b1a0927200c6e8b6e1b3860cb7bff131d
CRs-Fixed: 803932
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This is prima to qcacld-2.0 propagation
Send correct deauth reason code when keep alive timeout occurs.
Change-Id: I752bdf055e2ad53b4d08786ded448ce9d48b45f3
CRs-Fixed: 772435
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Release 4.0.10.136
Change-Id: I2a51ab6b3af4611a9a5f5a49f31654f788be6bde
CRs-Fixed: 688141
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
prima to qcacld-2.0 propagation
After roaming downgrading is not happening because while tl is fetching
pkt from hdd_tx_fetch_packet_cbk, flag wmmAcTspecValid is TRUE because of
(AddTs done for initial association). This needs to reset during roaming.
So now hdd is making wmmAcTspecValid FALSE at the time of the connection
complete a.k.a hdd_wmm_connect during roaming. Though hdd is not involved
for Add TSpec during roaming still hdd_wmm_sme_callback needs to called
to set this flag.
Change-Id: I235f6b61a9b682244da59845b50535a3d62c5a47
CRs-Fixed: 697180
|