| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
| |
Change-Id: Ia5473127cb182222071a97724abc35d36eb7b496
|
| |
|
|
| |
Change-Id: Ibf8540f6dcc130caee51d5f2d1bd5d72e04fcc76
|
| |
|
|
| |
Change-Id: I02ac0011792131de2231c9f84b559fa45886711b
|
| |
|
|
| |
Change-Id: Ic73bd6dae2e292f16c89c417b29c05c695491962
|
| |
|
|
|
|
| |
* Based on stock
Change-Id: I3fc845515894263752edc6596d0aece2ca49a94f
|
| |
|
|
|
|
|
|
|
| |
Tune the BMPS listening interval for the moderate
power level to achieve power savings during wifi
idle.
Change-Id: Id6b00f8ba663b33de5fcae9d03370395d45b2134
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
|
| |
|
|
|
|
|
|
|
|
|
| |
If active mode offload(gActiveModeOffload=1) is enabled then all applicable
data offload/filtering is enabled immediately in FW once config is available in
WLAN driver and FW caches this configuration accross suspend/resume.
CRs-Fixed: 880416
Change-Id: Ib0207cee6a1812a81354f638e6bf55441d52c6e1
Bug: 31537089
Signed-off-by: Srinivas Girigowda <sgirigow@codeaurora.org>
|
| |
|
|
|
|
|
|
|
|
| |
In present code, RPS is enabled in RX path which is causing
inconsistency in TCP DL throughtput case. Enabling Rx thread
Which is providing consistency in throughtput with minimal
power impect.
CRs-Fixed: 1018382
Change-Id: Ieac0fec9399375b723c0f350639f861bdbc143a2
|
| |
|
|
|
| |
Bug: 23556386
Change-Id: Ia8adb569b7899be99e3ee7452dcb1b2a73683925
|
| |
|
|
|
|
|
|
| |
Modify gMaxLIModulatedDTIM and gEnableModulatedDTIM.
Bug: 28106261
Change-Id: Id3ae5d57a52cfc72e19f16cdca044131ffd57ef8
Signed-off-by: Srinivas Girigowda <sgirigow@codeaurora.org>
|
| |
|
|
|
|
| |
* Matches stock
Change-Id: Iebd028437fe91962cdd6f569f6b1026606afe310
|
| |
|
|
|
|
| |
* This is done in system/core/rootdir/init.rc
Change-Id: I0d82d528493fa8c2b0e6bd0769966a205c01b090
|
| |
|
|
|
|
|
|
|
|
|
| |
Configures wpa_supplicant to use p2p0 mgmt iface instead of dynamically
created p2p-p2p0-XYZ virtual interfaces for P2P group.
Bug: 65524423
Test: CtsVerifier WifiDirect
Test: Manual verification
Change-Id: I70582f114ab69684424aac41fbeb78fa7cc04847
Merged-In: Id9add59aa122adb7f39c80252061a8f5d3cc55cd
|
| |
|
|
|
|
|
|
|
| |
* Setup legacymm livedisplay HAL too
* Even without a calibration XML, we can get HSIC control
This partially reverts commit 516bc1a1e80dd9468c327343d74d6c79511deef7.
Change-Id: I560bc5a69384eb766e339f8657657ade00abd229
|
| |
|
|
| |
Change-Id: I70ca90a0c8f657ccc45a1afda7ebea45f5df7cf1
|
| |
|
|
|
|
| |
* If we're going to keep this around, it might as well be valid
Change-Id: I0578379e87a1aa3150a758f5648f4c260ea4a0e5
|
| |
|
|
|
|
|
| |
* Copy them directly to vendor/firmware instead
* While we're at it, remove unused WCNSS_qcom_wlan_nv
Change-Id: I3a7df2fbbad90e5a07838eaf99453e4acdd9e1f1
|
| |
|
|
|
|
| |
* Volume up events still report just fine without this
Change-Id: I28584893534b7d3c301e710b51b37364f2fdad7b
|
| |
|
|
|
|
|
|
| |
* This is based on the following CAF patches:
- https://review.lineageos.org/201884/
- https://review.lineageos.org/201885/
Change-Id: I2e666bbe703f6ad82980296cbba98fd0175c5270
|
| |
|
|
|
|
|
|
| |
* Everything in this change is now global
This reverts commit 9e6b73c1f4d25ab99cdbde5464369e407ed77530.
Change-Id: I9f3b3796942a5826b02fa1e1904fd962f7327976
|
| |
|
|
|
|
| |
* The common impl supports this sysfs now
Change-Id: Ic93f9e7565dd6c12cc8ef50385b3e458319dd9f1
|
| |
|
|
| |
Change-Id: Ie0eda66034bbf37765c248b43d2da5ad8ea0ed9a
|
| |
|
|
|
|
|
|
| |
rmt_storage should not be automatically killed by Android during
shutdown since it is needed to actually process the peripheral
shutdown
Change-Id: I58a29fe86955f0abad944727bb4208dad89f495d
|
| |
|
|
|
|
|
|
|
| |
* Remove cpusets for oneshot services
* Move qseecomd into foreground cpuset
* Fix creation of camera-daemon cpuset
* Fix comment about cpusets
Change-Id: I7f92d0bd96239ff10559277a6fce30980c455f40
|
| |
|
|
| |
Change-Id: Ic4fbe88be1a06540e571fb1414173001f345a2ef
|
| |
|
|
|
|
|
|
| |
* Remove all rules and rewrite policy from scratch.
We had some unecessary rules here that were leftover
from before we updated to nougat blobs.
Change-Id: I7f97a696ce642bcc340e3eae442ce867af18ca02
|
| |
|
|
|
|
| |
* This is now supported by the frameworks as of 222262
Change-Id: Ia2b3995bce3a52feec561a7f3ab6bd898a2fa73b
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Fix battery bar level thresholds to make more sense
* Fix battery bar level order to be left-to-right
* Cleanup various spacing and style issues
* Remove unneeded notice file
* Remove dead/useless code
* Simplify existing code
* Cleanup unnecessary comments
* Update a few comments
* Use O_WRONLY in write_{int,str}
* Check for alloc failures in get_scaled_duty_pcts
* Update deprecated version_{major,minor} flags
* Update copyrights and author
Change-Id: I710d4da7f935205beec0e357d5de05b55b687882
|
| |
|
|
| |
Change-Id: Iecbefe61ffece88a6866ff2c8a1e79d99e535208
|
| |
|
|
|
|
|
|
| |
* This is used in init, even if it's not used in blobs
This partially reverts commit 27773579159879b4d326b17c48f178d52b085214.
Change-Id: Ibbe4d81913fb0f014bd1fdf0b4d27fa1f74e7154
|
| |
|
|
|
|
|
|
|
| |
On device bootup, ASAN was reported IPACM has the
heap-use-after-free issue on IPACM_Netlink.cpp file
when bootup. The fix is to move the debug prints
before the events are getting posted and processed.
Change-Id: Id0a1b4e85d2d3fb027729ff6e98e25d26a9e638a
|
| |
|
|
|
|
|
|
| |
Fix the security issue in IPACM ConntrackClient.
Change-Id: Ia586d9916fc6391ffce436fba9b1ceae1220bc48
Signed-off-by: Skylar Chang <chiaweic@codeaurora.org>
Acked-by: Shihuan Liu <shihuanl@qti.qualcomm.com>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This is now enabled by default in the kernel QCA
driver.
For proper function, we need to enable
support for TDLS external control in
wpa supplicant too.
Change-Id: I79144457b55d98321034929677d72ecd20bc7872
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch enables runtime PM for the qcacld-2.0
driver which supports it since it is inline
with LA.UM.5.5 since LA.BF64.1.2.3.
Since the default delay in runtime PM
(500ms) is very aggressive and leads
to wifi instability in some cases,
we increase it by 1000ms to 1500ms.
That delay provides power savings
without impacting stability.
Change-Id: I254c2f990cf111dc54015629ecd3281cefcd6208
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This addresses the wifi instability.
The problem is that the full QPower mode
(value 2) is too aggressive and causes
the driver to be in a bad state even when
it is needed.
As solution, I've tested the value 1 which
represents the legacy power management
mechanics and the value 5 which represents
duty cycling QPower.
Both fixed the stability issue.
I am choosing duty cycling QPower since
it has a smart logic which toggles
QPower based on an inactivity timer
and has a better power impact than
the legacy power management mechanics.
That matches the value Google uses
for Bullhead.
Change-Id: Ifeef1e367770c974f053371d4693876780f1cc41
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Enabling Beacon Early Termination in order
to speed up beacon processing drastically.
This will lead to power savings.
Also tune the BET wake interval to a value
of "8" to reduce the frequency of wakeups.
"8" is a safe choice since I could detect values
of higher than "10" to cause issues in some usecases.
Change-Id: I943f01b7599d8f7e42b1bc9249b8fc9b783e4345
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
|
| |
|
|
|
|
|
|
| |
Implement a smarter system on whether to connect
to the provided 2Ghz or 5Ghz network.
Change-Id: Idd41e3d9a5e2bde3ea9b4873dffc4274d1530324
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
|
| |
|
|
|
|
|
|
|
| |
Setting the beacon missed count threshold
with a difference of 10 to improve stability
of the WIFI connection.
Change-Id: Iea459ead3505bcea1a41f9b0e713957092bf4745
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
|
| |
|
|
|
|
|
| |
* Our notifcation LED supports pulsing
* While we're at it, update the comment for new caps
Change-Id: I272dea5484bc004d559ba8318f5e309d1b109a20
|
| |
|
|
|
|
|
|
| |
* We're not getting any more updates, we don't need that
* We need to keep the qseecomd hex editing so extracting
from lineage zips from a clean slate will work
Change-Id: If9b884bcfae21ecd8ff86b097d06d4b9f995e95a
|
| |
|
|
|
|
|
|
| |
* Move all services to system-background
* Add camera-daemon cpuset
* Remove foreground boost cpuset
Change-Id: Ife69bbd413208f11c273c080b1beff901921e8f4
|
| |
|
|
|
|
|
| |
* These don't work, and they didn't work on stock
* While we're at it, remove the unused large window prop
Change-Id: I9cfcff48ecacdc9a1f085e60b8e3c01f55e05284
|
| |
|
|
|
|
|
| |
* qcom_diag -> oem_2950
* While we're at it, move it to vendor and under IMS
Change-Id: Id1a8858247a1e9cb38ebcf43c57ba3e98b4c4d06
|
| |
|
|
|
|
|
|
| |
* qti-telephony-common is crashing, and nothing in CNE or IMS
references any of these blobs anyways
* Nothing references the intents in imssettings.apk
Change-Id: Ia787eb51da7ff86c3ec9cb17c36a07c0484ea3d2
|
| |
|
|
|
|
|
| |
* Just so all our kanged blobs are from the same place
* Package version: OPM1.171019.011-RZR-180509.5038
Change-Id: Ib8ad272a5ef50cf19708aeeb02fcb3deb5f06f7e
|
| |
|
|
|
|
| |
* Package version: OPM1.171019.011-RZR-180509.5038
Change-Id: If1989b07174393491be3767f2c56bf1abaed2f5a
|
| |
|
|
|
|
| |
This partially reverts commit 4e090d4ecd69142de901000ea6ca47da6e01fd3a.
Change-Id: Ic96711cb64b264d38b37db7873fe4cb1b9eb14f2
|
| |
|
|
|
|
|
|
| |
* HDR mode in HAL3 does nothing
This reverts commit bd27d0093df32553d11243c0bb8fafcd5209aa84.
Change-Id: Iaa1a6d5e5383c60e53ade15afdc2e457d0ad6ca3
|
| |
|
|
| |
Change-Id: I6bb12aad39d42b76927ea4c97ec640c1f001ec3c
|
| |
|
|
|
|
|
| |
Set TARGET_USES_MKE2FS := true to use new EXT4 tool chain
Bug: 62421233
Change-Id: Id6e5c68b2d7854e5330ad87443b974361a4bff9d
|