Test: regression test on the board
Bug: 38184704
Change-Id: Ife559125850277865451f92ea862e7ec47fc9da7
Merged-In: Ife559125850277865451f92ea862e7ec47fc9da7
Cherry-picked from master
Signed-off-by: Kevin Rocard <krocard@google.com>
Bug: 63911257
Test: mm
Test: cts binaries for sample drivers on walleye
Test: vts binaries for sample drivers on walleye
Change-Id: Ifc1f0381e5ac8d63152ad18635dafced10aadfe8
The files in arm/arm64 are symlinks and pinning symlinks is not
supported. Pin the target vdex file intead.
Bug: 73990433
Test: manual
Change-Id: I1edadcc3be6d88cc07c599cba024babc5abe0633
It's causing surfaceflinger denials and does not exist on other
devices. Grant kernel read access to vendor/firmware's new type.
denied { search } for comm="surfaceflinger" name="firmware"
scontext=u:r:surfaceflinger:s0 tcontext=u:object_r:vendor_firmware_file:s0
tclass=dir
denied { read } for comm="surfaceflinger" name="a530_pm4.fw"
scontext=u:r:surfaceflinger:s0 tcontext=u:object_r:vendor_file:s0
tclass=file permissive=0
Test: boot Taimen without denials.
Bug: 68213100
Change-Id: I8b070a0aae59e12391c881cec8a46b6b4dbe1c67
Add wcnss_filter to Bluetooth HAL section of device.mk.
This is done for three reasons:
1. wcnss_filter is part of Bluetooth HAL.
2. Once wcnss_filter is added to device.mk, entire BT section in
vendor/qcom/msm8998/proprietary/device-vendor.mk can be removed.
3. Switching to QCOM BT HAL becomes easier, just modify device.mk.
Bug: 73968979
Test: Manual build and test with BT A2DP headphones
Change-Id: Iecb725eb360bda624e98b21db154ebf3e9e17381
It's causing surfaceflinger denials and does not exist on other
devices. Grant kernel read access to vendor/firmware's new type.
denied { search } for comm="surfaceflinger" name="firmware"
scontext=u:r:surfaceflinger:s0 tcontext=u:object_r:vendor_firmware_file:s0
tclass=dir
Test: boot Taimen without denials.
Bug: 68213100
Change-Id: Ib5e1187a09ba59907c29e3de51f7189d25d42b49
Add dependency to install move_widevine_data.sh to /system/bin.
The script is only needed for existing devices that
are running Widevine DRM and are upgrading to Pi or later
Android releases.
Test: Upgrade from O-MR1 to Pi, Netflix & Play Movies
can play back offline content downloaded in O-MR1
Test: adb shell to verify script is in /system/bin
Test: adb shell to verify widevine data is moved to
/data/vendor/mediadrm
Test: use audit2allow to verify no avc:denied for widevine drm service
bug: 73656300
Change-Id: Ie42a8229c4d20f71725ba7e8f5c2b267976ed294
This denial seems to be caused by a race with the process that labels
the files. While we work on fixing it, hide the denials.
Bug: 68864350
Test: Built policy.
Change-Id: I3dc7f1a27714d81a42109d46b31b368c36e7fcff
The version actually being served is 1.1
Bug: 73772212
Test: vts_treble_vintf_test --gtest_filter=VtsTrebleVintfTest.ServedHalsAreInManifest
passes on walleye.
Change-Id: I84da55d2d1528afe4a8a3e9856b87c0ea15dd5f8
Rename proc.uidcpupower to sys.uidcpupower so system server can
read and set the property
Test: adb shell settings put global sys_uidcpupower 1
verify via getprop sys.uidcpupower
Change-Id: I6b78e755a444f6dce048f06f053d3002aa060b64
AtFwd hal is implemented as an android service and
it stars from system partition, that is, it's not
a traditional vendor hal and hence it should be
declared in the framework manifest and not in the
device manifest.
This commit introduces new device specific framework
manifest (framework_manifest.xml) and moves AtFwd hal
from device to framework manifest.
Bug: 68662838
Test: vts_treble_vintf_test atcmdfwd service is served
Change-Id: I9df05206fbe067ddf19a96fc52aba6b2cc922172
(cherry picked from commit b2895ddbcf1cc0b8986b4a80a43a3f0b6eb9fd66)
