`devname_value` is a local variable; if `property_get` succeeds, we'll
return a pointer to it. Returning a `std::string` instead sidesteps this
problem.
Bug: 190757198
Test: TreeHugger
Change-Id: If9ca733dd21128706f2a9f62e8460b1286631aa5
Merged-In: If9ca733dd21128706f2a9f62e8460b1286631aa5
Add skeleton on HAL for Gnss raw measurement injection
Bug: 190757198
Test: manual test
Merged-In: I9b58043d5ed321aa71ff4f23031df251ae89c407
Change-Id: I9b58043d5ed321aa71ff4f23031df251ae89c407
Using the current user for the command to disable
BT before the test and enable after the test.
This ensures the proper command is used for
devices running in secondary user
Bug: 202873250
Test: run vts -m VtsHalBluetoothV1_0TargetTest
-t PerInstance/BluetoothHidlTest#InitializeAndClose/0_default
Change-Id: Ica83f4959a87b8fb498b535572738e4e790c6ddb
V6 and V7 use actual device audio configuration
when testing I/O streams functionality. On low-end
TV devices this can cause the test to run for
significant time due to lower CPU speed and the
amount of multichannel configurations.
From testing on real devices, the timeout value
of 5 minutes seems to be adequate.
Bug: 211945822
Test: atest VtsHalAudioV6_0TargetTest
Change-Id: Ie233e014f9f3650f58319c4d5f5a836cfdea87ba
(cherry picked from commit 57633e5a78)
Merged-In: Ie233e014f9f3650f58319c4d5f5a836cfdea87ba
Certain OEMs' new S launches with vendor freeze cannot implement
multi-camera API.
Bug: 189053514
Bug: 211709637
Change-Id: I71b46f1e8343a5848cbe3fa55bbcd427c87b05dd
(cherry picked from commit 30a198ded0)
(cherry picked from commit 2b2846855f)
Certain OEMs' new S launches with vendor freeze cannot implement
multi-camera API.
Bug: 189053514
Bug: 211709637
Change-Id: I71b46f1e8343a5848cbe3fa55bbcd427c87b05dd
(cherry picked from commit 30a198ded0)
Add atrace hal to the readtracefs group
Bug: 214591300
Test: adb shell perfetto (Use config with atrace_categories)
Change-Id: Id9e06f88539b59480d5cf57a4ba67cef4676c1d5
Merged-In: Id9e06f88539b59480d5cf57a4ba67cef4676c1d5
run VtsHalGraphicsComposerV2_2TargetTest will timeout failed
when there is dual displays. use `std::unordered_set` Display type,
`std::vector<Display>(mDisplays.begin(), mDisplays.end())` as
getDisplays return value, mDisplays elements order may not be
same as the order of hotplug events during boot matters.
We should store `mDisplays` as `std::vector`, since the order of
hotplug events during boot matters in determining the primary display.
Bug: 209409863
Test: run vts -m VtsHalGraphicsComposerV2_2TargetTest
Signed-off-by: Kui1 Wu <wukui1@xiaomi.com>
Change-Id: I628ad593640f28fb8a916822a91ac2c578cc299d
Update the VTS tests so that attestation keys are not created with
another purpose.
Bug: 197096139
Test: VtsAidlKeyMintTargetTest
Merged-In: Ib6e4ad98cbe5c3015138854679b11fa0e683ade9
Change-Id: Ib6e4ad98cbe5c3015138854679b11fa0e683ade9
Ignore-AOSP-First: cross-merge from aosp/master
V6 and V7 use the audio policy configuration file for generating
the set of test profiles. The generator must only take into
account profiles of mix ports that are routed to device ports of
attached devices. Otherwise the HAL can rightfully refuse
to create a stream for the profile when the device isn't attached.
Bug: 205808571
Test: atest HalAudioV6_0GeneratorTest
Test: atest HalAudioV7_0GeneratorTest
Test: atest VtsHalAudioV6_0TargetTest
Test: atest VtsHalAudioV7_0TargetTest
Change-Id: I798f978c1c0bc32831c20e4b08338b462d8eea16
Merged-In: I798f978c1c0bc32831c20e4b08338b462d8eea16
The profiles generator wasn't loading correctly from the Audio
Policy Manager config file profiles with unpopulated attributes,
which is typical for dynamic profiles.
Bug: 204314749
Test: atest HalAudioV7_0GeneratorTest
Change-Id: I514a4e03da165cacb5dbaaa16470130895681484
Merged-In: I514a4e03da165cacb5dbaaa16470130895681484
Test that specifying RESET_SINCE_ID_ROTATION results in a different
unique ID value.
Test: VtsAidlKeyMintTargetTest
Bug: 202487002
Change-Id: I2aed96514bf9e4802f0ef756f880cac79fa09554
Ensure that if TEST_PATTERN_MODE is listed in physical request keys, the
corresponding physical camera must support the privacy related test
pattern modes.
Test: Run Camera VTS with Cuttlefish emulator
Bug: 204108650
Change-Id: I8f378ab642c7c010a2ba73a8e89e65c91ba780eb
The multi-camera Grf requirement is now only applicable to rear facing
camera.
Test: Run test on Pixel devices
Bug: 204252005
Change-Id: Idb7735eebdc104f6ee6a5946b50c11aea809eb85
As per documentation the default auto rotate&crop mode must
be overriden before passing the capture request back to the
camera provider.
Bug: 204407427
Test:
atest
VtsHalCameraProviderV2_4TargetTest:PerInstance/CameraHidlTest#processCaptureRequestPreview/0_internal_0
atest
VtsHalCameraProviderV2_4TargetTest:PerInstance/CameraHidlTest#processMultiCaptureRequestPreview/0_internal_0
atest
VtsHalCameraProviderV2_4TargetTest:PerInstance/CameraHidlTest#processCaptureRequestBurstISO/0_internal_0
atest
VtsHalCameraProviderV2_4TargetTest:PerInstance/CameraHidlTest#flushPreviewRequest/0_internal_0
Change-Id: I1e1d387a04add8be9a281fcc3f78a867990248b3
The flags field in UserInfo is a int32_t that contains multiple
UserFlags 'or'ed together. We should not parse it as enum.
Test: atest android.hardware.automotive.vehicle@2.0-utils-unit-tests
Bug: 202520478
Change-Id: Ie7e81a8a5f39f6070e35f2e77bce88a211fd526b
Fix VTS for getAllowedNetworkTypesBitmap and
setAllowedNetworkTypesBitmap, which should not rely on
getRadioHalCapabilities.
Bug: b/199809900
Test: atest VtsHalRadioV1_6TargetTest
Change-Id: I0b50e7e6c0d406a62f7a1e2ea8ac9bc3a71d8aeb
Merged-In: I0b50e7e6c0d406a62f7a1e2ea8ac9bc3a71d8aeb
This reverts commit 6f36dd6e31.
Differing only by vsync period was not a requirement on Android R, and
it shouldn't be added after-the-fact to an existing HAL version.
There are devices on the market that switch seamlessly between configs
that differ in resolution/dpi, and therefore have those configs in a
common group.
Bug: 200184776
Merged-In: Id0bfc67e55d5139fddb2b359cabafd9281c33734
Change-Id: Ib6e6a8c5f45c9829ea30d031e1cfa9a4fb775535
In some implementations P2P device interface is created under
primary interface(wlan0 by default). In those implementations
p2p device name is predefined in system property wifi.direct.interface.
And the interface is created by supplicant with primary interface as
the parent interface. The naming of p2p device interface is
p2p-dev-<parent interface> ("p2p-dev-wlan0").
With STA+STA feature, wlan0 interface gets deleted in certain scenarios
and wlan1 becomes the active interface. In such scenarios P2P fails to
create the interface as parent interface wlan0 is deleted.
To fix the issue update the p2p device interface from system property
based on the current active wlan interface. ie First get the p2p parent
interface name from p2p device interface name set in property.
Check if the parent interface derived from p2p device interface name is
the current active interface. If not, get the current active interface
and update the name as p2p-dev-<ifname> ("p2p-dev-wlan1").
This helps HIDL/supplicant to get an active wlan interface from p2p
device interface name and succeed in creating the p2p interface under the
active interface.
Bug: 203434193
Test: Manual - Tested STA+STA which ended up deleting wlan0 interface.
Then ran p2p tests and confirmed that p2p scan and connection
works.
Change-Id: I522cec02a662c057e21d434f3ed98c7e7a4ca8f0
Merged-In: I522cec02a662c057e21d434f3ed98c7e7a4ca8f0
(cherry picked from commit 7f2822aff7)
This CL makes following changes to fix VTS failures reported on Seahawk
reference platform.
* Correct the definition of RawStreamConfig and its size constant.
* Update all test cases to use a stream configuration reported by the
target camera device.
* Update CameraStreamExternalBuffering test cases to take internally
allocated buffers into account.
Bug: 192460757
Bug: 199626993
Test: atest VtsHalEvsV1_1TargetTest
Change-Id: Ia03775ae543617ba5057e91bbbb5aed9221d1a30
10s is not enough for low-end devices, extending waiting time to
avoid false alarm for low-end devices.
Bug: 201184673
Test: atest VtsHalWifiSupplicantP2pV1_4TargetTest
Change-Id: I9baa53a462b97738e6dc471cf06c2b9230b92c1c
(cherry picked from commit ef3f77f831)
Stopping entire framework can cause other essential services to be
stopped. When wifi is stopped, it does not interact with any of the
wifi HAL's.
Bug: 201266536
Bug: 201184673
Test: atest --iterations 10 \
VtsHalWifiSupplicantV1_0TargetTest \
VtsHalWifiSupplicantV1_1TargetTest \
VtsHalWifiSupplicantV1_2TargetTest \
VtsHalWifiSupplicantV1_3TargetTest \
VtsHalWifiSupplicantV1_4TargetTest \
VtsHalWifiSupplicantP2pV1_0TargetTest \
VtsHalWifiSupplicantP2pV1_2TargetTest \
VtsHalWifiSupplicantP2pV1_4TargetTest
Change-Id: Ia4a38c2e942681f323cf76941713c429e14870cc
(cherry picked from commit 3a5858a711)
cp: aosp/1428863 to sc-v2-dev branch as b/202788915 mentioned
This used to be present on the old host side VTS tests, port the feature to
target side since host side VTS tests are deprecated.
Also,
i) Added a separate test: VtsHalWifiV1_4TargetTest to test the wifi
chip methods. Putting them in VtsHalWifiApV1_4TargetTest will prevent
these from running on devices without AP feature.
ii) Ensured all the non-NAN, non-RTT tests disable framework for
testing. NAN/RTT tests uses framework to check if the corresponding
package manager feature exists.
Bug: 166529516
Bug: 202788915
Test: atest \
VtsHalWifiApV1_0TargetTest \
VtsHalWifiApV1_4TargetTest \
VtsHalWifiV1_0TargetTest \
VtsHalWifiV1_4TargetTest
Change-Id: I05aab6992277601633a0f926a8262c4c27402e93
Merged-In: I05aab6992277601633a0f926a8262c4c27402e93
The stride was not correct for the fuzzer.
And the FormatConvert have some restructions
on the width and height.
- YUYV
width and height must be even nmber
- YU12
width mush be divisible by 16
height must be even number
Bug: 202641239
Test: FormatConvertFuzzer_copyYV12toBGR32
Change-Id: I45ebea3e22854bdad037abb742fbdbe364b19ec5
