Sending a null hidl_string from native code to Java is currently not
a supported operation and will cause a crash. As originally written,
shallowCopy would set a hidl_string to null as soon as a non-empty
stringValue was followed by an empty stringValue. Since freeze frames
come with a string value (the DTC) and live frames doesn't, this would
cause all diagnostics to crash while reading VehiclePropValue from parcel
immediately after dispatching a freeze frame. Fix by resetting empty strings
to a global non-null but zero length buffer.
Bug: 37435530
Test: diagnostic_injector.py
Change-Id: Ie79174dd7ab3d266f1e3addefa5ba9e2128ea9f8
Right now there's no way to distinguish multiple clients, thus we cannot
associate IVehicleCallback from subscribe and unsubscribe method. As a
temporary solution we will use pointer to IBinder from underlying proxy
object until HILD will have some API to do that.
Test: verified unit tests run, checked that subsequent 'subscribe' calls
do not create new client objects in VHAL
Bug: b/37254245
Change-Id: I029564b7b03e4821287c2110d0929a5c4cbce0ce
This property is supposed to be used mostly from e2e tests (a simple
test-case is provided in separate CL)
Test: make -j && runtest -x packages/services/Car/tests/vehiclehal_test/
Bug: b/36510399
Change-Id: I09b24f22ab328eee1ef6add60901ed03bf046874
Files relying on transitive include of utils/Log.h (and things that it
includes) from MQDescriptor.h
Test: pass
Merged-In: Iff316b21bef556bb026378b7f89e97ded3febef4
Change-Id: Iff316b21bef556bb026378b7f89e97ded3febef4
Files relying on transitive include of utils/Log.h (and things that it
includes) from MQDescriptor.h
Test: pass
Change-Id: Iff316b21bef556bb026378b7f89e97ded3febef4
Also rename DefaultVehicleHal to EmulatedVehicleHal and
extract emulator related code to VehicleEmulator class
Test: mm -j32 ; verified Car Service worked
Change-Id: I34361fdec6f94629cf7ef6c35ff56ef9ce78b855
Move default values closer to config declration to add properties
with less pain in future.
Test: mm -j32; verified if Car Service works
(cherry picked from commit 24ade17518)
Merged-In: I2f18837658149eb657f44d81e34eb8e6e497a25c
Change-Id: Ie4285581fa9d871b366b0ed2f08fa8073739a0a9
Also, removed per property ACL as it was based on client UID
Test: mm -j ; verified Car Service works
Bug: b/36649684
Change-Id: I7625a013dda8e3ebce5b86a253a8c1befe267c11
(cherry picked from commit 72aebeb3ec)
Move default values closer to config declration to add properties
with less pain in future.
Test: mm -j32; verified if Car Service works
Change-Id: I2f18837658149eb657f44d81e34eb8e6e497a25c
Also, removed per property ACL as it was based on client UID
Test: mm -j ; verified Car Service works
Bug: b/36649684
Change-Id: I7625a013dda8e3ebce5b86a253a8c1befe267c11
Use this infrastructure to move OBD2_LIVE_FRAME support over to the Vehicle HAL 2.0 layer.
This is useful in DefaultVehicleHal 2.0 to let the Emulator layer send/receive information about properties defined by later VehicleHal versions.
Additionally, support the notion that a custom property thusly defined might have logic more intricate than simply a property Id and an area Id,
via a custom handler object that supports get() and set() operations. A default handler object simply falls back to a stored VehiclePropValue, but
more complex implementations are possible.
Test: manual testing in Python:
import vhal_consts_2_0 as c
from vhal_emulator import Vhal
v = Vhal(c.vhal_types_2_0)
v.getConfig(c.VEHICLE_PROPERTY_HVAC_TEMPERATURE_SET)
print v.rxMsg()
v.setProperty(c.VEHICLE_PROPERTY_HVAC_TEMPERATURE_SET, c.VEHICLE_ZONE_ROW_1_LEFT, 70)
print v.rxMsg()
v.getProperty(c.VEHICLE_PROPERTY_HVAC_TEMPERATURE_SET, c.VEHICLE_ZONE_ROW_1_LEFT)
print v.rxMsg()
v.getProperty(0x11e00d00, 0)
print v.rxMsg()
plus
runtest Obd2LiveFrame.java
Change-Id: I9c342e6e2c9ff581beacd238a42780ca10f96d9c
0xdeadbeef is exactly 32 bits, so it is interpretted as an unsigned
integer which was causing a warning. Since it's a test value, I changed
it to 0xdead which is interpretted as a signed integer.
Test: pass
Change-Id: Ibb75b23461700d77158e2337b8a472e87629ad61
Implemented RecurrentTimer class that can be used in VHAL
implementations to trigger an event every X amount of time.
This make sense for continous properties.
Also added linkToDeath for VHAL listeners.
Test: unit tests added and ran
Change-Id: I4a2e0607ff6b15b3e7babbe3fbb9fff8d4e28838
Test: build and flash Mojave board with TARGET_USES_CAR_FUTURE_FEATURES=true.
Test: check that the property is displayed in the logs during startup.
Bug: 35628104
Change-Id: I8b5327ae255cd94ac78a0436aac06b32e3168c28
Test: verified events are now gets propagated to CarSensorManager
(toggle day/night mode in Emulator UI)
Change-Id: Ifc4a755c5f6d31f41ab26ebc873c66c43d9e234a
Fix: b/35926101
Add properties to HVAC_POWER_ON config string
Return NOT_AVAILABLE for properties in HVAC_POWER_ON config string when
power is off.
Change-Id: Icfa06aa169345e973d2e152aa6dbd8c14f3a5d17
- Create base communications interface
- Refactor socket calls to use SocketComm class
Test: Use python scripts and custom emulator to test communications
Change-Id: Ia401587223035e748991516a2285cc31cb71a9c9
the getService() and registerAsService() methods of interface objects
now have default parameters of "default" for the service name. HALs
will not have to use any service name unless they want to register
more than one service.
Test: builds
Bug: 33844934
Change-Id: I1fc92a29178befc1b2232cfcc4b13058d57cd798
Major changes / decisions:
- created empty IVehicle@2.1 interface to distinguish HAL impls
from client's (Car Service) perspective;
- created default 2.1 implementation that currently delegates all
calls to existing 2.0 implementation
- had to refactor a little bit 2.0 thus components could be esily
imported in 2.1
TOOD:
- move OBD and VMS code to 2.1
- decouple "DefaultVehicleHal.cpp"
- rename "DefaultVehicleHal" to Emulator (let's call things by their names)
b/34716856
Test: todo
Change-Id: Ib23650ca1277f0dfb24e5c789d65a19dce8b1abc
We need google shims on the vendor partition because they are providing
an implementation of a vendor defined interface. They were written by
google just as a courtesy/to make the transition easier. They're
basically a set for vendors to assemble their hal implementations
from.
Bug: 34135607
Test: marlin persist.hal.binderization on/off
Change-Id: I2e2af5af39264cf290259755bb9b2eb9827a21f5
