API guidelines recommend adding ability to set the executor
Bug: 156046799
Test: atest HdmiControlServiceTest
Change-Id: I1f6dde6efc498a112ac7da2c61c36fcc2afa6638
* changes:
Implement a bare-bones UDFPS overlay
Add interfaces for controlling the UDFPS overlay
Add isUdfps to FingerprintService
Add onFingerDown/onFingerUp to FingerprintService
* changes:
Clear identity when checking BiometricService#isStrongBiometric()
17/n: Make generateChallenge/revokeChallenge ClientMonitors
16/n: Introduce ClientMonitor#FinishCallback
15/n: Combine EnumerateClient into InternalEnumerateClient
14/n: Remove the use of groupId
13/n: Remove the use of DaemonWrapper
As part of the move to make all HAL operations a ClientMonitor,
this change makes generateChallenge return the challenge via a
callback.
Added generateChallengeBlocking() so that existing synchronous clients
can continue to work with minimal changes (only renaming is necessary),
and so that it's explicit that the call may take some time (up to
hundreds of milliseconds on some devices).
1) LockSettingsService#resetLockout path
2) Fingerprint/Face enroll path
Bug: 157790417
Test: Enroll on fingerprint/face devices
Test: ResetLockout on fingerprint/face devices
Change-Id: I99efc1c72c8ac994c56d6d2deaf3e0131e5b2dac
Upstream AOSP change doesn't have a newline at the location
this change addresses, causing a merge conflict
Test: Build, presubmit
Bug: 159464174
Merged-In: I957dfe35aacc3e397611d8ab0b9ad19a1130a3dd
Change-Id: I7614d420ad55129b04e0dfa9310e39d715ed2da7
Update the GeomagneticField coefficients according to the World
Magnetic Model 2020-2025 values and specify the base time in UTC since
calls to System.currentTimeMillis() will return a UTC based time.
Signed-off-by: Cory Hockenbury <chockenbury@gmail.com>
Change-Id: I7614d420ad55129b04e0dfa9310e39d715ed2da7
GroupId was never used. FingerprintService#enroll sets groupId to
always be the same as userId. This change removes it from (almost)
everywhere.
We should update BiometricUtils and subclasses so that new HIDLs
do not take space for groupId. Since the framework persists the
info and we do not want to deal with database migration, let's
leave it there until the devices which shipped with groupId are
EOL.
Bug: 157790417
Test: Enroll fingerprints with a build before this change. Update
to a build with this change - still able to authenticate, rename,
remove previously enrolled fingerprints
Change-Id: I21391f5ffb71b27d7f9d4cfd4e6253b6001276dc
Update the GeomagneticField coefficients according to the World
Magnetic Model 2020-2025 values and specify the base time in UTC since
calls to System.currentTimeMillis() will return a UTC based time.
Test: CTS test passes
Signed-off-by: Cory Hockenbury <chockenbury@gmail.com>
Merged-In: I7614d420ad55129b04e0dfa9310e39d715ed2da7
Change-Id: Ica9182053359e5387da57bb0b3b93347040160ed
This is already semi-combined in the Fingerprint/FaceService layer,
for example authenticate() already updates the active group. This
change makes the remainder of user-specific operations (resetLockout,
enroll, etc) atomic from a caller's perspective.
Also fixed some fingerprint naming (passing userId through groupId
parameter since groupId is not used). GroupId is going to be removed
in a follow-up CL.
Removed unused rename() from IFaceService.aidl / FaceService.java
Bug: 157790417
Test: For fingerprint/face, do each of the following
Test: Enroll, rename, auth, remove each for multiple users
Test: 1) Set up work profile with shared challenge
2) Lock out owner and work profile
3) Unlock keyguard via password
4) BiometricPromptDemo for work profile and owner have
lockout reset
Change-Id: I71d33cccba40e0df09e0faa8a4d3973e77470b93
Do not disallow reprocessable session to use physical streams.
At the same time, we cannot claim reprocessing to/from physical streams
work because:
1. It's undefined behavior to reprocess an input buffer into a physical
stream of different physical camera, or an input physical stream buffer
into an output buffer of different physical camera.
2. If the reprocessing input buffer comes from different physical cameras
with different sizes, because the current HAL API doesn't support sending
different size buffers through input stream, there isn't a
non-vendor-specific solution available.
Bug: 157123506
Test: vendor testing
Change-Id: I1ccf2b3918c1cb475b1baec10d35c6785b25208d
Allow physical stream's field-of-view to be larger than that of the
logical stream, thus enabling more robust depth-from-stereo or motion
tracking.
Test: Build
Bug: 153583897
Bug: 157676445
Bug: 157138779
Bug: 155393103
Change-Id: Ibc059396e9d5e7db80b6c2632d26f48774aad4d4
* changes:
11/n: Remove unnecessary things from BiometricAuthenticator
10/n: Split ClientMonitor functionality into appropriate subclasses
9/n: Remove MetricsLogger and Constants from biometrics
8/n: Move acquire message ignoring to face subclasses
7/n: Move Fingerprint/Face AuthenticationClients to their own file
6/n: Decouple remainder of lockout stuff
5/n: Virtualize biometric performance stats
4/n: Start moving fingerprint lockout to its own class
3/n: Move internal cleanup to its own ClientMonitor subclass
2/n: Clean up ClientMonitor
1/n: Move modality-specific files into new subdirectory
This interface was created before the existence of BiometricService.
Since the sensor registration, arbitration, etc all exist in
BiometricService layer, most of the things here can be removed and
split into separate classes.
Bug: 157790417
Test: Builds
Change-Id: I91ebc13959943b86b93d4fff46ee99050f2cb7bc
