When there is no HAL connection or HAL malfunctioning, thermal status
will stay in an invalid value, which would be returned when client
register status callback.
This CL will opportunistically assume system is in "NONE" throttling
when in above states such as no HAL connection or HAL malfunctioning.
Bug: 155772239
Test: atest ThermalManagerServiceTest
Test: atest PowerManagerTest
Test: atest thermalmanager-test
Change-Id: I9fceb5ef6f8b9a23983b7b93200f381eaf23a77b
This reverts commit d64e9452e4.
Reason for revert:
Those two metrics are added for Openroaming service, but they can not be used finally, so remove them.
Bug: 146669261
Test: No, just remove proto fields.
Change-Id: I601927c443b13a5d2d52e60ad4c910204156a102
- Added onFixedRotationStarted/onFixedRotationFinished callback in
DisplayWindowListener
- onFixedRotationStarted shall be called before onTaskAppeared for PiP
- When onTaskAppeared is received in PipTaskOrganizer, we defer the
entering PiP transition if fixed rotation is ongoing
- When onFixedRotationFinished is received in PipTaskOrganizer and the
entering PiP transition is deferred, schedule an immediate
transition to PiP to make sure all the expected callbacks from
PipTaskOrganizer are still being sent
Video: http://go/recall/-/aaaaaabFQoRHlzixHdtY/bb8HjJvMZKtpN8YTPKZXmj
Bug: 153861223
Test: manually enter PiP from Play Movies / YT fullscreen
Test: atest ActivityRecordTests \
RecentsAnimationControllerTest \
ActivityTaskManagerServiceTests \
PinnedStackTests
Change-Id: I0dea905d610e2387af56b611be5f93518cc9a153
This reverts commit d638cd6164.
Bug: 157106728
Bug: 115619667
Reason for revert:
Looked into this some with the Camera test case in b/157106728. The wait()/notify strategy isn't going to work here because we have a 3-lock system vs. 2. We have the WM, UGM, and AM locks involved. Both the WM and UGM lock are held in the call to ActivityManagerInternal#checkContentProviderUriPermission and in the subsequence call that ActivityManagerService#checkContentProviderUriPermission AMS calls back into WMS and UGMS for various reasons which requires their locks...wait()/notify strategy allows us to break out of 1 lock, but things get really messed up if we try to use that strategy to break out of 2 locks.
Change-Id: Ic0f30f6c1b24334105a00790488d2758390e85a1
This will prevent the stack from becoming separated and confused in some situations. The root cause was that the BubbleStackView returned false by default in its touch handler, which allowed the bubble views to grab touch events they shouldn't have and start animations. However, those bubbles trying to start the animations should have been prevented by the active controller logic, but it wasn't due to a missing check. Might as well fix both causes!
Also, cancel the path animator before starting a new one just in case.
Fixes: 155499044
Test: manual
Change-Id: Idbdb68dfe3cccc69ae3709faad8d3b7078b354a9
- specifying the order of camera open and session configuration.
- explaining the use of CONTROL_ZOOM_RATIO with concurrent operation of cameras.
Bug: 151891611
Test: make doc-comment-check-docs
Change-Id: I3ee3f114f7295570aa6af5dbe35bb32db555811b
Signed-off-by: Jayant Chowdhary <jchowdhary@google.com>
This fixes the issue where the last bubble doesn't get an exit animation, saves memory, and is less unnecessarily risky when you have no bubbles.
Test: add and remove a lot of bubbles
Bug: 134514227
Change-Id: If3e3164e8a76a0155496b05e5ba8994d8da61c1a
This ignores failures in the other parsing comparison tests
in favor of a specific test that runs through the system APKS on
the device and verifies that no exceptions are thrown.
Bug: 155935153
Test: manual add a broken APK and verify failure in
SystemPartitionParseTest
Change-Id: I4a02b3dcd8ab993f27c6fc52bb4048d8f2eab273
We were resetting the expiration clock every time a finger went down, but were not resetting it when fingers went up. As a result the user had a super narrow window to get all the fingers up before the gesture canceled.
Bug: 156880174
Test: atest GestureManifoldTest
Change-Id: I81dd9cd67729cfae38952d44295c3340c5426028
Change-Id: I1b6323ac413bcbd46fd9292fbb4cf7ebdf2b6246
This reverts commit 5de8369e20.
Reason for revert: Found source of the bad uids. Adding wtf logs back.
Change-Id: I45434eed78d0b2d5c8df9ff1557ee4bd1e0dd4d0
Test: atest AppEnumerationTests
Bug: 156863453
Bug: 157066842
