The recently added TextUtils.formatSimple() can efficiently format
simple strings, roughly 6.5x faster than using String.format().
This change starts using this new method in hot codepaths identified
from pprof traces, and this CL alone should reduce system_process CPU
usage by 0.24%. Linked bug has detailed pprof screenshots.
Bug: 170978902
Test: atest FrameworksCoreTests:android.text.TextUtilsTest
Exempt-From-Owner-Approval: trivial refactoring
Change-Id: Ie5a21a24be6dfcdb29768fe235e20f050fc111c7
* changes:
Apply fixes for EfficientStrings.
Apply fixes for EfficientStrings.
Apply fixes for EfficientStringsChecker.
Apply fixes for EfficientCollections.
Trivial refactor for consistent naming.
Expand formatSimple() to support widths.
Refinement of EfficientStringsChecker.
The server won't dispatch the legacy content insets, stable insets, and
visible insets to the client. The insets would be computed from the
insets state by the client.
This CL also fixes the insets scaling issue in compatible mode.
Bug: 149813814
Fix: 169940916
Fix: 165412978
Test: atest WindowAddRemovePerfTest InsetsAnimationControlImplTest
ActivityRecordTests DisplayPolicyLayoutTests
InsetsPolicyTest InsetsStateControllerTest
Change-Id: I83570973f587a6abf887752494f750b46e3e484d
The recently-built Error Prone checker has found many instances where
we're always paying the cost of StringBuilder concatenation, even in
the typical cases where preconditions are successfully met.
Benchmarks have shown that even when replacing these with varargs
formatter strings, the default case is 20x faster.
Bug: 170978902
Test: none
Exempt-From-Owner-Approval: trivial refactoring
Change-Id: If8c00bc73467bfb91ec16c162969c9d26ca53646
The new logic sends the requested visibility to server even when the
client doesn't have a control. This makes the requested visibility
reliable at the server side.
This CL also changes getRequestedState to getRequestedVisibility in
WindowState, because the frames in getRequestedState are not reliable.
Fix: 161247175
Fix: 166819574
Test: atest DisplayPolicyLayoutTests DisplayPolicyTests InsetsPolicyTest
InsetsSourceProviderTest InsetsStateControllerTest
WindowStateTests InsetsAnimationControlImplTest
InsetsControllerTest WindowAddRemovePerfTest
ActivityRecordTests
Change-Id: I7d32dacaa1302bab61386f5bd69aaa7dde953ebe
Update AlwaysOnHotwordDetector.Callback#onAvailabilityChanged API
documentation to reflect its current behavior.
Test: N/A
Bug: 170681090
Change-Id: I6b16b527a521af66367caaeeb28cc73ae22b0272
Those annotations could be inferred by some tools (like Kotlin), but the
https://checkerframework.org/ doesn't check inherited annotations
complaining about all equals() invocations that get nullable argument.
The change was generated by running
find . -name \*.java | xargs sed -i 's/public boolean equals(Object /public boolean equals(@Nullable Object /'
in the frameworks/base directory and by automatically adding and
formatting required imports if needed. No manual edits.
Bug: 170883422
Test: Annotation change only. Should have not impact.
Exempt-From-Owner-Approval: Mechanical change not specific to any component.
Change-Id: I5eedb571c9d78862115dfdc5dae1cf2a35343580
If Blast was enabled, WindowStateAnimator would create a container layer
and then a blast layer as a child. This caused some issues with
reparentChildren and relative Z, so instead of trying to continously fix
those issues, just use one layer. If blast is enabled, it will create a
single SurfaceControl with type blast. Otherwise, it will create a
BufferQueueLayer.
This means that we have to be careful with setting any properties on
the layer since the blast adapter could also be modifying the layer.
This is in the right direction since we will move all transactions for
the layer to the client, eventually.
Test: YT and Chrome with and without Blast
Change-Id: I43efc37663696ba8e54d341a03531f0048ee6a14
In order to provide a better user experience, it would better know the
reason why a save dialog was not shown and improve the autofill experience.
So it is necessary to add the reason why a save dialog was not shown to
FillEventHistory.
Bug: 158328375
Test: atest FillEventHistoryTest
Test: atest InlineFillEventHistoryTest
Change-Id: I50c1e40334066d3f4844426c4b03294079a967b2
This CL removes
- mSeq
- System UI flags used to communicate between WMS and System UI
- redundant AIDL methods
- redundant fields and methods
- redundant tests
- PolicyControl
This CL also
- refines the format in DisplayPolicy#dump
- sends a boolean to InputManager to indicate if System UI is in a low
profile mode instead of sending the legacy system UI visibility
Bug: 149813814
Fix: 169105126
Test: presubmit
Test: dumpsys window displays
Test: See if the layout of ImmersiveModeConfirmation is as expected
Change-Id: I8c8df509355bebc9b560af57d5458614557bcd2f
* changes:
Demote AlwaysOnHotwordDetector to SystemApi
Sessionize VoiceInteractionManagerService
Sessionize the SoundTriggerService layer.
Use standard API in SoundTriggerTestApp
Fix bug in dumpsys
Require identity information in SoundTrigger.java
Associate an originator identity to sessions
Extract permission checking as a separate aspect
Correctly handle HAL death
In certain languages, "Until Tue, 9AM" and "Unil 3pm" are formatted
differently, so separate them into two unique strings for better translation.
Test: manual
Fixes: 165796742
Change-Id: Ib7c44714191acca427bcee0a4d88e1cb50a90afc
AlwaysOnHotwordDetector gets demoted from public to SystemApi and
new permissions are enforced on its methods that imply usage of the
microphone via sound trigger.
Bug: 163865561
Change-Id: I113d69e569962b20d9d7dd1c6815daa6fb0650c5
This change associates an identity with every
VoiceInteractionManagerService session.
Change-Id: I6c87c2976a4409b8dda4480234d794db91cee374
Bug: 163865561
The frame won't be changed if there is no IWindow#resized or
IWindowSession#relayout. So it can be retrieved from these methods
directly instead of another binder transaction.
And because some parameters are usually used together for layout,
the parameters are consolidated into a new ClientWindowFrames.
That reduces changing the interface in the future if the frame
related information needs to be changed.
Also refine the resize handling in ViewRootImpl to make it easier
to read.
There should be no behavior change by this modification.
Bug: 161781274
Test: WmTests, DialogFrameTests
Change-Id: I9f711ad2023442046fa8582944320b98e7c4ecfa
Now both notification pipelines require DismissedByUserStats when
removing a notification.
Also, refactored the DismissRunnable that gets called whenever a
notification is manually swiped away and dismissed from the
NotificationShade. Now, ExpandableNotificationRowController injects a
OnDismissCallback with a #onDismiss method that will get called whenever
a notification is manually swiped by the user OR is clicked with the
AUTO_CANCEL flag. Since it's injected it's easier to switch between the
new and old pipeline's OnDismissCallback (one which interacts with the
old pipeline's NotificationEntryManager, and another with the new
NotifCollection). Now this dismiss runnable doesn't need to be passed
around starting from inflation.
Test: atest SystemUITests
Change-Id: Iec985ce2c502462ee35cf86c1e0332168e578823
The Ranking object of a notification update has lastAudiblylertedMs of
-1 if the notification did not buzz because a buzz happened within the
last 5s. This causes the bell icon to disappear on a conversation
where a person posted twice within 5s. This is resolved by copying the
lastAudiblyAlertedMs from the NotificationEntry's existing Ranking when
setting a new one, if the new Ranking's value is -1.
Bug: 156887617
Test: Update an alerting notification within 5s of update which caused an alert; bell should stay even though buzz is skipped.
Change-Id: Ibb72ab27c88cfa57add781e2746fff0eee5de6be
* For the case where the pinned inline suggestion triggers an
pending intent through the auth flow, and it returns a dataset
to be autofilled, previously we would replace the existing
dataset with the returned dataset. However, this is causing
several potential issues:
1. if the returned dataset doesn't contain inline presentation
the the pinned icon will not show again
2. if the returned dataset contains inline presentation but not
the pending intent, then the pinned icon will show up again
but tapping on it will not launch the pending intent
3. if the returned dataset contains the inline presentaion and
the pending intent, then when we "autofill" it, it'll fire
the pending intent directly as opposed to filling in the
value
* We fix the issue by not replacing the old dataset if the dataset
is a pinned inline suggestion.
* One caveat of the approach is that: a dataset can potentially
have multiple fields, and it's possible that some of the fields'
has inline presentation and some don't. It's also possible that
some of the fields' inline presentation is pinned and some isn't.
So the concept of whether a Dataset is pinned or not is
ill-defined. Here we say a dataset is pinned if any of the field
has a pinned inline presentation in the dataset. It's not ideal
but hopefully it is sufficient for most of the cases.
* An alternative approach is to have the autofill provider telling
whether they want to replace the old dataset or not, through
a new field in the returned Bundle. But that requres an API change
so is infeasible at this time.
Test: atest android.autofillservice.cts.inline
Test: atest android.autofillservice.cts.augmented
Test: atest android.autofillservice.cts.LoginActivityTest
Test: atest android.autofillservice.cts.AuthenticationTest
Bug: 159367101
Change-Id: I6d162aeb88a4655989c1aa315df8304c0980ac60
