The goal of this CL is to remove RELAYOUT_RES_DRAG_RESIZING_DOCKED and
RELAYOUT_RES_DRAG_RESIZING_FREEFORM, which is a step to make relayout an
oneway binder call.
When the resize mode is changed at the server side, the server will send
the new mode to the client via IWindow#resized, so the client doesn't
need to obtain the resize mode from the flags returned from relayout.
Bug: 161810301
Test: Perform drag-resize and drag-move on a freeform task and see if
there is any unexpected behavior.
Change-Id: I450abc113932b6c1dc0ce0e2b76bebdf85a60777
This will be used by clients such as Gboard to ensure they are
responding to translation events for the correct activity. Ideally we
would include the activity token, in case there are multiple activities
being translated in an app, but that isn't supported right now and it
requires changes to EditorInfo (in Input Method Framework) to get the
activity token.
Also fixed an issue where the callback methods are invoked twice when
the IME itself is the translated app is registering a callback for
itself. Note: This happens in UiTranslationManagerTest where the test
activity and IME have the same uid, but shouldn't happen outside of
tests.
Bug: 223979041
Test: atest CtsTranslationTestCases
Change-Id: Id99f501b065b92fe3358c2ae525175ad3ea4b119
If the fill dialog is available for the field, the FillRequest flags
will contain FLAG_ACIVITY_START. The field is hidden originally but
it needs to be get from AutofillService to avoid creating unnessary
fill dialog presentation. Rename the flag because the it's not clear
to combine the relationship with fill dialog.
Bug: 223472039
Test: manual. Build and boot pass.
Change-Id: I52c367e75100f6af2ab782f58755bc5727b1dd72
Move the callback as the last paramter to registerOnBackInvokedCallback
Test: Existing
Bug: 224562951
Change-Id: I8608a03b53baa04bedee81d2e86b2339abd6c594
We introduce a seqId and use it to replace the
RELAYOUT_RES_BLAST_SYNC flag. The server code block that
produces RELAYOUT_RES_BLAST_SYNC is guarded by
mNextRelayoutUseSync, and sets it to false when executed.
This block is the only place that sets it to false. We replace
every place where it is set to true, with incrementing a seqId
stored on the integer. We change this block to not check
mNextRelayoutUseSync but instead check if the seqId has changed
since we reached the block. This is evidentally equivalent conditions
and in these cases we send the client the seqId. In the else block
where we wouldn't have sent RELAYOUT_RES_BLAST_SYNC we send -1.
On the client side, we replace the USE_BLAST_SYNC conditional
with a check if the SeqId we received has increased. It's clear
that this will never execute in cases where the seqId is -1. Likewise
its clear that if the seqId has increased, it will execute. Since the
seqId increases in all cases where the server would have returned
RELAYOUT_RES_BLAST_SYNC this change is a noop by itself.
Bug: 161810301
Bug: 175861051
Bug: 175861127
Bug: 200285149
Change-Id: Ib2e45cae7709519251e4bfcc2b8adc689c590a00
Instead of using the count to keep track of sync information in VRI, use
the SurfaceSyncer to handle this. This also means changing SV since SV
would also increment the VRI counter. This cleans up the logic since
code that's used for the SurfaceSyncer can be re-used in VRI.
Test: seamless rotation, launching apps
Bug: 200284684
Change-Id: Ia3c1a13b42ceba29a75425295e41e543a00436a3
This adds the committed state to the display.
The `state` is used to set the correct power state (e.g. turning display on/off), however there was no way to know when this power state change finished. This new "committedState" aims to fulfill this gap.
DisplayListeners receives a `onDisplayChanged` callback whenever the power state change is fully committed, and can get this information from DisplayInfo.
This will be used in a follow up cl to improve unfold latency performances.
Bug: 223808403
Bug: 220690092
Bug: 197515205
Test: atest LocalDisplayAdapterTest && manual
Change-Id: Ia2fbab768b696c59578d974c88157ee3c8f3ddde
We've received several feedbacks that the current tapTimeout is too short,
making it very hard to initiate handwriting. To unblock us from bug bash,
this CL temporarily use long lick timeout for handwriting timeout.
Once UX decided the final value for handwriting timeout, we will create
another CL.
Bug: 217603538
Test: atest FrameworksCoreTests:android.view.stylus.HandwritingInitiatorTest
Change-Id: Icb9f49b8b2a49c6c2a13fa377c20a3d1b3bd782c
The custom close action will replace the system close action and
the button in the menu will be colored red to be able to tell it
apart.
If one of the custom actions matches the close actions it will be
removed to avoid duplicates resulting from the custom close action.
If the app doesn't close the PiP after five seconds the system will
continue to close the PiP as normal.
Bug: 218487423
Test: manual
Change-Id: Ic3c520c50ac0e3d2f0e73cf2f71d47a4f548f861
Introduces layer mirorring of a single task, given the task is both
identified (future work - dependency on setup flow in SysUI) and the
flag is enabled (currently disabled).
Error handling (tearing down the VirtualDisplay if task recording setup failed)
will come in a future change
Bug: 216625226
Test: atest WmTests:DisplayContentTests
Test: atest WmTests:ContentRecorderTests
Change-Id: I16abd55ce75537ff8fbbb379e4970668ea02a230
When WindowManagerService.mInTouchMode is changed, we will set the new
mode to InputManagerService, and clients will be notified via
ViewRootImpl.WindowInputEventReceiver#onTouchModeChanged. We don't need
the flag to bring the information to the client.
This is also a step to make relayout an oneway binder call.
Bug: 161810301
Test: Invoke View#requestFocusFromTouch on one window and see if another
window can receive OnTouchModeChangeListener#onTouchModeChanged.
Change-Id: Ic149c95bee89b2be83a85eb4858131224b63a7c8
Handwriting initiator accidentally changes the View's handwriting area.
As a result, it stops working after startHandwriting is called for the
View once. This CL fixed the issue by not exposing the internal
handwriting area Rect object in the View.
Bug: 222117485
Test: manually tested
Change-Id: I5cb6231361fbd7566edddb1aebf7b99330c3a8d8
(cherry picked from commit b1ba2dae29)
