Re landing with the following changes:
We initially cloned the SurfaceControl handle to avoid locking when
accessing the SurfaceControl from the position listener callbacks
running from RT workers. But this meant the last reference to the
layer handle would only be released by GC. If SurfaceViews are
created and destroyed rapidly, we would be at the mercy of GC to
release buffers.
Original change:
There are three transaction queues that can submit SurfaceView
changes.
1. Buffer updates via BBQ apply token
2. SCC apply token
3. ViewRootImpl BBQ apply token
It makes sense for most SurfaceView changes to be synchronized with
ViewRootImpl draws since the caller can optionally synchronize the
change with main window content.
This change eliminates the tmp transaction that is applied directly
via the SCC apply token and instead applies them with the ViewRootImpl
draw transaction.
Also take the opportunity to scope down mSurfaceControlLock usage.
Test: atest SurfaceViewSyncTest
Test: go/wm-smoke
Test: run mem tests via forrest
Bug: b/217973491, b/221631942
Change-Id: Idba712d146e62d7346920dc4f060cba92d47fada
Bug: 204794759
Rotary scrolling should trigger the stretch edge effect. Mouse
scrolling should NOT trigger the stretch edge effect. This
adds support for rotary scrolling while keeping mouse scrolling
from triggering the edge effect.
Test: new CTS tests. manual testing on emulator
Change-Id: Ibad25bcbc0740f2af2042baf67f7606f15b8c57e
While we investigate the root cause of this bug, turn the crash into a
warning.
Bug: 218409359
Test: none; we don't have repro steps for the bug yet
Change-Id: I6d699ea839726adc634a200a0057e316d59ecaa9
I discovered this error in documentation while I needed to capture "PasteAsPlainText" in an android app.
I suspect it might be missing in other folder too, but I’m not familiar enough with Android to be sure
Change-Id: I77af9b06eb602c4ba60b0744cc02f7449bc4f500
There are three transaction queues that can submit SurfaceView
changes.
1. Buffer updates via BBQ apply token
2. SCC apply token
3. ViewRootImpl BBQ apply token
It makes sense for most SurfaceView changes to be synchronized with
ViewRootImpl draws since the caller can optionally synchronize the
change with main window content.
This change eliminates the tmp transaction that is applied directly
via the SCC apply token and instead applies them with the ViewRootImpl
draw transaction.
Also take the opportunity to scope down mSurfaceControlLock usage.
Test: atest SurfaceViewSyncTest
Test: go/wm-smoke
Bug: 217973491
Change-Id: I0b5741b376e042537c37ec34644e0a048eb4bfe0
1) Update the LineBreakConfig class to be immutable.
2) Do not return null in the PrecomputedText.Params#getLineBreaiConfig API
Bug: 216638444
Test: atest TextViewTest; atest MeasuredTextTest; atest PrecomputedTextTest; atest TextViewPrecomputedTextTest; atest StaticLayoutLineBreakingVariantsTest
Change-Id: I07766137ff6639c7d4acaad07dbcf11a2841cdb0
One of the optimizations ag/12911059 did was calling
SelectionModifierCursorController.show() before
startActionModeInternal(). The rationale was that if we start the action
mode first, SelectionModifierCursorController.show() would end up
invalidating the action mode twice unnecessarily, once for each handle.
However, with this optimization, we are calling
SelectionModifierCursorController.show() even when onCreateActionMode
returns false.
Reverted this particular optimization to fix the issue.
Added a test which was failing without this fix but passing with it.
Fixes: 199380016
Fixes: 214341747
Test: atest TextViewActivityTest
Merged-In: I793f76a23978cbbbbde2d16e8a522615174bcdd5
Change-Id: I793f76a23978cbbbbde2d16e8a522615174bcdd5
(cherry picked from commit 11bd644822)
Does not remove Support Library artifacts from docs classpath (ApiDocs.bp)
because they are still used in development/samples, which is not as easy
to migrate as javadoc.
Bug: 158779503
Test: make docs
Exempt-From-Owner-Approval: Mass find/replace for androidx migration
Change-Id: Icf7f53ec36a0e970413352e2ebf40ce9d60ed17e
One of the optimizations ag/12911059 did was calling
SelectionModifierCursorController.show() before
startActionModeInternal(). The rationale was that if we start the action
mode first, SelectionModifierCursorController.show() would end up
invalidating the action mode twice unnecessarily, once for each handle.
However, with this optimization, we are calling
SelectionModifierCursorController.show() even when onCreateActionMode
returns false.
Reverted this particular optimization to fix the issue.
Added a test which was failing without this fix but passing with it.
Fixes: 199380016
Fixes: 214341747
Test: atest TextViewActivityTest
Change-Id: I793f76a23978cbbbbde2d16e8a522615174bcdd5
Replaced "than" (comparative) with "then" (temporal) for the setTextSize JavaDoc. No functional change.
Change-Id: I84b5d1b250cb7a5d925cfbd0a86a2bfef55ade24
ACTION_SET_SELECTION is needed on all TextView nodes to reset the
cursor. But some non-edit texts are selectable and need to be
identified by a11y services. ACTION_SET_SELECTION has been on these
nodes for years so to avoid the complexity of breaking this, add a
node property.
Services should use ACTION_SET_SELECTION for nodes where
property=true
Bug: 62058901
Test: atest AccessibilityTextTraversalTest, TextViewTest cts and
core (TextViewTest#testCutShouldNotThrowException keeps failing but
it's unrelated to this change)
Change-Id: Ia825dd806824e663e7fa7d274f51c6bf44f0c2c4
This change propagates FLAG_WIDGET_IS_COLLECTION_CHILD and
FLAG_USE_LIGHT_BACKGROUND_LAYOUT to nested RemoteViews (either added by
addView, or set as portrait/landscape/sized RemoteViews) if that flag is
also present on the parent RemoteView.
For FLAG_WIDGET_IS_COLLECTION_CHILD, this prevents a PendingIntent from
being set on a RemoteView that is a child of a collection item RemoteView.
For FLAG_USE_LIGHT_BACKGROUND_LAYOUT, this ensures that nested
RemoteViews use their light background layout if this is set on the
parent.
Bug: 214288099
Test: RemoteViewsTest
Change-Id: I8da92d0d338631a99ee718a136bf9674827437bb
