The UI Thread is whatever we happen to be attached to, or the current thread if we are unattached.
Bug: 38180075
Test: Repro from bug. go/wm-smoke.
Change-Id: I3f75882aa13de2b781c71ebbc7b09888981521b3
in time taken to read cpu stats.
This should avoid super long durations > 10 secs that are occasionally
seen in system logs.
Bug: 62104524
Test: manual
Change-Id: I9ed985a61125a207b6d4f6297091bd11b4c78616
- Fixed WebView < > and API calls.
- Improved description of virtual views.
- Described how to set boundaries of virtual views.
- Improved AUTOFILL_FLAG_INCLUDE_NOT_IMPORTANT_VIEWS doc.
Bug: 37567048
Test: ran 'm -j doc-comment-check-docs' and checked resulting HTML
Change-Id: Ic0d1e9ff2703c87d4007f0092a2f8dfe0efca6db
Use the looper from the TextView's thread for the helper
Bug 62043115
Test: Manual, type on edit field and select text
Change-Id: I501430a500016a81963a9f9fa636474b708b9b36
- Happens at a channel level, and system services have to opt in
Bug: 38428796
Test: runtest systemui-notification, manual
Change-Id: I5592b4e32c177844583aee25379390199eef0a58
There were a couple of problems with the new paddingHorizontal
and layout_mareginHorizontal attributes. For one thing, the behavior
of layout_marginHorizontal needed to change with respect to marginStart/End.
Instead of the implemented behavior where Horizontal took precedence over
start/end, the behavior is being changed such that start/end can override
horizontal. This makes it consistent with the way that the attributes work
for padding.
Also, the attribute docs were not correct. For one thing, they needed to be
updated to match the new behavior for marginHorizontal. Also, the docs for
the padding attributes (including the docs for the existing "padding") were
not correct for the behavior as-implemented (specifically with respect to the
precedence of the attributes where paddingStart/End are concerned).
Bug: 37756178 double-check logic of horizontal/vertical attributes wrt start/end attributes
Test: Updated cts tests, submitting at the same time
Change-Id: I85a102549022cbec7d7b5c76f31ac985db103372
Bug: 38509779
Bug: 32295793
Setting to null causes problem seen in b/38509779
Using weakreference will cause problems if the application
declares AuthenticationClient as a weak reference as well and would
very likely break a lot of existing apps.
Test: manual
Change-Id: Iedbdd76f3a402b766ce8a799bf76324d1164a0ec
Per bug, updating code snippet to clarify that it's an example of an
implementation of the method. Oscar & Felipe, please check that I did
it right!
Also fixed a couple of HTML syntax errors while I had the file open
(badly formatted escape-characters that Chrome figured out, but other
browsers might choke on) and a spelling error.
Revised Javadoc is staged to:
http://go/dac-stage/reference/android/view/View.html#autofill(android.view.autofill.AutofillValue)
Test: make ds-docs
Bug: 38347106
Change-Id: I587a66c53fd5ebeeb6108529723d2d7a74c61cf7
An invalid, 'null' resource ID is defined as 0. Apps often use -1.
Add a helper method that makes checking valid IDs easy and more
centralized.
Eventually make it public API.
Bug: 38393777
Test: manual
Change-Id: I969ec4a45e86bdab3d7f57d357d475b77c7f8a78
Allows creating a HW Bitmap from the drawing
commands of a RenderNode.
Bug: 38507414
Bug: 37698012
Test: Sample in HwAccelerationTest
Change-Id: I57c60b2c8bf5194f4412ad4b7f1c1f35e2e4c757
Primarily used by tests to be more robust, since reading raw logcat
data recently became very flaky.
Bug: 37915178
Test: cts-tradefed run commandAndExit cts-dev -m CtsOsTestCases -t android.os.cts.StrictModeTest
Change-Id: I3f12508bb6206c53005356b5d8d9ba57aac2436e
Port the previous change that was merged DO NOT MERGE for an upstream
branch that filters non-exported activities from ResolverActivity.
Bug: 34147053
Test: by previous CL
Change-Id: I2d0a018afd5e9c876f82fc5f551f1edc4b2305c8
The restriction on system property key length has been lifted.
Update the invoke-with code to first check the full-length property.
Then fall back to the truncated version for backwards-compatibility.
Bug: 37276041
Test: m
Test: manual with long package name (Maps)
(cherry picked from commit 4c88a8c684)
Change-Id: Ic27a26a676f7ac5a998a9bea14259725e89b7e37
When using SYSTEM_UI_FLAG_LAYOUT_FULLSCREEN and
fitsSystemWindows="true", there are cases when the
inner insets will change but the content insets will
remain the same.
One case where this happens is when using
windowSoftInputMode="adjustResize|stateHidden",
bringing the keyboard up, hitting the Recents button,
then going back into the app.
By issuing a requestLayout when the base inner insets
change, we ensure that the inner layout is properly
updated, avoiding an empty space where they keyboard
used to be.
Test: Run app from b/37743589#comment6
BUG: 37743589
Change-Id: I7615c51b8f4eee3971b9b95e638011c1d885b3a6
Change-Id: I97f791b61f9b4f7ed33305345bf3d92394b40ae4
Fixes: 38202759
Test: cts-tradefed run commandAndExit cts-dev -m CtsAppSecurityHostTestCases -t android.appsecurity.cts.EphemeralTest
Test: Manual. Create sample app that replaces the launcher to test ability to see ephemeral apps.
See Ia9081d92ae9aea50d863455be770eecd0c73be1a for multi-selection
counts.
Test: Manually checked logging happens as per go/tron-howto and verified
nothing is broken in related classes by running:
bit FrameworksCoreTests:android.widget.TextViewActivityTest
bit CtsWidgetTestCases:android.widget.cts.TextViewTest
Bug: 32572232
Change-Id: I4ceab136ab73a20c6bc56732f7606ed103fe64d3
Fix a bug where the internal date/time picker views reused view IDs
causing state save/restore bugs when placed within AlertDialogs and
other places. Since the pickers already save/restore their state at a
higher level leaving this enabled was redundant.
Bug 32654446
Test: manual
Change-Id: I3df2fc932ac5296ab6eb0a5013dddef8d1117635
