As of today we register Ime consumer in IMM onWindowGainedFocus().
However, it does nothing functionally when ViewRootImpl#sNewInsetsMode is disabled.
This CL adds a check to not register consumer at the first place.
This may also be causing a leak. Added a TODO to revisit that before we
enable the flag.
Bug: 131181940
Test: atest ViewRootImpl
Test: atest InsetsController
Change-Id: I6b2b62d10670d63edafeb1b24c25b26c7721f3ad
QuotaController was inadvertently updating all Timers for a particular
user whenever any process state crossed the FOREGROUND_SERVICE
threshold, instead of only updating the Timer for the specific UID.
Also adding more data to QuotaController's dump to make future debugging
easier.
Bug: 129117282
Test: atest com.android.server.job.controllers.QuotaControllerTest
Test: atest CtsJobSchedulerTestCases
Change-Id: Ic8d9e6478e61cc62318ae5651f0526e41a71de8d
smallestScreenWidthDp configuration should be set to smallest possible size.
This could cause unexpected resources to be loaded based on the incorrect configuration. So it
should be set as minimum size of screenWidthDp and screenHeightDp to account for rotation.
Test: reproduce test steps in bug
Test: atest CtsWindowManagerDeviceTestCases:DisplayTests
Bug: 129521230
Change-Id: I07403294e8fe6e6ad7aa718c8a5948aff4fafae1
(cherry picked from commit 7671d7042d60944813f36a03b74199c5a3d737ad)
Many processes are incorrectly listed under UID 0 (e.g. surfaceflinger)
because some "/proc/$PID/stat" files are with user "root".
The user of the folder "/proc/$PID/" is the correct one to use.
Bug: 119218763
Test: adb shell dumpsys batterystats -c | grep surfaceflinger
Change-Id: Ibfbe9b04c884bc9bde383629bcbba73f48ed34ee
(cherry picked from commit 7ead3ee46f)
This change is an attempt at delaying the install complete callback
to the installer on a DONT_KILL install until the package indicates
that it as completed updating its classpath with the new APKs.
Bug: 80269951
Bug: 109751013
Test: manual test of dynamic delivery/instant apps
Merged-In: I689ec523522da37987cff9b1a67eaae9e5633ffb
Change-Id: I689ec523522da37987cff9b1a67eaae9e5633ffb
(cherry picked from commit abfc054c90)
Use notification colors even on night mode (enforcing contrast of
course)
Test: pull down shade on LS and launcher
Fixes: 130756560
Change-Id: I3f3f6e426163e316906692ef8b9c416ca94242ee
It's possible the Bundle has not been initialized when determining if
the ANGLE dialog box should be shown, due to how early that check is
performed. This change will verify the Bundle is not null before using
it and fall back to using the Context if it is.
Bug: 130185493
Test: atest CtsAngleIntegrationHostTestCases
Test: Enable Toast Message and then atest CtsAngleIntegrationHostTestCases
Change-Id: I39f48bdf20616298c66b2bb36082149cb24e908c
(cherry picked from commit 5da6a95eab)
The signature field is deprecated, but must still be populated. The
thought was that for APEX files [which is a new concept and thus
could use the new way to handle signatures] we could remove this
capability, but, it turns out Play relies on it.
Change-Id: I7096567082d4f798be1d3e6007122750453665ae
Fixes: 130827182
Test: Manual.
Test: Create sample app that dumped the signatures of an APEX
Test: Note that the signatures before this change were 'null' and set to something after this change
When apps are sending EXTRA_CHOOSER_TARGETS, the wrong value is
getting recorded for metrics tracking. Use the correct offset.
Bug: 130613626
Test: Retest share through Imgur app
Change-Id: I0bccbfb064796f547b2e7b474f5aa06993ac3920
This moves everything over to using a shared data class
for Estimates in SettingsLib which will facilitate a cl
that will help ensure greater consistency across
surfaces where battery estimates are shown.
Test: Tests pass
Bug: 124030091
Change-Id: I0b7f1f3a806255ff4804a00e6d90a7846c484484
- Added a new value for config_longPressOnPowerBehavior:
"5" means "launch assist intent"
- Added global settings for power key longpress and "very
long press" (3.5 sec), allowing these functions to be
tweaked at runtime by privileged code
Bug: 130415373
Test: adb shell settings put global power_long_press 5
# ^ set longpress to assist
adb shell settings put global power_very_long_press 1
# ^ set vlongpress to globalactions (so you can still
# power off the device)
Test: atest frameworks/base/core/tests/coretests
Change-Id: Ic611b2850dc98fddd7662d32b0a60ca1e8231efb
Move the profile switching logic into the scrollable list area. Fix
visual artifact due to reusing drawable.
Bug: 130353935
Test: atest ChooserActivityTest + visual inspection
Change-Id: I0666b0e61ea696017c2da4854e407d502b7b556c
To help evaluate the new ranking algo we want to know if the package for
a direct share was also offered as one of the ranked apps.
Logging the position if it is part of the top apps, otherwise logging
-1.
Also added a test case I missed for HashedStringCache and removing the
test for hashing resulting in the same hashed string when called twice
from the ChooserActivityTest since we now cover that in the
HashedStringCacheTest.
Bug: 130658734
Test: Added more test cases in ChooserActivityTest and did manual
testing
Change-Id: I0e34a6bf64114d94197f62b8219652c33c03a410
CL [1] introduced SystemUI to support updating IME window status per
display. (i.e. IStatusBarService#setImeWindowStatus left TODO item to
add displayId parameter to select the target display.)
For IMMS, we applied this API base on current token's displayId, and refined
the code flow in Sysui to reset IME window state for non-target display's
NavBar when single IME switch to another display case.
And make sure some cases in IMMS should reset IME window status:
- When current method unbind.
- When current top window focus display is not same as current token's display.
(That means the input session may connected but IME window not
yet attached. Note that it doesn't include external display without
system decoration and show IME window on default display case since it
is intentional behavior and we still need to update status for this
case.)
Also added testSetImeWindowStatusWhenImeSwitchOnDisplay in
NavigationBarFragmentTest to enhance the ability of verifying external
navigation bar.
[1]: 24e7a9fdb5
Bug: 127309955
Bug: 117478341
Test: atest NavigationBarFragmentTest
Test: manual as below steps:
- Pre-condition:
1. Enable desktop mode.
2. Create simulated display.
- Use case 1):
1. Launch an activity with input field on simulated display.
2. Bring up the IME there, expect back key icon on external
NavBar will changed when IME bring up.
3. Tap on primary display or launch an activity there.
(or pressing home key)
4. Expect IME will hide & back key icon on external NavBar will
set back.
- Use case 2):
1. Launch activity with input field on default display.
2. Tapping EditText in activity on default display.
3. Expect IME will bring up here and back key icon on default
display will changed when IME bring up.
4. Launch activity without input field in external display.
(i.e. clock app)
5. Expect Both Default & external display's nav bar back key icon
will set back.
Change-Id: Ia414b8aea631e295cccd6f6da44d04bad16545c7
When running cancel animation for drag shadow image, the system cannot
start another drag operation, which is problematic when the system would
start the other drag operation quickly.
Bug: 130313958
Test: Cancel drag operation by calling cancelDragAndDrop
Change-Id: I5d6650a0ce9a4cd80bbdb1beabc9e514349ccadc
This will sort the share activities based on the APS sorting.
We add a constructor for ResolverListController which takes an
AbstractResolverComparator, so that ChooserActivity may pass in
the APS comparator if it is enabled and available.
Test: Manually tested on APS sorter that did no sorting.
Test: atest frameworks/base/core/tests/coretests/src/com/android/internal/app
Bug: 129014961
Change-Id: I542254ffb0debad45bcd8d5073cc3f3e1bafc616
Signed-off-by: George Hodulik <georgehodulik@google.com>
