(Previously called "full" idle is now referred to as "deep" idle.)
We now keep separate enabled states for light and full idle
modes, and shell commands dealing with them now include an
argument to specify which they should impact.
Change-Id: Idf6d3d4bd85f800ebc034d97f57863351138fb4a
The menu key was not unlocking the keyguard properly
and didn't hide the notifications.
Bug: 26539875
Change-Id: I9dd89767f00d29c15d1d4e840a1ab22d59e58c8d
If the general setting is not to hide sensitive information
we allow the user to dismiss it.
Bug: 26328503
Change-Id: Ibea5d94e544291f00d8a2923b8cbf14170510bd8
Public notification titles were incorrectly modified when
the notification wasn't autoredacted (i.e screenshot).
This also lead to the doze mode not working correctly anymore
as notifications were colored.
Bug: 24866646
Change-Id: I3d4cdb0e13cfbf3fb8082b930936260d7c4749af
Ignored! Removed! A non-broadcast!
Any attempt to send these will be silently (except a log)
ignored.
Change-Id: Ia19357d4c90fdcd18e45f2b41c57e654ef94edb1
When delete is called, now the STH unloads the corresponding models.
Fix bug with keyphrase where a delete call didn't unload (or stop)
the keyphrase model.
Bug: 27279380
Change-Id: Ia34f713d2aecef4102c0f0ccc57b8d2e5febe4bb
Moved user sensitive recents data to the credential encrypted directory
/data/system_ce. Since this directory is not available before the user
is unlocked, and unavailable task ids are needed to be known before
assigning an id to a new task, this change adds a file in /data/system_de
for each user, to store the ids of the tasks present in /data/system_ce.
Bug: b/24569398
Change-Id: I340d1abe23e2594e2cee4bff89a697bd8dc61fc7
This follows up to my previous CL [1], which forgot to update
BridgeIInputMethodManager.java (again!).
[1] I56934f18e30d90fcdf77bcbb0c35a92a5feb1b82
05c25f8a3a033816ac25aa5cd7db5b1ab495bc3f
Bug: 25373872
Change-Id: I0643069fd0b677115c89c340b7a23f1d12bd7988
Previously accounts were copied in onUserStarted. At that stage only
encryption-aware authenticators could run.
Bug: 27274434
Change-Id: I4f2dbdc09e81087d960067e7930ce3b9cff829ab
All depth formats are considered as non-high-res. Also compare depth
minimum frame durations instead of regular minimum frame durations
for depth formats.
Bug: 26687093
Change-Id: I65b691fe450077b0e3ace15f6ac0289b41caf42f
Alternate lock screen shows UNLOCK button in the middle of an
empty screen before any user has been created. See the screenshot
in the bug.
This fixes it. The approach is similar to the check in
KeyguardViewMediator.handleHide().
Bug: 27109342
Change-Id: I25536364b2c456d6fd55c79de7f25d2b0667e26f
Activity manager normally singals window manager that it is okay to
start the app transition process for an activity we are launching
during the resume process. However, the activity will not be resumed
if the activity is been launched in a non-focusable stack
(e.g. pinned stack). So, the app transtion will not start and it will
not be visible until the app transition time's out in window manager.
Activity manager now tells window manager to execute app transitions
if we are launching an activity in non-focusable stack.
Bug: 27224884
Change-Id: I35245fceb2637b927e7885b2c02daa1d55deaabb
Don't leave safe more if persist.sys.audit_safemode is set, unless the
current build date is newer than the specified value.
This allows us to keep the device in safe mode across reboots until an
OTA has been applied or user data is wiped.
Bug: 26902605
Change-Id: I781c3059ea8d4fb2f0c923e4488b1932d69678d3
(cherry picked from commit ee2b49283261f7b4518ec7c98f568f883f817c9d)
Clear animation when destroying window surface.
Because we are not stepping in animations for windows if they don't
have a surface, we can't use it as a detached wallpaper target
animation.
When the surface was destroyed while animating,
winAnimator.mAnimating was still true, and never got false, because
we didn't call stepAnimation on the WindowAnimator anymore.
Thus, we were constantly animating, using a lot of battery (surface
flinger was constantly pushing frames) and impacting overall system
performance.
Bug: 27276043
Change-Id: I8b0943da0908c85740dc1113aedb573fb440f162
Packages that are entitled to schedule jobs on behalf of other uids
are not subject to the limit. Also break the JobStore's monolithic
set of jobs into per-uid slices for efficiency and orthogonality.
Bug 27150350
Change-Id: I8f5f718bf200d55f9977a6fc53b7f617e7652ad9
This is a safe refactoring that changes nothing.
In order to improve the keyboard dismissal lags [1][2], we have used
IMMS#windowGainedFocus() as a combined event to do startInput() in
certain situations.
To make the intent of those CLs clear, this CL renames
IMMS#windowGainedFocus() to IMMS#startInputOrWindowGainedFocus(). Note
that these are @hide internal IPC protocols. Hence this change is never
observable to application developers.
[1] I8494cbd6e19e2ab6db03f2463d9906680dda058b
a82ba54b0bbc3ff41f29db3998806cb45b261d58
[2] Icb58bef75ef4bf9979f3e2ba88cea20db2e2c3fb
7663d80f6b6fd6ca7a736c3802013a09c0abdeb9
Bug: 25373872
Change-Id: I56934f18e30d90fcdf77bcbb0c35a92a5feb1b82
See thread: "key presses & Ambient"
Bluetooth media button presses and Stem buttons were being rejected in ambient. This change allows those keys while disabling touch events while in ambient.
BUG: 25964771
BUG: 26524301
Change-Id: I914edd447c38678a06d27677be4597c0ee384635
(cherry picked from commit 7528f57c4ddbde1c6d0a2a352bd8a1af43711013)
