This fixes the logic on the death handlers for notification listeners,
and doesn't unbind from the listener services so that the system will
bring them back up again.
Bug: 12587702
Change-Id: I44ce250e0e1c2583836dc823d9a333dabec51df9
The dreams manager also manages dozing. It has a minimal footprint
so there is no real reason to disable the component (it just makes
debugging more difficult).
Improved the documentation of the config_dreamsSupported resource
to clarify exactly what it controls.
Bug: 12494706
Change-Id: I78244846f7c1ddfd11bc1605af59b0db91337971
When the ActivityView is part of the home activity special checks
must be made. Things like don't move the home stack to the back
when the ActivityView activity is resumed.
Fixes bug 13119389.
Change-Id: I3a6040c9824dfd4b8ee97d58d131b14a519b470a
A task is scheduled for deletion after the final activity has
been removed and has animated away. But if another activity is then
added to the task the deletion flag must be reset.
Also added improved debugging.
Fixes bug 12987986.
Change-Id: I207ea6e9592a9e036d67aa5d1465b4acc5bdd120
Using the alpha value to trigger a resizing of the DimLayer was a
Bad Idea. The alpha value should reflect the true alpha value and
not be used to trick the code. Actually changing the size and
position is a Better Idea.
Fixes bug 13101776.
Change-Id: I11c16b8276919ea85960fe87bb17c0956ce8a3b1
When a doze component has been specified in a config.xml resource
overlay, the power manager will try to start a preconfigured dream
whenever it would have otherwise gone to sleep and turned the
screen off. The dream should render whatever it intends to show
then call startDozing() to tell the power manager to put the display
into a low power "doze" state and allow the application processor
to be suspended. The dream may wake up periodically using the
alarm manager or other features to update the contents of the display.
Added several new config.xml resources related to dreams and dozing.
In particular for dozing there are two new resources that pertain to
decoupling auto-suspend mode and interactive mode from the display
state. This is a requirement to enable the application processor
and other components to be suspended while dozing. Most devices
do not support these features today.
Consolidated the power manager's NAPPING and DREAMING states into one
to simplify the logic. The NAPPING state was mostly superfluous
and simply indicated that the power manager should attempt to start
a new dream. This state is now tracked in the mSandmanSummoned field.
Added a new DOZING state which is analoguous to DREAMING. The normal
state transition is now: AWAKE -> DREAMING -> DOZING -> ASLEEP.
The PowerManager.goToSleep() method now enters the DOZING state instead
of immediately going to sleep.
While in the doze state, the screen remains on. However, we actually
tell the rest of the system that the screen is off. This is somewhat
unfortunate but much of the system makes inappropriate assumptions
about what it means for the screen to be on or off. In particular,
screen on is usually taken to indicate an interactive state where
the user is present but that's not at all true for dozing (and is
only sometimes true while dreaming). We will probably need to add
some more precise externally visible states at some point.
The DozeHardware interface encapsulates a generic microcontroller
interface to allow a doze dream for off-loading rendering or other
functions while dozing. If the device possesses an MCU HAL for dozing
then it is exposed to the DreamService here.
Removed a number of catch blocks in DreamService that caught Throwable
and attempted to cause the dream to finish itself. We actually just
want to let the process crash. Cleanup will happen automatically if
needed. Catching these exceptions results in mysterious undefined
behavior and broken dreams.
Bug: 12494706
Change-Id: Ie78336b37dde7250d1ce65b3d367879e3bfb2b8b
When switching USB modes there are often spurious connect and disconnect events
that occur before reenumeration is complete. There is currently a 1000ms timer
to "debounce" the disconnect events. But with some USB accessories, this timeout
is not long enough, which results in an endless cycle of attempts to enter
USB accessory mode when the phone is connected.
To fix this, we now wait up to 10 seconds for the host to successfully configure
the device when entering USB accessory mode before giving up.
This is separate from the existing debounce timer, so the behavior of the
USB state change broadcasts are not affected.
Bug: 12877769
Change-Id: I7aa61f8a618546d749a7ddfc97bf103029a73d03
Converts surfaces from transparent to opaque and opaque to transparent
without creating a new surface. Uses the new SurfaceControl.setOpaque
method.
Fixes bug 12387406.
Change-Id: I669c064e622e211b00b1585183a488d5b3f4b778
PackageManager does not need to create a hash set
when comparing two sets of signatures, where each set
contains one signature. This is the common case for
most applications.
Bug:13029313
Change-Id: I61883d85960db7b376e234d7abae6149c4d20245
Required replumbing and updating the TaskStack bounds calculations
to match the new separation of TaskStack from DisplayContent.
Fixes bug 12780687.
Change-Id: I061c92831c06f4eb3e673ad6296e721b0c0f3202
If an apptoken's removal has been delayed then it will be in both
mExitingAppTokens and the stack's tasks. When rebuilding the windows
the windows from that apptoken will get added twice. This change
skips adding those windows from apptokens that have already been
added from mExitingAppTokens.
Change-Id: I55f94b7dbef36aeacb390b91fbe82426eaac5458
Check explicitly for null listeners in NMS, throwing
IllegalArgumentException (on the small list of exceptions
that survive RPC boundaries) with a message.
Normally this situation is caused by listeners that attempt to
perform NM-related actions before they are bound. Check for
this case in the base NLS class and avoid the call to NM if we
know it will fail.
Although it's tempting to throw an IllegalStateException on the
client side, preserve the existing semantics for backwards-compatibility
purposes. That is, silently fail (or return null) - and provide a
log warning.
Bug:12805707
Change-Id: I0d92fd0d460a8592e8a23fd8fd718ae2ba3bd4c7
Modified ActivityView to inject touch events it receives back into
its activity container. The container then injects the event into
the input system along with the display id of the underlying virtual
display.
Enhanced the input system to support concurrent dispatch of touch
events on multiple displays which is required for this to work.
Change-Id: I9cf1870db3be6f99a52ed9a1e3ceafe42c940093
This change simplifies the process of initializing a SystemService
by folding the onCreate() step back into the constructor. It removes
some ambuiguity about what work should happen in the constructor and
should make it possible for services to retain most of their final
fields after refactoring into the new pattern.
Change-Id: I25f41af0321bc01898658ab44b369f9c5d16800b
- Remove task after removing activity. Removing the task before
removing the activity led to problems when the task was still needed.
- In WindowManager wait for activity manager to remove the task.
Previously was automatically removing the task when the last activity
was removed.
Fixes bug 12957434.
Change-Id: I7f6b1f09a2e2a845f177f337769075c9376f8bdb
Just like stacks and displays, activities and tasks need to stick
around until animations have completed.
Change-Id: I54fe8f6855d60cbc3a25cbc6e762defd5ac50bf5
Changing Configuration first and then testing for changes yields a
result indicating no change.
Fixes bug 12904769.
Change-Id: If7e39e843f15b1143d9877497d595511afabd020
Allowing the detach of ActivityStack from DisplayContent to happen
immediately was causing all sorts of problems associated with not
having a Display to complete the animations.
Waiting for animations to complete before either the detach or the
display removal fixes those problems.
Change-Id: I8a5663bfac5c3c1084ff4fcc451e0e38e8080265
Now that activities and tasks can be detached from a display do not
keep their state within the DisplayContent class. All such state is
now moved to the TaskStack class with this change.
Change-Id: Ie6b1811e9ec3f545782bad8c3461bd4f2d336b93
At startup, we check with PackageManager whether a system service is
available before attempting to load it. A system service is available
if its associated feature (similar to hardware features) is present.
This does not remove unavailable services from the compiled jar.
Change-Id: I13571805083aa4e65519a74acb52efd17b9fb3d7
PendingIntents and IntentSenders can now be launched. Still does not
work once the host activity has been paused and resumed.
Window manager TaskStacks now exist independently of Displays and app
windows persist after Displays are removed below them. Attaching the
stack to a new Display does not yet restore the windows to it.
Fixes bug 12747909.
Change-Id: I509007ee23fda400b353f483cf6ecce08177763b
