This introduces four generic thread that services can
use in the system process:
- Background: part of the framework for all processes, for
work that is purely background (no timing constraint).
- UI: for time-critical display of UI.
- Foreground: normal foreground work.
- IO: performing IO operations.
I went through and moved services into these threads in the
places I felt relatively comfortable about understanding what
they are doing. There are still a bunch more we need to look
at -- lots of networking stuff left, 3 or so different native
daemon connectors which I didn't know how much would block,
audio stuff, etc.
Also updated Watchdog to be aware of and check these new
threads, with a new API for other threads to also participate
in this checking.
Change-Id: Ie2f11061cebde5f018d7383b3a910fbbd11d5e11
The window manager now keeps track of the overscan of
each display, with an API to set it. The overscan impacts
how it positions windows in the display. There is a new set
of APIs for windows to say they would like to go into the
overscan region. There is a call into the window manager to
set the overscan region for a display, and it now has a
concept of display settings that it stores presistently.
Also added a new "wm" command, moving the window manager
specific commands from the "am" command to there and adding
a new now to set the overscan region.
Change-Id: Id2c8092db64fd0a982274fedac7658d82f30f9ff
If a rotation occurred while the electron beam surface was showing,
the surface may have appeared in the wrong orientation. We fix this
problem by adjusting the transformation matrix of the electron beam
surface according to the display orientation whenever a display
transaction occurs.
The rotation itself is allowed to proceed but it is not visible
to the user. We must let this happen so that the lock screen
is correctly oriented when the screen is turned back on.
Note that the electron beam surface serves two purposes.
First, it is used to play the screen off animation.
When the animation is finished, the surface remains visible but is
solid black. Then we turn the screen off.
Second, when we turn the screen back on we leave the electron beam
surface showing until the window manager is ready to show the
new content. This prevents the user from seeing a flash of the
old content while the screen is being turned on. When everything is
ready, we dismiss the electron beam.
It's important for the electron beam to remain visible for
the entire duration from just before the screen is turned off until
after the screen is turned on and is ready to be seen. This is
why we cannot fix the bug by deferring rotation or otherwise
getting in the way of the window manager doing what it needs
to do to get the screen ready when the screen is turned on again.
Bug: 7479740
Change-Id: I2fcf35114ad9b2e00fdfc67793be6df62c8dc4c3
Calling blank() on Surface Flinger to turn the screen off is not
enough to ensure that the content is blanked to all virtual displays.
What's more, the black surface left in place by the ElectronBeam may
not completely hide the content (particularly if the display orientation
changes). To fix this for real, we'll want to move the display power
management code from the power manager into the display manager
but we don't have time for that.
As a work around, force all displays to show an empty layer stack
with no surfaces on it while blanked.
Bug: 7311959
Change-Id: I870c985f9e76f3f2322e5d83cdbbed9ed15b9f10
WindowManager was notifying DisplayManager of content if any window
existed on a display. Now the window must be visible and we must not
be showing a Dream or the Keyguard.
Bug: 7214060.
Change-Id: I9ce4a49aabfbac22ff1e39a837199ce35b9f7503
Add a setting to globally disable Wifi display.
Fixed a bug where the wifi display broadcast receiver
was running on the wrong thread.
Removed the wifi-display QuickSettings dialog, all functionality
has been moved to Settings.
Bug: 7178216
Bug: 7192799
Change-Id: I9796baac8245d664cf28fa147b9ed978d81d8ab9
The window manager is no longer responsible for telling the
input system about the display viewport.
Change-Id: I932882bae55decef55f25093bb2a7ebac1620bb1
The API is quite simple. There are a few extra functions
on DisplayManager to scan, connect and disconnect from
wifi displays and get status, and a single protected
broadcast sent when the status changes.
Change-Id: Ic91dbab5ee818e790b27fa32e1a1e93788793be0
Tell the display manager whenever a given logical display
contains interesting windows. If so, then the display
manager arranges to show that content on a physical display,
otherwise it ignores the logical display and makes its
associated primary physical display mirror the default
display.
Assign DisplayContents when Displays are added, remove them when
Displays are removed, and update the DisplayInfo when Displays
change.
Change-Id: I36e08ec538055acabe1e24cdd12c40de4e47a158
Added more complete support for logical displays with
support for mirroring, rotation and scaling.
Improved the overlay display adapter's touch interactions.
A big change here is that the display manager no longer relies
on a single-threaded model to maintain its synchronization
invariants. Unfortunately we had to change this so as to play
nice with the fact that the window manager wants to own
the surface flinger transaction around display and surface
manipulations. As a result, the display manager has to be able
to update displays from the context of any thread.
It would be nice to make this process more cooperative.
There are already several components competing to perform
surface flinger transactions including the window manager,
display manager, electron beam, overlay display window,
and mouse pointer. They are not manipulating the same surfaces
but they can collide with one another when they make global
changes to the displays.
Change-Id: I04f448594241f2004f6f3d1a81ccd12c566bf296
Split the DisplayManager into two parts. One part is bound
to a Context and takes care of Display compatibility and
caching Display objects on behalf of the Context. The other
part is global and takes care of communicating with the
DisplayManagerService, handling callbacks, and caching
DisplayInfo objects on behalf of the process.
Implemented support for enumerating Displays and getting
callbacks when displays are added, removed or changed.
Elaborated the roles of DisplayManagerService, DisplayAdapter,
and DisplayDevice. We now support having multiple display
adapters registered, each of which can register multiple display
devices and configure them dynamically.
Added an OverlayDisplayAdapter which is used to simulate
secondary displays by means of overlay windows. Different
configurations of overlays can be selected using a new
setting in the Developer Settings panel. The overlays can
be repositioned and resized by the user for convenience.
At the moment, all displays are mirrors of display 0 and
no display transformations are applied. This will be improved
in future patches.
Refactored the way that the window manager creates its threads.
The OverlayDisplayAdapter needs to be able to use hardware
acceleration so it must share the same UI thread as the Keyguard
and window manager policy. We now handle this explicitly as
part of starting up the system server. This puts us in a
better position to consider how we might want to share (or not
share) Loopers among components.
Overlay displays are disabled when in safe mode or in only-core
mode to reduce the number of dependencies started in these modes.
Change-Id: Ic2a661d5448dde01b095ab150697cb6791d69bb5
Cleaned up the implementation of Surface and SurfaceSession
to use more consistent naming and structure.
Added JNI for all of the new surface flinger display API calls.
Enforced the requirement that all Surfaces created by
the window manager be named.
Updated the display manager service to use the new methods.
Change-Id: I2a658f1bfd0437e1c6f9d22df8d4ffcce7284ca2
Reverting to the previous stub as the display adapter registration
and the logical to physical mapping is not at all what we are going
to need moving forward.
Fixed up the service initialization order so that the display manager
service has a context from the start.
Change-Id: I717f2f1099c7a77180ef207c371ec8329258850a
This puts in most of the infrastructure needed to allow us to
switch between different densities at run time. The main remaining
uses of the global are to initialize the Bitmap object (not sure
what to do about that since it doesn't have anything passed in
the constructor to get this information from), and being able to
load drawables if we need a different density than what was preloaded
by zygote.
Change-Id: Ifdbfd6b7a5c59e6aa22e63b95b78d96af3d96848
The purpose of this change is to remove direct reliance on
SurfaceFlinger for describing the size and characteristics of
displays.
This patch also starts to make a distinction between logical displays
and physical display devices. Currently, the window manager owns
the concept of a logical display whereas the new display
manager owns the concept of a physical display device.
Change-Id: I7e0761f83f033be6c06fd1041280c21500bcabc0
