Rewrote interceptKeyBeforeQueueing to make the handling more systematic.
Behavior should be identical except:
- We never pass keys to applications when the screen is off and the keyguard
is not showing (the proximity sensor turned off the screen).
Previously we passed all non-wake keys through in this case which
caused a bug on Crespo where the screen would come back on if a soft key
was held at the time of power off because the resulting key up event
would sneak in just before the keyguard was shown. It would then be
passed through to the dispatcher which would poke user activity and
wake up the screen.
- We propagate the key flags when broadcasting media keys which
ensures that recipients can tell when the key is canceled.
- We ignore endcall or power if canceled (shouldn't happen anyways).
Changed the input dispatcher to not poke user activity for canceled
events since they are synthetic and should not wake the device.
Changed the lock screen so that it does not poke the wake lock when the
grab handle is released. This fixes a bug where the screen would come
back on immediately if the power went off while the user was holding
one of the grab handles because the sliding tab would receive an up
event after screen turned off and release the grab handles.
Bug: 3144874
Change-Id: Iebb91e10592b4ef2de4b1dd3a2e1e4254aacb697
This change fixes a bug in the dispatcher where the window manager
policy would incorrectly receive a key repeat count of 0 in the case
where the key repeat was generated by the hardware or driver.
Long-press on HOME was broken as a result.
Repeating keys could also get stuck down.
Bug: 3159581
Bug: 3208156
Change-Id: I1145487cfcc41a7850dba4cafc63c4a5951ace5b
Two issues:
1. First, due to an inverted conditional in the input dispatcher, we were
reporting touches as long touches and vice-versa to the power manager.
2. Power manager user activity cheek event suppression also suppresses touch
events (but not long touch or up events). As a result, if cheek event
suppression was enabled, touches would not poke the user activity timer.
However due to the above logic inversion, this actually affected long
touches. Net result, if cheek suppression was enabled in the power manager
and you held your thumb on the screen long enough, the phone would
go to sleep!
Cheek event suppression is commonly turned on when making a phone call.
Interestingly, it does not seem to get turned off afterward...
This change fixes the logic inversion and exempts touches from the cheek
suppression. The reason we do the latter is because the old behavior
was actually harmful in other ways too: a touch down would be suppressed
but not a long touch or the touch up. This would cause bizarre behavior
if you touched the screen while it was dimmed. Instead of brightening
immediately, it would brighten either when you lifted your finger or
300ms later, whichever came first.
Bug: 3154895
Change-Id: Ied9ccec6718fbe86506322ff47a4e3eb58f81834
This patch makes the dispatcher drop all of its state when it is
disabled (when the screen turns off). This ensures that the dispatcher
does not get stuck thinking a pointer is still down if the screen
turned off while the user was touching the display (such as a fat touch
while hitting the power button).
Bug: 3098344
Change-Id: If50ef5804870aa1acd3179fd4b40e3cda58dd39d
Fixed a bug where we would lose the first touch point when swiping out of
the virtual key area.
Fixed a bug where we would not send an ACTION_MOVE event in cases where
individual pointers went down/up and the remaining pointers actually moved.
This is important since many applications do not handle pointer movements
during ACTION_POINTER_DOWN or ACTION_POINTER_UP. In the case of
ACTION_POINTER_UP the movement was completely lost since all pointers were
dispatched using their old location rather than the new location.
Improved motion event validation to check for duplicate pointer ids.
Added an input source constant that was missing from the NDK api but
defined in the framework api.
Added a timestamp when reporting added/removed devices in EventHub.
Bug: 3070082
Change-Id: I3206a030f43b7616e2f48006e5a9d522c4d92e56
We now poke user activity twice: once upon dequeueing an event
for dispatch and then again just before we dispatch it. The second
poke is to compensate for the fact that it can take a few seconds to
identify the dispatch target (if the application is responding slowly)
but we want to keep the display from going to sleep for X amount of time
after the app gets a chance to actually receive the event. This mirrors
pre-Gingerbread behavior.
Removed some unnecessary code that filters user activity pokes when sending
events to KeyGuard. We don't need this because KeyGuard already tells the
power manager to disable user activity.
Bug: 3101397
Change-Id: I8c3a77601fdef8f584e84cfdd11aa79da0ff51db
This change adds a new window type for secure system overlays
created by the system itself from non-secure system overlays that
might be created by applications that have the system alert permission.
Secure views ignore the presence of secure system overlays.
Bug: 3098519
Change-Id: I8f8398f4fdeb0469e5d71124c21bedf121bd8c07
Fixed some issues with Monkeys turning off their own screens. Ook ook!
Added some more comments to explain what's going on.
Change-Id: Id2bc0466161a642a73ef7ef97725d1c81e984b12
Added the concept of a "trusted" event to distinguish between events from
attached input devices or trusted injectors vs. other applications.
This change enables us to move certain policy decisions out of the
dispatcher and into the policy itself where they can be handled more
systematically.
Change-Id: I4d56fdcdd31aaa675d452088af39a70c4e039970
This change fixes several issues where events would be dropped in the
input dispatch pipeline in such a way that the dispatcher could not
accurately track the state of the input device.
Given more robust tracking, we can now also provide robust cancelation
of input events in cases where an application might otherwise become
out of sync with the event stream due to ANR, app switch, policy decisions,
or forced focus transitions.
Pruned some of the input dispatcher log output.
Moved the responsibility for calling intercept*BeforeQueueing into
the input dispatcher instead of the input reader and added support for
early interception of injected events for events coming from trusted
sources. This enables behaviors like injection of media keys while
the screen is off, haptic feedback of injected virtual keys, so injected
events become more "first class" in a way.
Change-Id: Iec6ff1dd21e5f3c7feb80ea4feb5382bd090dbd9
Added dumpsys reporting to EventHub.
Made the formatting a bit clearer.
Added 'Locked' suffix to some internal methods of EventHub.
Change-Id: Ic449560bcce378f6361895d27c66854e9724abb0
This feature is currently used to enable dragging the start and end
selection handles of a TextView at the same time. Could be used for
other things later.
Deleted some dead code in ArrowKeyMovementMethod and CursorControllers.
Change-Id: I930accd97ca1ca1917aab8a807db2c950fc7b409
Redesigned the input dispatcher's ANR timeout mechanism so it is much
closer to Froyo's policy. ANR is only ever signalled if the dispatcher
is waiting on a window to finish processing its previous event(s) and
there is new pending input.
In the old code, we tracked the dispatch timeout separately for each
input channel. This was somewhat complicated and also resulted in the
situation where applications could ANR long after the user had pushed
them into the background.
Change-Id: I666ecada0952d4b95f1d67b9f733842b745c7f4b
As part of this change, consolidated and cleaned up the Looper API so
that there are fewer distinctions between the NDK and non-NDK declarations
(no need for two callback types, etc.).
Removed the dependence on specific constants from sys/poll.h such as
POLLIN. Instead looper.h defines events like LOOPER_EVENT_INPUT for
the events that it supports. That should help make any future
under-the-hood implementation changes easier.
Fixed a couple of compiler warnings along the way.
Change-Id: I449a7ec780bf061bdd325452f823673e2b39b6ae
This change is essentially a rewrite of the main input dispatcher loop
with the target identification folded in. Since the input dispatcher now
has all of the window state, it can make better decisions about
when to ANR.
Added a .5 second deadline for processing app switch keys. This behavior
predates Gingerbread but had not previously been ported.
Fixed some timing inaccuracies in the ANR accounting that could cause
applications to ANR sooner than they should have.
Added a mechanism for tracking key and motion events that have been
dispatched to a window so that appropriate cancelation events can be
synthesized when recovering from ANR. This change helps to keep
applications in sync so they don't end up with stuck buttons upon
recovery from ANRs.
Added more comments to describe the tricky parts of PollLoop.
Change-Id: I13dffca27acb436fc383980db536abc4d8b9e6f1
Added the MotionEvent.FLAG_WINDOW_IS_OBSCURED flag which is set by the
input manager whenever another visible window is partly or wholly obscured
the target of a touch event so that applications can filter touches
accordingly.
Added a "filterTouchesWhenObscured" attribute to View which can be used to
enable filtering of touches when the view's window is obscured.
Change-Id: I936d9c85013fd2d77fb296a600528d30a29027d2
Finished the input device capability API.
Added a mechanism for calibrating touch devices to obtain more
accurate information about the touch contact area.
Improved pointer location to show new coordinates and capabilities.
Optimized pointer location display and formatting to avoid allocating large
numbers of temporary objects. The GC churn was causing the application to
stutter very badly when more than a couple of fingers were down).
Added more diagnostics.
Change-Id: Ie25380278ed6f16c5b04cd9df848015850383498
The code now takes into account the time when the event was generated.
So with a 60Hz throttling rate, it ensures that a motion event will not
be delayed due to throttling by any more than 1/60th of second past
the time it was generated.
Change-Id: Iaea1d4f76e79036b4a18873485230731c3cd63c3
Previously, the input dispatcher assumed that the input channel's
receive pipe file descriptor was a sufficiently unique identifier for
looking up input channels in its various tables. However, it can happen
that an input channel is disposed and then a new input channel is
immediately created that reuses the same file descriptor. Ordinarily
this is not a problem, however there is a small opportunity for a race
to arise in InputQueue.
When InputQueue receives an input event from the dispatcher, it
generates a finishedToken that encodes the channel's receive pipe fd,
and a sequence number. The finishedToken is used by the ViewRoot
as a handle for the event so that it can tell the InputQueue when
the event has finished being processed.
Here is the race:
1. InputQueue receives an input event, assigns a new finishedToken.
2. ViewRoot begins processing the input event.
3. During processing, ViewRoot unregisters the InputChannel.
4. A new InputChannel is created and is registered with the Input Queue.
This InputChannel happens to have the same receive pipe fd as
the one previously registered.
5. ViewRoot tells the InputQueue that it has finished processing the
input event, passing along the original finishedToken.
6. InputQueue throws an exception because the finishedToken's receive
pipe fd is registered but the sequence number is incorrect so it
assumes that the client has called finish spuriously.
The fix is to include a unique connection id within the finishedToken so
that the InputQueue can accurately confirm that the token belongs to
the currently registered InputChannel rather than to an old one that
happened to have the same receive pipe fd. When it notices this, it
ignores the spurious finish.
I've also made a couple of other small changes to avoid similar races
elsewhere.
This patch set also includes a fix to synthesize a finished signal
when the input channel is unregistered on the client side to
help keep the server and client in sync.
Bug: 2834068
Change-Id: I1de34a36249ab74c359c2c67a57e333543400f7b
Added a new asynchronous injection mode and made the existing
synchronization mechanism more robust.
Change-Id: Ia4aa04fd9b75ea2461a844c5b7933c831c1027e6
Added several new coordinate values to MotionEvents to capture
touch major/minor area, tool major/minor area and orientation.
Renamed NDK input constants per convention.
Added InputDevice class in Java which will eventually provide
useful information about available input devices.
Added APIs for manufacturing new MotionEvent objects with multiple
pointers and all necessary coordinate data.
Fixed a bug in the input dispatcher where it could get stuck with
a pointer down forever.
Fixed a bug in the WindowManager where the input window list could
end up containing stale removed windows.
Fixed a bug in the WindowManager where the input channel was being
removed only after the final animation transition had taken place
which caused spurious WINDOW DIED log messages to be printed.
Change-Id: Ie55084da319b20aad29b28a0499b8dd98bb5da68
This factors out the boiler-plate code from the sample
app to a common glue code that can be used for everyone
writing this style of app: a dedicated app thread that
takes care of waiting for events and processing them.
As part of doing this, ALooper has a new facility to allow
registration of fds that cause ALooper_pollOnce() to return
the fd that has data, allowing the app to drive the loop
without callbacks. Hopefully this makes some people feel better. :)
Also do some other cleanup of the ALooper API, plus some
actual documentation.
Change-Id: Ic53bd56bdf627e3ba28a3c093faa06a92be522b8
Removed old input dispatch code.
Refactored the policy callbacks.
Pushed a tiny bit of the power manager state down to native.
Fixed long press on MENU.
Made the virtual key detection and cancelation a bit more precise.
Change-Id: I5d8c1062f7ea0ab3b54c6fadb058c4d5f5a9e02e
Set a default orientation of ROTATION_0.
Added some more careful checks based on whether we have valid
absolute axis information from the driver.
Reset key repeating during configuration changes since the keyboard
device may have been removed.
Change-Id: I685960828acffcb17595fc5683309e8064a76714
Target identification is now fully native.
Fixed a couple of minor issues related to input injection.
Native input enabled by default, can be disabled by setting
WindowManagerPolicy.ENABLE_NATIVE_INPUT_DISPATCH to false.
Change-Id: I7edf66ed3e987cc9306ad4743ac57a116af452ff
Added ANRs handling.
Added event injection.
Fixed a NPE ActivityManagerServer writing ANRs to the drop box.
Fixed HOME key interception.
Fixed trackball reporting.
Fixed pointer rotation in landscape mode.
Change-Id: I50340f559f22899ab924e220a78119ffc79469b7
Added more tests.
Fixed a regression in Vector.
Fixed bugs in pointer tracking.
Fixed a starvation issue in PollLoop when setting or removing callbacks.
Fixed a couple of policy nits.
Modified the internal representation of MotionEvent to be more
efficient and more consistent.
Added code to skip/cancel virtual key processing when there are multiple
pointers down. This helps to better disambiguate virtual key presses
from stray touches (such as cheek presses).
Change-Id: I2a7d2cce0195afb9125b23378baa94fd2fc6671c
Refactored the code to eliminate potential deadlocks due to re-entrant
calls from the policy into the dispatcher. Also added some plumbing
that will be used to notify the framework about ANRs.
Change-Id: Iba7a10de0cb3c56cd7520d6ce716db52fdcc94ff
The old dispatch mechanism has been left in place and continues to
be used by default for now. To enable native input dispatch,
edit the ENABLE_NATIVE_DISPATCH constant in WindowManagerPolicy.
Includes part of the new input event NDK API. Some details TBD.
To wire up input dispatch, as the ViewRoot adds a window to the
window session it receives an InputChannel object as an output
argument. The InputChannel encapsulates the file descriptors for a
shared memory region and two pipe end-points. The ViewRoot then
provides the InputChannel to the InputQueue. Behind the
scenes, InputQueue simply attaches handlers to the native PollLoop object
that underlies the MessageQueue. This way MessageQueue doesn't need
to know anything about input dispatch per-se, it just exposes (in native
code) a PollLoop that other components can use to monitor file descriptor
state changes.
There can be zero or more targets for any given input event. Each
input target is specified by its input channel and some parameters
including flags, an X/Y coordinate offset, and the dispatch timeout.
An input target can request either synchronous dispatch (for foreground apps)
or asynchronous dispatch (fire-and-forget for wallpapers and "outside"
targets). Currently, finding the appropriate input targets for an event
requires a call back into the WindowManagerServer from native code.
In the future this will be refactored to avoid most of these callbacks
except as required to handle pending focus transitions.
End-to-end event dispatch mostly works!
To do: event injection, rate limiting, ANRs, testing, optimization, etc.
Change-Id: I8c36b2b9e0a2d27392040ecda0f51b636456de25
