InputChannels are normally duplicated when sent to a remote process
over Binder but this does not happen if the recipient is running within
the system server process. This causes problems for KeyGuard because the
InputMethodManagerService may accidentally dispose the channel
that KeyGuard is using.
Fixed the lifecycle of InputChannels that are managed by the IME
framework. We now return a duplicate of the channel to the application
and then take care to dispose of the duplicate when necessary.
In particular, InputBindResult disposes its InputChannel automatically
when returned through Binder (using PARCELABLE_WRITE_RETURN_VALUE).
Bug: 8493879
Change-Id: I08ec3d13268c76f3b56706b4523508bcefa3be79
Many media files and source code files were marked as executable in Git.
Remove those.
Also a shell script and python script were not marked as executable.
Change-Id: Ieb51bafb46c895a21d2e83696f5a901ba752b2c5
The input system needs to know about the window that has
focus, even if it is on a secondary display. So now we
send it the list of all windows and indicate which display
they are on. We filter the list of windows as necessary
when delivering touch events.
To keep things simple, monitor input channels and input
filters are not supported except on the main display.
We also do not pass the display id to applications; it is
only used inside the input system for now.
Properly scale touch coordinates based on the viewport.
This will be needed to ensure that touch works on external
display as well as when the internal display is being used
to simulate a different resolution.
Change-Id: I1815579a52fcc852c519b5391fc7ab8767c2dc59
Moved some duplicate code from SearchPanelView and LockScreen
over to SearchManager to avoid creating yet another copy of it
in PhoneWindowManager.
Bug: 6594275
Change-Id: Ib4ebcd6817639d17548952ab2ce7cb876c05777c
For comparison purposes, port the legacy velocity tracker
algorithm as it behaved prior to ICS.
Bug: 6413587
Change-Id: I7e8e56584dcdb1a3c660ca9d8f9c5bd5d868e449
This algorithm better tolerates certain kinds of errors in the
touch input than the least squares strategy but it may underestimate
the velocity of accelerating movements. This algorithm is
mainly of interest as a baseline for testing and comparison with the
least squares algorithm, which remains the default.
Bug: 6413587
Change-Id: I8ddb50084e44875e234717907e5b06d03f59788c
This change is very useful for testing purposes because it makes it
easy to compare different implementations to see how they behave.
There is no change to the current default strategy.
Bug: 6413587
Change-Id: I4d8567aa4160571ba9fa397ce419882cd9366749
Some input devices do not generate ACTION_MOVE events while all
pointers have stopped, thereby lulling the VelocityTracker into
a false sense of complacency. Before handling the following sample,
reset the VelocityTracker state so as not to be influenced by
earlier samples before the pointer stopped. The velocity after
stopping is assumed to be discontinuous.
Bug: 6413587
Change-Id: I6387bc036ff141d083d3d17a89e37eeaa3188349
Fixed a few bugs related to the id-to-index mapping for
pointer coordinates.
Tightened the bounds on the resampling time interval to
avoid predicting too far into the future.
Only lerp X and Y components of motion events.
Alter the future to satisfy past predictions. (Rewrite touch
events to conceal obvious discontinuities.)
Added a system property to control whether resampling is enabled
for debugging purposes.
Bug: 6375101
Change-Id: I35972d63278bc4e78148053a4125ad9abeebfedb
If a fallback key is generated using a key plus a modifier,
then it's possible we might get a different fallback key
generated if the modifier has changed. PhoneWindowManager
needs to remember which fallback is last generated for a
given key code so that it can apply the same fallback action.
When generating cancellation events, it's important to have
preserved the policyFlags of the original event. Otherwise
we may not dispatch the cancellation properly. For example,
some actions are not performed if the POLICY_FLAG_TRUSTED
is not specified.
Remember the metaState associated with a key event so we can
include it when canceled.
Tell the policy when a fallback is being cancelled so that it
can clean up its state.
After a SEARCH shortcut is invoked, clear the flag indicating
that a shortcut is pending. This is to prevent SEARCH from
getting stuck down in the case where we might forget to send
the up. (Shouldn't happen anymore after the prior fixes.)
Bug: 5616255
Change-Id: I68f0a9679c7af464eaf31c099f2aa50b53fecf1f
The window manager policy made some incorrect assumptions about the
meaning of the Configuration.keyboard field. We need to be more
careful about distinguishing between built-in and external keyboards.
Most of this change is to move the determination of the parts of
the Configuration related to input devices into the WindowManagerService
leveraging new features of the InputManagerService to good effect.
Then we plumb through the flag that indicates whether a device
is internal or external so that we can be more particular about
how the lid switch effects changes to the Configuration.
Bug: 6424373
Change-Id: I36a1c22ade35e578955465a25940a33f227b9763
Bluetooth devices can be renamed by the user. Make the
input system aware of the user-specified name and transparently
pass it down to applications. This enables the keyboard
layout picker Settings UI to use device names that are
consistent with what the user set in the Bluetooth UI.
Bug: 6363157
Change-Id: I8eea26ce2c69c2a3f09c8de02e9e847610e0419c
Automatically choose a default value for the 'number' property
based on the characters that the key can generate.
Don't generate any character when ctrl, alt or meta is
pressed unless the behavior exactly matches the modifier keys
that are pressed.
Simplified the basic keyboard layouts taking into account the
new features.
Bug: 6110399
Change-Id: Ibc0f0b50c2dcf3f962a33ac77c24d2993b77637d
Added support for mapping both scan codes and HID usages to
KeyLayoutMap and KeyCharacterMap. Keyboard overlays can
now influence how key events are mapped to key codes.
Bug: 6110399
Change-Id: I6619fd2d3e1337c55928f89869dbc45b535c7ccf
Added the concept of a keyboard layout overlay, which is
a key character map file that has "type OVERLAY".
Added support for loading keyboard layout overlays from
resources dynamically. The layouts are reloaded whenever they
are changed in the Settings application or an application
is installed. This is somewhat more aggressive than necessary
so we might want to optimize it later.
Before system-ready, the input system uses just the generic
keyboard layouts that are included on the device system image.
After system-ready, it considers the user's selected keyboard
layout overlay and attempts to load it as necessary. We need to
wait until system-ready before doing this because we need to
be in a state where it is safe to start applications or access
their resources.
Bug: 6110399
Change-Id: Iae0886d3356649b0d2440aa00910a888cedd8323
Added a getVibrator() method to InputDevice which returns a Vibrator
associated with that input device. Its uses the same API as the
system vibrator which makes it easy for applications to be modified
to use one or the other.
Bug: 6334179
Change-Id: Ifc7f13dbcb778670f3f1c07ccc562334e6109d2e
This change allows the InputManager to keep track of what input
devices are registered with the system and when they change.
It needs to do this so that it can properly clear its cache of
input device properties (especially the key map!) when changes
occur.
Added new API so that applications can register listeners for
input device changes.
Fixed a minor bug in EventHub where it didn't handle EPOLLHUP
properly so it would spam the log about unsupposed epoll events
until inotify noticed that the device was gone and removed it.
Change-Id: I937d8c601f7185d4299038bce6a2934fe4fdd2b3
Added handling for EV_MSC / MSC_SCAN which typically reports
the HID usage associated with a key. This will enable key maps
to map keys with HID usages that Linux does not natively recognize.
Removed keyCode and flags fields from EventHub RawEvent since
they don't necessarily make sense in isolation now that we
pay attention to HID usage codes too.
Removed the fallback code for mapping keys and axes. In practice,
an input device should be self-sufficient. We should not ever
need to look at the built-in keyboard's key map. In fact, there
usually isn't a built-in keyboard anyhow. This code was originally
working around a problem where we weren't loading the key map
for touch screens with virtual keys, which has long since been fixed.
Change-Id: I0a319bdec44be9514f795526347397e94d53a127
Instead of each application loading the KeyCharacterMap from
the file system, get them from the input manager service as
part of the InputDevice object.
Refactored InputManager to be a proper singleton instead of
having a bunch of static methods.
InputManager now maintains a cache of all InputDevice objects
that it has loaded. Currently we never invalidate the cache
which can cause InputDevice to return stale motion ranges if
the device is reconfigured. This will be fixed in a future change.
Added a fake InputDevice with ID -1 to represent the virtual keyboard.
Change-Id: If7a695839ad0972317a5aab89e9d1e42ace28eb7
The purpose of the input device descriptor is to make it possible
to associate persistent settings for each input device, such as the
keyboard layout.
The descriptor is a hash of the information we have about the
device, such as its vendor id, product id, unique id, name,
or location.
Bug: 6110399
Change-Id: Idb80f946819b3f0dbf4e661bb0a753dbc2b60981
These keys are specific to Japanese hardware keyboard which can be
used by input method.
Patch ported from AOSP, with the addition of EISU and KANA mappings.
Change-Id: I647473cdd257458e3b9d134b0fc623eae946c3e0
ZipUtils is needed by build/tools, move it from libandroidfw
(frameworks/base) to libutils (frameworks/native).
Change-Id: I2b4b7adcdf68eb25ee7cba5dd3b69eadf0523af3
Some native activities experienced ANRs when the input consumer
deferred an input event due to client-side batching. If the
input channel was fully emptied then the client had no way of
knowing that it should wake up to handle the deferred input event.
This patch also fixes some lock issues in the native activity
input queue implementation. In at least one error case, it
was possible for a function to exit without releasing the lock.
Bug: 6051176
Change-Id: I4d9d843237e69b9834f8d8b360031b677fcab8c3
