- Added ActivityOption to mark a starting activity as a taskOverlay
activity. That is the activity will always be the top activity of the
task and doesn't cause the task to be moved to the front when it is added.
- Only set the starting window state of the ActivityRecord to shown if
window manager actually showed the starting window for the activity.
Avoids incorrectly trying to remove starting window for an activity that
didn't show any.
- When starting additional activity in a task, transfer the starting
window from the top most activity with a starting window. It is possible
the top most window does have a starting window like in the case of the
forcedResized activity.
- Only ensure visiblity of an activity we are starting in a task whose top
activity is a task overlay. They need to start in the visible-paused state
and not the resumed state which just causes extra churn in the system.
- Always add additional starting activities in a task with an overlay
activity below the overlay activity.
Bug: 28751186
Change-Id: I3624a4313ae9c406d42c67a3537f67ad685791af
We don't need to set up JIT profiles and register usage etc when
the package context we're trying to construct doesn't request code.
This will correct accounting for packages which are only used for
resources.
bug: 28519185
Change-Id: I849675efa76c8100ae937de478b52254babe384c
We need to make IIntentSender oneway... but when the system is
calling that for itself, it needs to be able to return a result code.
Solution: instead of directly calling the interface, we have a new
IPC through the activity manager. If the thing being used is the
activity manager impl, it can do the synchronous send and return
the result directly in place. If not, you only get asynchronous
sending and thus never a failure result back (too bad for you!).
Change-Id: I4096e5b00063e8dba66230585a2dfe67e35e8092
am: d29c26073e
* commit 'd29c26073e97e4c6e7825641cf6e76720df395e3':
docs: Updates to multi-window and related docs.
Change-Id: Ic4be60debb41d074d717bfb0989125299428831a
am: c88130c572
* commit 'c88130c5724227b3ba7ef0b5ef4476fedabca650':
docs: Updates to multi-window and related docs.
Change-Id: I308c988e3a93737a478f9f2445b512e86f218643
Framework edition
Add a matching onAttachFragment method to Fragment to match the
fragment host version.
Bug 28760393
Change-Id: I5f50b3446449cae7110da6b4e468ee80f413e1e5
Clarified behavior when activity is resized or put in fullscreen
mode, per b/28580007 . Also (per email from o-o) removed misleading
statement about when onStop() might or might not be called.
Both changes can go live now (multiwindow update applies to DP1 & 2,
and onStop() clarification applies to all versions of API), so I'll
submit as soon as this is approved.
See first comment for doc stage location.
bug: 28580007
Change-Id: Ib008f24e5796ec7810b67c91e512e679680d4afd
When app has several windows and activity is relaunched + we try to preserve
main window - other windows just stayed around until removed by timeout or
replaced by app. There was a problem when one of the windows registered
broadcast receiver and set its own timer to remove it. In this case all
receivers were removed by framework because windows were considered leaked
and apps' timer caused crash when trying to remove registered receiver.
This CL removes all windows expect the main one, which we're trying to
preserve in this case.
Bug: 28337135
Change-Id: Ib8790cc8c61801f11d871ba3803bb0ebc3d3be01
ThreadedRenderer was never the right place for this anyway, and
ApplicationLoaders can provide both the full library search path (not
just the extracted native library dir) as well as the application loader
namespace.
Bug: 28213888
Change-Id: Ibcc0a9178da4dba6f3f3105932fdac1a1d0261af
Framework edition
Fix a bug where child FragmentManagers moving too lazily into the
CREATED state and beyond caused child fragments to not be
attached/created when expected.
Bug 25019275
Change-Id: I04ff0d3bcb693178a6ee3057da591392defdbcf8
When installing an APK that supports multiple ABIs, the ABI installed
can be forced to the secondary ABI [i.e. On devices that support both
32 and 64 bit variants, the 32-bit version can be forced when it's
the secondary ABI.] In this case, instrumenting the class always tried
to use the primary ABI. Instead of blindly using the primary ABI and
dropping the secondary ABI, we propagate both ABIs and make a
decision on which one should be chosen.
Bug: 28406240
Change-Id: I7ebb2fd264d2281912afd30f6d73ccb460f9cf85
- This should decrease the time it takes to fetch the recent task list,
and also reduce the number of tasks we store on-disk.
- Since a user can still open a fair number of tasks within the
time-window, and this list includes tasks that are excluded from the
UI, we are just reducing it to about half for the time being.
Bug: 28318359
Change-Id: Ie21b3d3d09a567007015fbbecaf914123c71cc06
Callbacks
- DataUsageCallback renamed to UsageCallback
- DataUsagePolicy removed; passing in params directly to register method
- making it an abstract class
- passing in (networkType, subscriberId) that reached its threshold
- renaming onLimitReached to onThresholdReached to match existing naming
- only monitor single network,subscriberId
- no monitoring of specific uids; using device or user wide instead
Tags
- only owner uid can read its tags
- exposing only TAG_NONE to match service side
BUG: 27530098
Change-Id: I2b2664da71806868a1e937d2bf4d1f234637509b
When multiple activities within the same process
try to handle requests for AssistStructure, the
singleton mLastAssistStructure tends to trash
the old structure when a second window's request
comes in.
This change passes in a sessionId so that the
cache is only cleared if the session id changes.
Bug: 28348867
Change-Id: I07efcd933db7e48aefd25a1c95493b71bbcffe4b
The white list of permitted directories is applied for every
classloader namespace by libnativeloader. LoadedApk no longer
needs to do that.
Add mDataDir to permitted paths in the case if for any reason
it is outside of the whitelisted dirs.
Bug: http://b/28639227
Change-Id: I5b61ac64e281ecee3e3e9b1fe56cf04dc6ee7b25
Move the logic for dispatching multi-window mode change events to
fragments and the activity window out of the public API onFoo methods
meant to be overridden by apps. This prevents problems if an extending
class forgets to call super.onFoo in these methods and removes the
need for doing so.
Bug 28667205
Change-Id: Ibf4c543e22b95753e1e4812bbd3e81f7648de54d
It's a protected broadcast, so sending it directly from 'am' is
no longer an option. This is needed for CTS as well as being
generally useful during app development.
Bug 28406044
Change-Id: I101915a8c6f19454330a8db2079a75023c112582
It is risky/complicated to try to put the activity showing on-top of
the lock screen in the right fullscreen configuration with the current
architecture. So, we just dismiss for now.
Bug: 28195260
Change-Id: I29a0350993ce8fe548d4a465b06d877cde151c78
The code assumed mapping from a permission to an app op
is one to one but this is not always the case. For example,
READ_SMS is mapped to OP_READ_SMS and OP_READ_ICC_SMS which
resulted mapping the READ_SMS permission to the OP_READ_ICC_SMS
instead of OP_READ_SMS resulting in a failure to find the op
name given the permission.
This breaks the AppOpsManager.permissionToOp() API for READ_SMS
returning null instead of OPST_READ_SMS. The consequence of this
is that the apps that proxy permission protected operations may
let the operations for READ_SMS go through as they would get a
null app op, i.e. no app op while there is one and it can be
disabled for the caller.
bug:28620132
Change-Id: I92f8ef375ae2122b7266c50653ce73f3d90f4b28
When relaunching an Activity, the DecorView and ViewRootImpl is re-used,
along with the IBinder activity token. This means that when we create
new resources, we end up just updating the base Activity resources.
However, we would set the configuration to the new requested override
config and then proceed to update existing Resources references, which
would return immediately due to the equality check for configuration
updates.
This change pushes the setting of the new override config to the
updateResourcesForActivity method, which will properly update Resources
references that are held in the re-used DecorView.
Bug:27915587
Change-Id: I113007a40fa464b8a234b073dcf851e16fee0177
You can't back up content in e.g. getNoBackupFilesDir() even if your
app explicitly tries to.
Bug 28321431
Change-Id: Ifa2a4bc518de03aba4c8809e60d8bb90fce767f8
