1. Retry validation 3 times instead of 10.
Reducing the number of retries speeds up dead network detection
and saves battery on networks that are don't have Internet access
(or block our connectivity checks).
Retrying 10 times is overkill. Each attempt sends 5-10 DNS
requests, waits 10 seconds to transfer only 3 or 4 packets, and
can last up to tens of seconds. Also, in MR1 we now retry every
10 minutes.
2. Make periodic revalidation only try once.
When validating periodically to see if a non-working connection
is working again, only make one attempt every 10 minutes, not
three.
3. Rename "retries" to "attempts" to make things a bit clearer.
Bug: 18922569
Bug: 18869302
Change-Id: I5f70b420c4f1bac0c33bb9ab6afb322902cdac4b
If there is a window with the accessibility focus we want to click
on the accessibility focused view in this window. The logic to
compute the bounds of the window was using the wrong window id,
hence getting an incorrect result. As a consequence in some cases
the user could not click on accessiiblity focused controls in the
nav bar.
bug:18889611
Change-Id: I89aee3ae2ffe27fe29819049c287a7155154c65b
If createUser returns null, we run into a NullPointerException
in createAndInitializeUser. This can happen when the policy
DISALLOW_ADD_USER is set.
Bug: 19121141
Change-Id: Id4bda832a243fa42c31eb5a176ecaa248aee68f5
When the restriction 'DISALLOW_SHARE_LOCATION' is applied or removed on a user
it is important to send out notifications to content observers of
LOCATION_PROVIDERS_ALLOWED.
Bug: 18995405
Change-Id: I8b3910a423a012ea9d15470eec101723c9f0eaf7
Previously the time gap between the first and the second key event
for repeated keys(press-and-hold) can be bigger than the threshold,
which violates CEC spec. This CL rectifies it by managing its own
state rather than relying on Android key handler.
Bug: 19117830
Change-Id: Iedfa48f9ab826252e8616d1c3b7491e8b8333c81
Currently we risk getting a SecurityException in a number of places,
where getUserData is called for a different user than the calling user.
To avoid this, the caller is cleared in a helper function.
Bug: 18662452
Change-Id: Ibc131c602e52d9f013fe739a9c18e693181ded67
This CL introduces a logic that, upon receiving <Inactive Source>
from the active source or a corresponding MHL subcommand, lets
the service return to one of internal inputs.
Introduced to handle it is a new type for HdmiDevice (INACTIVE)
that will be passed to input change listeners. The callback
is expected to transform to other mechanism such as intent,
to let TV app to decide which input to switch to, which will
be one of non-HDMI input that was viewed previously.
Bug: 19008579
Change-Id: I1922f4cd20e9220411061bb9d9fbe5fbc5676d48
Saves users from having to switch TV input manually to a playback device
after it boots up.
Bug: 19046176
Change-Id: I0a4fae1036138ce2e19cbac9fdd9dda41bdcf89f
Symptom:
In some scenario, there might have two home tasks
in the home stack. If the top home task was finishing
and expected to return to home stack, no activity
will resume.
Root Cause:
Unable to perform resumeHomeStackTask() because the
resumeTopActivityLocked() has been prevented to invoke
recursively.
Solution:
Just resume the next activity if already in home stack
Change-Id: I24ecbcac6cee4ddbd90cdd27c4f4915cffb9b354
Reason: Executing service com.google.android.syncadapters.contacts
/.SyncHighResPhotoIntentService
Make the code more robust when destroying services, so that if
the nesting count gets out of sync we don't just hang.
Change-Id: If117d5ef242e7c148fd9576bd89a1a092583d6ad
Makes sure the previous HDMI input gets switched to upon receiving
<Inactive Source> from the current active source device.
Bug: 18986597
Change-Id: I7120e81d75b5fd4cafd4b483e9b232b96e50f597
It is possible for empty activities to be leaked if their last
window is removed before animations finish. This change keeps tasks
and activities from having removal deferred if they have no windows.
Fixes bug 19047432
Change-Id: If6562f46bbfcac9ba987e2c834a1d55a9a8f3766
The uid on the task we are restoring from another device are
meaningless on the current device. Change updates the uid of any
task we restore to match the package uid on the current device.
Bug: 19014377
Change-Id: I285fbe9f509cae3d2e9a212a59e88f34b4db20e5
1) Inform MHL HAL of OPTION_MHL_SERVICE_CONTROL at system
ready/standby event.
2) Call OPTION_CEC_SET_LANGUAGE with current system language info
for HAL to be able to respond to <Get Menu Language> while in standby
mode.
Bug: 19054079
Change-Id: I92c9f392a94f70e11157d20a33163ef2d8bc832f
HDMI Spec says TV shall try to allocate TV() logical address, unlike
other kinds of devices that are recommended to use the previously
allocated one.
Also included are small changes:
- reply with 'can not provide source' if TV has no source to provide
- return the present device info if we have one at the given path
Bug: 19045081
Change-Id: Icf17cb581f0ef2274e252ec5665f80bb1c08b494
Don't say we're disconnected from a legacy type until there are no outstanding requests for it.
bug:18946574
Change-Id: I8e45c4a7558f7ced0840b71c50081989ba13c1c7
1. Remove debug for bug 17721767.
2. When adding activity windows based on base layer skip over
InputMethod windows.
Change-Id: I3600e9acd997002ce87d34df2af577cc7b648480
If a volume change occurred because it was sent through the MediaController directly
it wasn't showing the system's volume UI. This fixes things so the volume slider gets
shown for all volume requests with the SHOW_UI flag set.
bug:19036130
Change-Id: I1f7cb2e03386e2f3f7bc448cb4c38a71573775ff
