During package installation, we were parsing the APK twice; once
in the context of the PackageInstaller and once in the context
of the PackageManager. Instead, the installer should just pass
the certificates to be used further in the process.
If the PackageManager doesn't receive certificates [or, if there's
an error using them], it will fallback to re-parsing the APK.
Bug: 27502465
Change-Id: I94ce551af54eaa9916228e933134debe50867d21
On the lock screen the inline controls and gear behind the
notifications may be activated by accident. This CL alters them
to close after a certain period of time has occurred without
interaction.
Bug: 27335311
Change-Id: Ib3b0575571076a5912e23c7d2da5209e9a523fc8
I had a multiplier for this to make it easier, but it wasn't enabled.
This CL enables it and also fixes a bug I noticed where this doesn't
account for the distance moved being on the appropriate side for displaying
gear.
Change-Id: I7a516a425a556d7f5aa0627de5c64bb9211ac81e
MANAGE_USERS permission is not required if calling userId is the same as
requested user id. Theoretically this allows any 3P app to read UserInfo
state including PII fields like name and icon. The change clears PII fields
if the caller doesn't have MANAGE_USERS permission.
Bug: 27705805
Change-Id: Ic69c8cc6aafb7ac72b4fc2b9691cb8e4bef3fb2c
This forces everyone to go through sdcardfs, instead of letting them
around the back door.
Bug: 27992761
Change-Id: If1d4d5daa4b32bb3be7de10102bab24d63552b65
This adds a flag to WearableExtender (and Action.WearableExtender)
indicating that the attached ContentIntent will launch an activity
on Wear, which lets the platform know that it can (and should)
generate a sleek transition for the launching of said intent (as
opposed to the intent causing change in the background).
Bug:28019194
Change-Id: Id8b422ea52c5e941e38558aaf31ce2cf4bb7c6f3
Shouldn't animate the whole view, use the translatable views so that
they get clipped on tablet / landscape.
Additionally this changes the behavior in the case where the gear
might be showing, previously that would hide prior to the dismiss
all animation, UX prefers this to just have the notification cover
the gear and slide away or the gear fade away depending on the side
it's showing.
Bug: 28006639
Change-Id: I43027cb29410496de5b6e09c4a0c57a69484295c
Bug: 27821976
AMD held wakelock only during accelerometer data
acquisition, which allowed sleep to occur between its two
orientation measurements (http://go/anymotiondetector).
We now hold wakelock for its entire measurement, from
checkForAnyMotion() until it returns either STATIONARY or MOVING
or stop() is called.
Change-Id: Ib9c90de2c581d92f25212b91b9cc5e23642b5c77
Before this change, a PRE_BOOT_COMPLETED receiver could try calling
back into the system to ask for AccountManager details, only to be
told the user wasn't unlocked yet. If the broadcast code was probing
in a loop, it would force us to wait around for the 60-second ANR
timeout. Since typical devices have ~7 PRE_BOOT receivers, this
would delay BOOT_COMPLETED and other operations, like mounting the
SD card, for up to 7 minutes.
Bug: 28024024
Change-Id: Ibf8611e6fe94b0deb5ae5715c86f897ff6779088
Move flag to inner class to isolate Animation. This allows compile-
time initialization of Animation and subclasses.
Bug: 27265238
Change-Id: I3d95ab77e87de0c033f97545609461bdc9bd5d61
Removed the percentage shown in the header and
migrated the name to the subtext as contentinfo
was deprecated.
Change-Id: Ifd79a67cad8958049bd29b8eb4c9bcbb4822688b
- In PWM, make sure to read the height values after the new
configuration has been applied.
- Reset all navigation bar button icons when density changes.
- Adjust height of notification bar.
- Reload divider height values in SysUI and WM.
- Snap divider handle to a new position after loading the
new configuration, as the snap points change.
Bug: 26844819
Bug: 27450471
Bug: 27921696
Change-Id: I9e28f0c49f6367c5fcfac010e7a6e98a42e85996
When calculating the frame for non-fullscreen windows, we incorrectly
used the wrong bounds to calculate the frame, which lead to wrong
positioning.
To fix this, we use the inset bounds, which we consider the source
of truth for all layout related aspects, to calculate the frame,
and then offset everything by the difference between the inset bounds
and the task bounds to position them correctly.
Bug: 28012565
Bug: 27860956
Bug: 27441808
Change-Id: I90d45054e0bcce78d021ad2cd20e5ef7f79ded3d
