Stabilize mapping between ringer-mode=silent and zen=priority
by keeping track of two ringer modes:
- Internal ringer mode: Used for underlying stream muting
- External ringer mode: Reported to clients
The mapping between external ringer mode + zen is:
- normal = all
- vibrate = all
- silent = priority (read-write) or none (read)
Changes include:
- Remove "zen check" from audio service, back to audio
service having no knowledge of zen.
- Maintain a new external ringer mode in audio service,
this is the ringer mode reported through AudioManager
to callers, also mapped to the change intent.
- Introduce a "ringer mode delegate" to the local
audio manager interface, responsible for observing
external / internal mode changes, and making changes
if necessary.
- Internal ringer mode changes are still interesting
to the volume dialog, wire up a callback through
the existing IVolumeController interface.
- On devices without vibration, the mapping is the same
but since no ringer mode change is possible, disable
the icon toggle and remove the mute icon when volume=0.
- On devices with vibration, volume down presses should
pulse the vibrate icon (and vibrate) as a hint that this
is as low as the device can go using the keys. Since
the mechanics are similar to the existing zen=none hint,
pull into shared helper.
- Log ringer mode changes to the zen log, include calling
package information for issue diagnosis.
- Include whether vibration is supported in the audio service
dump.
- Update the status bar icon policy to use the internal ringer
mode, not the external mode (for vibrate icon).
- Update the "Muted by <x>" logic, include current suppressor
in dumpsys, ensure suppression icon is enabled & !clickable,
regardless of zen mode.
Bug: 17884168
Bug: 15471679
Bug: 16824970
Change-Id: Ia7d3bb23ce6d1e37b24fb6521d1c1ab9bb8f60c0
Allows ManagedProvisioning to determine whether there's a
challenge and thus whether to disable NFC provisioning.
Other implementation option: new hidden boolean API method.
Can't think of benefit of new API method "isBlockInUse", other
than doesn't leak PDB size and is more explicitly tied to the
use case. Open to either impl if anyone has opinions on the matter.
Bug: 18508767
Change-Id: I28d2eb5a0837ff85cb91f140b17ce1dd843fe9d6
The return value of jniRegisterNativeMethods is checked only in
LOG_FATAL_IF, which defines to nothing in the LOG_NDEBUG
case.
Fake a use of the 'res' variable to shut off warnings when LOG_NDEBUG.
Change-Id: I8263610f327c56897f76796fe1fbc2b325b0559f
Signed-off-by: Bernhard Rosenkränzer <Bernhard.Rosenkranzer@linaro.org>
Apps delivered as multiple split APKs must have identical package
names, version code, and signatures. However, developers may want
to iterate quickly on a subset of splits without having to increment
the version code, which would require delivery of the entire app.
This change introduces "revision codes" which can vary between
split APKs belonging to the same app. An install is valid as long
as the normal version code is identical across all splits. Splits
can be added/removed to an app over time, but if a split is present
across an upgrade the revision code must not decrease.
Since system apps could have been updated with splits, only revert
to the built-in APKs if the version code is strictly greater than the
data version. Also fix bug to enable inheriting from system apps
when adding splits.
Bug: 18481866
Change-Id: I34d8e14c141a8eb95c33ffe24b4e52d6af5c8260
Remove an unused static function clang complains about
Change-Id: Id5875960e9f93348c7a3a912f42b5f457af2cdc0
Signed-off-by: Bernhard Rosenkraenzer <Bernhard.Rosenkranzer@linaro.org>
