There were 2 problems in SoundPool:
1 If not using the shared memory buffer mode, there was a problem when a sound channel was stolen. The new channel could sometimes not be added to the restart
list if the AudioTrack callback thread was stopped before the underrun callback was called.
The SoundChannel::play() method is modified so that SoundPool::done() is called immediately after stopping the channel. There is a possibility that done() is called a second time by the callback; in this case it will be added 2 times to the restart list but the second start request will be ignored as the first one will have reset the next chennel ID when processed.
2 There was a deadlock on SoundPool::mLock if SoundPool::stop() was called while a channel restart was pending:
SoundPool::stop() lock mLock -> SoundChannel::stop() -> SoundPool::done() -> SoundPool::addToRestartList() -> try to lock mLock == deadlock
A second mutex mRestartLock is added to protect the restart list mRestart. mLock is still used to protect mChannels list but mRestart is now used to
protect access to mRestart by restart thread and client thread.
The cause is very likely that the WaveGenerator *lpWaveGen returned by lpToneGen->mWaveGens.valueFor(lFrequency) just before calling lpWaveGen->getSamples(lpOut, lGenSmp, lWaveCmd) is invalid. The frequency lFrequency is not part of the frequencies in mWaveGens.
This can happen if a different tone is started while the callback function is active: The state is changed to TONE_RESTARTING and the call to prepareWave() at line 1226 will change the tone descriptor pointed to by mpToneDesc as well as the content of mWaveGens. However, mpToneDesc was cached in a local variable lpToneDesc when entering the callback and is not reloaded when exiting prepareWave(). This causes a mismatch between the tone frequencies listed in lpToneDesc and the frequencies present in mWaveGens.
This regression was introduced in change 973 when mpToneDesc was cached in a local variable.
* changes:
Move ExifInterface to android.media package so we can reference it from MediaScanner. Also hide public constructor and wrap common use cases as atomic operation to avoid race condition in jhead native codes.
Store sample rate on 32 bits instead of 16 bits in audio_track_cblk_t.
Removed sampleRate() methods from AudioTrack and AudioRecord: replaced by getSampleRate().
AudioTrack::setSampleRate() no returns a status.
This allows the camera service to hang onto the callback interface
until all callbacks have been processed. This prevents problems
where pending callbacks in binder worker threads are processed
after the Java camera object and its associated native resources
have been released.
Bug 1884362
Do not disable a route unless routes is zero, to avoid accidentally disabling
if the same route is enabled twice in a row.
Signed-off-by: Mike Lockwood <lockwood@android.com>
The problem is that setRingerModeInt() does not handle streams not affected by ringer mode: when enabling "Alarm in silent mode" while in silent mode, setRingerModeInt is called after alarm streamis removed from ringer mode affected streams, and nothing is done.
The fix constists in also processing streams that are not affected by ringer mode and to restore last audible volume for them. It does not arm reapplying the volume for streams the are never affected by ringer mode as we don't do this very often.
The other problem noted in the bug report (ringer volume always restored to non zero value when exiting silent mode even if set to zero before) is also fixed: a new parameter is added to setIndex() and setStreamVolumeInt() to explicitely request to store the volume index as last audible instead of doing it automatically if index > 0.
* changes:
Update MediaPlayer to allow setVideoSurface calls after prepare. Also allow passing a null surface. The API is now enabled to change the surface while the video is playing. This could allow orientation changes during playback or to allow the audio track from a video to play in the background. NOTE: There are still changes required to pmem driver to allow remapping shared physical memory into a process in order for this to work. This change only enables the API to send the appropriate calls when the lower level code supports it.
passing a null surface. The API is now enabled to change the surface while
the video is playing. This could allow orientation changes during playback
or to allow the audio track from a video to play in the background.
NOTE: There are still changes required to pmem driver to allow remapping
shared physical memory into a process in order for this to work. This
change only enables the API to send the appropriate calls when the lower
level code supports it.
This change is the first part of a fix for issue 1846343, :
- Added new enum values for input sources in AudioRecord and MediaRecorder for voice uplink, downlink and uplink+downlink sources.
- renamed streamType to inputSource in all native functions handling audio record.
A second change is required in opencore author driver and android audio input to completely fix the issue.
implementation to the mediaserver process. The permission check was previously
enforced only at the AudioRecord binder interface for clients not in the same
process. This change adds an additional check when the client tries to set
the audio source.
Bug 1868334
