| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
Eventually we may want to use uint64_t, but will need to confirm atomicity.
Bug: 12381724
Change-Id: Ia2c591d262d22b47b6f7dab4b9d9faa14b86d865
|
|
|
|
|
|
| |
in IAudioFlinger::createTrack and IAudioFlinger::openRecord
Change-Id: I09c644c80e92c8e744b1b99055988a2588b2a83d
|
|\
| |
| |
| |
| | |
* commit '77177fa20773d02b4f9c4147ecb98107f019fa7d':
Allow releaseBuffer after flush
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
After AudioTrack start checks for pending flush,
allow releaseBuffer on any previously obtained buffer.
For example, this can happen if the resampler has obtained
a buffer but not released the whole buffer yet.
Note that the resampler will be reading obsolete data.
Bug: 11285590
Change-Id: I0614fbb62e43604aac3089cce4b7797c87a306b5
|
|\ \
| |/
| |
| |
| |
| |
| | |
restore" into klp-dev
* commit '7f8c397378a7ee5abd395413be71388ad36d3ed2':
AudioTrack: fix head position after restore
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The head position transfered to the new track
by restoreTrack_l() must take into account the frames that
are dropped from the old track to avoid a non recoverable
offset in the playback head position returned to applications.
Bug: 11230062.
Change-Id: I51143a08b95e8f264ed709ae2054360315f2b8b1
|
|\ \
| |/
| |
| |
| |
| |
| | |
SRC" into klp-dev
* commit '11454092e4a94d3c8b4576c981595339abdfac0d':
Fix underruns when fast track denied due to SRC
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
OpenSL ES requests a fast track. If sample rate conversion is needed,
the request is denied by server, and a larger client buffer is used
to handle the higher latency of a normal track. However the client
notification period was calculated based on buffer being divided into
2 sub-buffers. That resulted in the notification period being too long.
The server pulls chunks that are smaller than half the total buffer.
So now the client uses 3 sub-buffers when there is SRC.
Also removed the 'defer wake' optimization because it was incorrect.
This optimization attempted to reduce the number of wakeups of client,
when server releaseBuffer knows that another releaseBuffer will be
following. But there is no way for the first releaseBuffer to predict
how soon the second releaseBuffer will occur. In some cases it was
a long time, and the client underran. So now the client is woken up
immediately if the total number of available frames to client is >=
the minimum number the client wants to see (the notification period).
Also fix bug where minimum frame count was not being used in the
calculation of notification period.
Bug: 10342804
Change-Id: I3c246f4e7bc3684a344f2cf08268dc082e338e2a
|
|\ \
| |/
| |
| |
| | |
* commit '908d3c09ca7f2ccb280aa5dc8d876099ff9a9d0f':
Implement Track::getTimestamp()
|
| |
| |
| |
| |
| |
| |
| | |
using a new timestamp latch in PlaybackThread, and
AudioTrackServerProxy::framesReleased() which returns mServer.
Change-Id: I1ebfba968c773faaab95648c272fd3ebd74718d6
|
|/
|
|
| |
Change-Id: Id3ccc183a03421330d0498faaa62a45915cdc3d6
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This fixes a regression that was introduced earlier
by commit 9f80dd223d83d9bb9077fb6baee056cee4eaf7e5
called "New control block for AudioTrack and AudioRecord".
That commit broke underrun reporting for fast tracks.
Also remove Track::mUnderrunCount, which counted the number of underrun
events, and was only used by dumpsys media.audio_flinger.
Now dumpsys media.audio_flinger reports the number of underrun frames,
Isolated underrun-related control block accesses via the proxy, so that
the server is not directly poking around in the control block.
The new proxy APIs are AudioTrackServerProxy::getUnderrunFrames() and
AudioTrackServerProxy::tallyUnderrunFrames(). getUnderrunFrames() returns
a rolling counter for streaming tracks, or zero for static buffer tracks
which never underrun, but do a kind of 'pause' at end of buffer.
tallyUnderrunFrames() increments the counter by a specified number of frames.
Change-Id: Ib31fd73eb17cbb23888ce3af8ff29f471f5bd5a2
|
|
|
|
|
|
|
| |
This is part of a series of CLs to clean up the shared memory
control block, by removing any fields that don't have to be there.
Change-Id: I6e51003a1293b6800258c31b22cff2eba42162e7
|
|
|
|
| |
Change-Id: I7b6d31e24531954ab1ecdf3ed56c19433700bd89
|
|
|
|
| |
Change-Id: Ieabd91acee92d0e84e66fbd358df5282b856306e
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- start() returns a status so that upper layers can
recreate a non offloaded track in case of error.
- Added states to handle offloaded tracks specific:
- waiting for stream end (drain) notification by
audio flinger
- allow pause while waiting for stream end notification
- getPosition() queries the render position directly from
audio HAL.
- disable APIs not applicable to offloaded tracks
- Modified track restoring behavior for invalidated
offloaded tracks: just send the callback and wait for
upper layers to create a new track.
- Added wait for stream end management in audio track client
proxy. Similar to obtainBuffer and should be factored in.
Change-Id: I0fc48117946364cb255afd653195498891f622bd
Signed-off-by: Eric Laurent <elaurent@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Added specialized playback thread class for offload playback,
derived from directoutput thread.
This thread type handles specific state transitions for offloaded
tracks and offloading commands (pause/resume/drain/flush..) to audio HAL.
As opposed to other threads, does not go to standby if the track is paused.
- Added support for asynchronous write and drain operations at audio HAL.
Use a thread to handle async callback events from HAL: this avoids locking
playback thread mutex when executing the callback and cause deadlocks when
calling audio HAL functions with the playback thread mutex locked.
- Better accouting for track activity: call start/stop and release Output
methods in audio policy manager when tracks are actually added and removed
from the active tracks list.
Added a command thread in audio policy service to handle stop/release commands
asynchronously and avoid deadlocks with playback thread.
- Track terminated status is not a state anymore. This condition is othogonal
to state to permitted state transitions while terminated.
Change-Id: Id157f4b3277620568d8eace7535d9186602564de
|
|
|
|
|
|
|
|
| |
Maintain unreleased frame count on client side also (was already there on server side).
Assertion failure instead of BAD_VALUE status for incorrect usage of APIs.
Clean up error handling code.
Change-Id: I23ca2f6f8a7c18645309ee5d64fbc844429bcba8
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
NOTE: this does _not_ include all private member variables added
to classes as part of offload support. Only public/protected functions
and stubs functions/variables needed to make the changes buildable.
- isOffloadSupported() added to audio policy service
A stub implementation is required to build, this always returns false
- setParameters() added to IAudioTrack
A stub implementation is required to build, this always returns
INVALID_OPERATION
- CBlk flag for stream end
- Change AudioSystem::getRenderPosition() to take an audio_output_t
so caller can specify which output to query
- Add AudioSystem::isOffloadSupported()
This is fully implemented down to the AudioFlinger function
AudioPolicyServer::isOffloadSupported() which is just a stub
that always returns false.
- Add EVENT_STREAM_END to AudioTrack interface.
STREAM_END is used to signal when the hardware has actually finished
playing all the data it was sent.
- Add event type enumeration to media player interface AudioSink callbacks
so that the same callback can be used to handle multiple types of
event. For offloaded tracks we also have to handle STREAM_END and
TEAR_DOWN events
- Pass audio_offload_info_t to various functions used for opening outputs,
tracks and audio players. This passes additional information about the
compressed stream down to the HAL when using offload.
For publicly-available APIs this is an optional parameter (for some of
the internal and low-level APIs around the HAL interface it is mandatory)
- Add getParameters() and setParameters() API to AudioTrack
Currently dummy implementations.
- Change AudioPlayer contructor so that it takes a set of bitflags defining what
options are required. This replaces the original bool which only specified
whether to use deep buffering.
- Changes to StageFright class definition related to handling tearing-down of
an offloaded track when we need to switch back to software decode
- Define new StageFright utility functions used for offloaded tracks
Currently dummy implementations.
- AudioFlinger changes to use extended audio_config_t.
Fills in audio_offload_info_t member if this info is passed in when
opening an output.
- libvideoeditor changes required to add the new event type parameter
to AudioSink callback functions
- libmediaplayerservice changes required to add the new event type parameter
to AudioSink callback functions
Change-Id: I3ab41138aa1083d81fe83b886a9b1021ec7320f1
Signed-off-by: Richard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: Eric Laurent <elaurent@google.com>
|
|
|
|
|
|
|
|
| |
An event flag can be more fault-tolerant in case of loss of synchronization,
as it cannot overflow. It also allows more bits to be used in the future.
See http://en.wikipedia.org/wiki/Event_flag
Change-Id: I01ca25d951eb263124da54bb4738f0d94ec4a48b
|
|
|
|
|
|
|
|
| |
Main differences between old and new control block:
- removes the mutex, which was a potential source of priority inversion
- circular indices into shared buffer, which is now always a power-of-2 size
Change-Id: I4e9b7fa99858b488ac98a441fa70e31dbba1b865
|
|
|
|
| |
Change-Id: If7e2bc9b2a216524ee9cbb68682e2634933b4973
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The proxy object will eventually be the only code that understands the
details of the control block. This should make it easier to change the
control block in the future.
Initial set of control block fields that are isolated:
- sample rate
- send level
- volume
Prepare for streaming/static separation by adding a union to the control
block for the new fields.
Fix bug in handling of max sample rate on a track. It was only checking
at re-configuration, not at each mix.
Simplify OutputTrack::obtainBuffer.
Change-Id: I2249f9d04f73a911a922ad1d7f6197292c74cd92
|
|\
| |
| |
| |
| |
| |
| | |
VideoFrame"
* commit 'ba6b1bc38e0c355277f69af286469adb5f02e876':
Initialize and copy mRotationAngle in VideoFrame
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
When thumbnails were generated they could be generated at random
angles as the mRotationAngle variable was not initialized to any
value. This variable would have to be explicitly overwritten to not
cause random rotation. Changed the implementation to initialize the
value to 0 (no rotation). mRotationAngle was also missing in the
copy constructor.
Change-Id: I67a5340fdd807c6ab3a3da5eecb09b5b9d5f4666
|
| |
| |
| |
| |
| |
| | |
This is part of a series to clean up the control block.
Change-Id: I7f4cb05aef63053f8e2ab05b286d302260ef4758
|
| |
| |
| |
| | |
Change-Id: I85d7d2f6381b251db5695202fec75128883a8662
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Finish removing CBLK_RESTORING and CBLK_RESTORED from control block flags,
and remove constant RESTORE_TIMEOUT_MS.
Also minor cleanup:
- Cache mCblk in local variable cblk and make cblk allocatable in a register.
- Use "iMem" for sp<IMemory>.
- Add missing error log to AudioRecord; it was already in AudioTrack.
This is part of a series to clean up the control block.
Change-Id: Ia5f5ab4763c392bc06a45851b167ddaee29e3455
|
| |
| |
| |
| |
| |
| | |
This is part of a series to clean up the control block.
Change-Id: Ifab1c42ac0f8be704e571b292713cd2250d12a3f
|
| |
| |
| |
| |
| |
| | |
This is part of a series to clean up the control block.
Change-Id: Ie474557db7cb360f2d9a0f11600a68f5a3d46f07
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Remove CBLK_RESTORING and CBLK_RESTORED from control block flags,
for AudioTrack only. They are still used by AudioRecord.
This is part of a series to clean up the control block.
Change-Id: Iae4798f5b527c492bdaf789987ff3a1dadd0cb37
|
| |
| |
| |
| |
| |
| | |
This is part of a series to clean up the control block.
Change-Id: I0265fece3247356b585d4d48fbda6f37aea8a851
|
| |
| |
| |
| |
| |
| | |
This is part of a series to clean up the control block.
Change-Id: Ic881a3560d9547cb63fcc0cefec87aa3da480e0d
|
| |
| |
| |
| |
| |
| | |
Use only one symbol per flag
Change-Id: Ia3582e2134abd60c896d11337face65383e79c7c
|
|/
|
|
| |
Change-Id: Ib28fd7b9ce951a6933f006e7f8812ba617625530
|
|
|
|
|
|
|
| |
This should help diagnose problems by allowing us to correlate
the logs with the dumpsys media.audio_flinger output.
Change-Id: I8c7c592b4f87d13b0f29c66ce7a2f301a0f063c9
|
|
|
|
| |
Change-Id: Ifd825590ba36996064a458f64453a94b84722cb0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fix indentation to be multiple of 4.
Make it easier to search:
sp< not sp < to
"switch (...)" instead of "switch(...)" (also "if" and "while")
Remove redundant blank line at start or EOF.
Remove whitespace at end of line.
Remove extra blank lines where they don't add value.
Use git diff -b or -w to verify.
Change-Id: I966b7ba852faa5474be6907fb212f5e267c2874e
|
|
|
|
|
|
| |
We no longer put the filename at start of file.
Change-Id: Ic435b159a23105681e3d4a6cb1ac097bc853302e
|
|
|
|
|
|
|
|
|
| |
Fortunately audio_track_cblk_t doesn't have a destructor, but for clarity
remove the double destruction.
Also add warning not to add any virtuals to audio_track_cblk_t.
Change-Id: I70ebe1a70460c7002145b2cdf10f9f137396e6f3
|
|
|
|
|
|
|
| |
This is part of the process of abstracting the control block
to make it easier to maintain.
Change-Id: Idb8f461e68dab3bcf268159cc0781651c6fb7094
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Always read and write track volumes atomically. In most places this was
already being done, but there were a couple places where the left and
right channels were read independently.
Changed constant MAX_GAIN_INT to be a uint32_t instead of a float.
It is always used as a uint32_t in comparisons and assignments.
Use MAX_GAIN_INT in more places.
Now that volume is always accessed atomically, removed the union
and alias for uint16_t volume[2], and kept only volumeLR.
Removed volatile as it's meaningless.
In AudioFlinger, clamp the track volumes read from shared memory
before applying master and stream volume.
Change-Id: If65e2b27e5bc3db5bf75540479843041b58433f0
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
except in the control block, where we don't have room.
In AudioFlinger::ThreadBase::TrackBase::getBuffer,
read the frame size from control block only once.
Change-Id: Id6c4bccd4ed3e07d91df6bbea43bae45524f9f4e
|
|/
|
|
|
|
|
|
|
|
|
| |
Add an API to control block for getting/setting send level.
This allow us to make the mSendLevel field private.
Document the lack of barriers.
Use 0.0f to initialize floating-point values (for doc only).
Change-Id: I59f83b00adeb89eeee227e7648625d9a835be7a4
|
|
|
|
| |
Change-Id: I84906ebb9dfcfa5b96b287d18364b407f02a30c1
|
|
|
|
| |
Change-Id: I942d43973c20a7ace8b0d3f78b4da97e45e996c6
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Record and playback objects (resp AudioRecord and AudioTrack)
are created using a channel mask, but this information is lost
in the mixer because only the channel count is known to
AudioFlinger. A channel count can always be derived from a
channel mask.
The change consists in:
- disambiguiting variable names for channel masks and counts
- passing the mask information from the client to AudioFlinger
and the mixer.
- when using the DIRECT ouput, only verifying the format of
the track is compatible with the output's for PCM.
Change-Id: I50d87bfb7d7afcabdf5f12d4ab75ef3a54132c0e
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The first fix (commit 913af0b4) is problematic because it makes threads
in mediaserver process block on the cblk mutex. This is not permitted
as it can cause audio to skip or worse have a malicious application
prevent all audio playback by keeping the mutex locked.
The fix consists in using atomic operations when modifying the control
block flags.
Also fixed audio_track_cblk_t::framesReady() so that it doesn't block
when called from AudioFlinger (only applies when a loop is active).
Change-Id: Ibf0abb562ced3e9f64118afdd5036854bb959428
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This change fixes the stability problems experienced when using
a bluetooth headset supporting both A2DP and SCO. Problems occur
when starting the video chat at which time the A2DP output is being
stopped to start SCO. At that time, active AudioTracks are invalidated
by AudioFlinger so that a new AudioTrack binder interface can be
recreated by the client process on the new mixer thread with correct parameters.
The problem was that the process to restore the binder interface was not
protected against concurrent requests which caused 2 binder interfaces
to be created sometimes. This could lead to permanent client deadlock
if one of the client threads was waiting for a condition of the first
created binder interface while the second one was created (as the AudioFlinger
would only signal conditions on the last one created).
This concurrent request situation is more likely to happen when a client
uses the JAVA AudioTrack as the JNI implementation uses simultaneously the
native AudioTrack callback and write push mechanisms. By doing so, the code
that checks if the binder interface should be restored (in obtainBuffer()) is
much more likely to be called concurrently from two different threads.
The fix consists in protecting the critical binder interface restore phase
with a flag in the AudioTrack control block. The first thread acting upon the binder
interface restore request will raise the flag and the second thread will just wait for
a condition to be signaled when the restore process is complete.
Also protected all accesses to the AudioTrack control block by a mutex to prevent
access while the track is being destroyed and restored. If a mutex cannot be held
(e.g because we call a callback function), acquire a strong reference on the IAudioTrack
to prevent its destruction while the cblk is being accessed.
Modified AudioTrack JNI to use GetByteArrayElements() instead of
GetPrimitiveArrayCritical() when writing audio buffers. Entering a critical section would
cause the JNI to abort if a mediaserver crash occurs during a write due to the AudioSystem
callback being called during the critical section when media server process restarts.
Anyway with current JNI implementation, either versions do not copy data most of the times
and the criticial version does not guaranty no data copy.
The same modifications have been made to AudioRecord.
Change-Id: Idc5aa711a04c3eee180cdd03f44fe17f3c4dcb52
|