OSDN Git Service

android-x86/frameworks-av.git
12 years agoMerge "Use pid_t not int"
Glenn Kasten [Wed, 8 Feb 2012 16:40:28 +0000 (08:40 -0800)]
Merge "Use pid_t not int"

12 years agoMerge "Don't double destruct audio_track_cblk_t"
Glenn Kasten [Wed, 8 Feb 2012 16:40:15 +0000 (08:40 -0800)]
Merge "Don't double destruct audio_track_cblk_t"

12 years agoMerge "AudioFlinger methods const and inline"
Glenn Kasten [Wed, 8 Feb 2012 16:39:39 +0000 (08:39 -0800)]
Merge "AudioFlinger methods const and inline"

12 years agoMerge "Remove dead mutex in AudioTrack/AudioRecord thread"
Glenn Kasten [Wed, 8 Feb 2012 16:34:33 +0000 (08:34 -0800)]
Merge "Remove dead mutex in AudioTrack/AudioRecord thread"

12 years agoMerge "Use virtual destructors"
Glenn Kasten [Wed, 8 Feb 2012 16:33:41 +0000 (08:33 -0800)]
Merge "Use virtual destructors"

12 years agoMerge "Improve performance for sp<> on stack"
Glenn Kasten [Wed, 8 Feb 2012 15:45:15 +0000 (07:45 -0800)]
Merge "Improve performance for sp<> on stack"

12 years agoMerge "AudioTrack declare more methods const"
Glenn Kasten [Wed, 8 Feb 2012 15:42:40 +0000 (07:42 -0800)]
Merge "AudioTrack declare more methods const"

12 years agoMerge "Use 0 not NULL for sp<> and wp<>"
Glenn Kasten [Wed, 8 Feb 2012 15:41:44 +0000 (07:41 -0800)]
Merge "Use 0 not NULL for sp<> and wp<>"

12 years agoMerge "Use bool instead of int"
Glenn Kasten [Wed, 8 Feb 2012 15:40:23 +0000 (07:40 -0800)]
Merge "Use bool instead of int"

12 years agoMerge "Declare more IAudioFlinger methods const"
Glenn Kasten [Wed, 8 Feb 2012 15:39:27 +0000 (07:39 -0800)]
Merge "Declare more IAudioFlinger methods const"

12 years agoMerge "Remove dead code"
Glenn Kasten [Wed, 8 Feb 2012 15:38:42 +0000 (07:38 -0800)]
Merge "Remove dead code"

12 years agoMerge "MatroskaExtractor: to support MPEG4 and MP3 codec."
Andreas Huber [Tue, 7 Feb 2012 18:43:08 +0000 (10:43 -0800)]
Merge "MatroskaExtractor: to support MPEG4 and MP3 codec."

12 years agoDon't call virtual function in destructor of SurfaceMediaSource
James Dong [Sat, 4 Feb 2012 14:19:50 +0000 (06:19 -0800)]
Don't call virtual function in destructor of SurfaceMediaSource

Change-Id: I3cbc2b1222335b61c814b5cdcfaefa495148b0ec

12 years agoMerge "Don't call virtual functions in the destructor for audio and camera source...
James Dong [Sat, 4 Feb 2012 13:54:53 +0000 (05:54 -0800)]
Merge "Don't call virtual functions in the destructor for audio and camera source classes"

12 years agoDon't double destruct audio_track_cblk_t
Glenn Kasten [Fri, 3 Feb 2012 18:24:48 +0000 (10:24 -0800)]
Don't double destruct audio_track_cblk_t

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

12 years agoUse pid_t not int
Glenn Kasten [Fri, 3 Feb 2012 19:10:26 +0000 (11:10 -0800)]
Use pid_t not int

Change-Id: Iad1c2fd4152e94080ad8c65c13ddf4519fc2ed27

12 years agoRemove dead code
Glenn Kasten [Thu, 26 Jan 2012 17:50:01 +0000 (09:50 -0800)]
Remove dead code

mFormat is unused in resampler
mClientTid is unused
local variable pid is unused in dump

Change-Id: Ib156e38029366620bfeff2a13e73471867155a5b

12 years agoAudioTrack declare more methods const
Glenn Kasten [Wed, 4 Jan 2012 20:41:44 +0000 (12:41 -0800)]
AudioTrack declare more methods const

Change-Id: I4999e984460893961d0d8092cff17f3cf07d7214

12 years agoDeclare more IAudioFlinger methods const
Glenn Kasten [Fri, 27 Jan 2012 00:25:10 +0000 (16:25 -0800)]
Declare more IAudioFlinger methods const

This is just documentation, as C++ method const-ness doesn't mean anything
for a binder API.  Instead, here const means "no side effects".

Change-Id: Iaa9cd2fe477db10ae9a40cac4f79f0faa9b4e5e6

12 years agoMerge "Don't call virtual functions in destructors for the writer classes"
James Dong [Fri, 3 Feb 2012 23:37:40 +0000 (15:37 -0800)]
Merge "Don't call virtual functions in destructors for the writer classes"

12 years agoUse bool instead of int
Glenn Kasten [Fri, 27 Jan 2012 20:33:54 +0000 (12:33 -0800)]
Use bool instead of int

The .h is not modified to avoid increasing data size.

Change-Id: Ide4a821a5b16424ffa03471dfff98dc3e9b5f751

12 years agoAudioFlinger methods const and inline
Glenn Kasten [Thu, 2 Feb 2012 22:06:11 +0000 (14:06 -0800)]
AudioFlinger methods const and inline

This saves 1063 bytes and probably improves performance.

Change-Id: I11cf0dfd925fbaec75e3d1b806852a538eae5518

12 years agoUse virtual destructors
Glenn Kasten [Mon, 30 Jan 2012 22:54:39 +0000 (14:54 -0800)]
Use virtual destructors

It turns out to be just a comment, as all except AudioMixer are RefBase.

There are only a few performance-sensitive cases where it's worth thinking
about whether you need a virtual destructor, and the headache usually
outweighs the benefit.

Change-Id: I716292f9556ec17c29ce8c76ac8ae602cb496533

12 years agoMerge "Keep AACEncoder around for a bit longer to have more time fix issues related...
James Dong [Fri, 3 Feb 2012 19:29:59 +0000 (11:29 -0800)]
Merge "Keep AACEncoder around for a bit longer to have more time fix issues related to video editor engine"

12 years agoUse 0 not NULL for sp<> and wp<>
Glenn Kasten [Fri, 20 Jan 2012 21:44:40 +0000 (13:44 -0800)]
Use 0 not NULL for sp<> and wp<>

Change-Id: Id1f0c89acefaceed6cb9ca7c165fce895e46d85b

12 years agoKeep AACEncoder around for a bit longer to have more time fix issues related to video...
James Dong [Fri, 3 Feb 2012 19:03:56 +0000 (11:03 -0800)]
Keep AACEncoder around for a bit longer to have more time fix issues related to video editor engine

o this should be reverted after the problem is fixed.
o related-to-bug: 5947347

Change-Id: Iaec4b59d2c99c975e83f0588a813e9a4bfcb7ee2

12 years agoMerge "Use audio_in_acoustics_t consistently"
Glenn Kasten [Fri, 3 Feb 2012 16:31:01 +0000 (08:31 -0800)]
Merge "Use audio_in_acoustics_t consistently"

12 years agoMerge "Use ToneGenerator::tone_type consistently"
Glenn Kasten [Fri, 3 Feb 2012 16:00:52 +0000 (08:00 -0800)]
Merge "Use ToneGenerator::tone_type consistently"

12 years agoUse NULL not 0 for raw pointers
Glenn Kasten [Sat, 28 Jan 2012 00:47:15 +0000 (16:47 -0800)]
Use NULL not 0 for raw pointers

Use if (p != NULL) instead of if (ptr)

Change-Id: Iaec3413a59ccbf233c98fcd918cc7d70ac5da9fa

12 years agoMerge "For performance, return large objects by reference"
Glenn Kasten [Fri, 3 Feb 2012 15:46:42 +0000 (07:46 -0800)]
Merge "For performance, return large objects by reference"

12 years agoMerge "No need to check a wp<> for 0 before promote()"
Glenn Kasten [Fri, 3 Feb 2012 15:45:13 +0000 (07:45 -0800)]
Merge "No need to check a wp<> for 0 before promote()"

12 years agoConstructor initialization and const fields
Glenn Kasten [Wed, 25 Jan 2012 23:28:08 +0000 (15:28 -0800)]
Constructor initialization and const fields

In constructors, initialize member fields in the initialization list
rather than constructor body where possible.  This allows more fields
to be const, provided they are never modified.

Also initialize POD fields in constructor, unless it's obvious they
don't need to be initialized.  In that case, put a comment instead.

Remove explicit clear() in destructors on fields that are now const.

Give AudioSessionRef a default constructor, so it's immutable fields can
be marked const.

Add comment about ~TrackBase() trick.

Initialize fields in declaration order to make it easier to confirm that
all fields are set.

Move initialization of mHardwareStatus from onFirstRef() to constructor.

Use NULL not 0 to initialize raw pointers in initialization list.

Rename field mClient to mAudioFlingerClient, and getter from client()
to audioFlingerClient().

Change-Id: Ib36cf6ed32f3cd19003f40a5d84046eb4c122052

12 years agoMerge "Cleanup thread types"
Glenn Kasten [Fri, 3 Feb 2012 15:26:28 +0000 (07:26 -0800)]
Merge "Cleanup thread types"

12 years agoMerge "Make AudioTrack control block volume field private"
Glenn Kasten [Fri, 3 Feb 2012 15:18:06 +0000 (07:18 -0800)]
Merge "Make AudioTrack control block volume field private"

12 years agoDon't call virtual functions in the destructor for audio and camera source classes
James Dong [Fri, 3 Feb 2012 02:04:02 +0000 (18:04 -0800)]
Don't call virtual functions in the destructor for audio and camera source classes

Change-Id: Ia74ffc1c0cbd7971697f5e3c476e340ec5c7727a

12 years agoDon't call virtual functions in destructors for the writer classes
James Dong [Thu, 2 Feb 2012 23:07:52 +0000 (15:07 -0800)]
Don't call virtual functions in destructors for the writer classes

Have not found any concrete bugs related to these calls yet, but we should avoid
calling virtual functions in destructors, regardless.

Change-Id: I2d47b79d3fb2d29f418619bee83aa147d232a5d4

12 years agoMerge "Fix const sp<>& in parameter list and return value"
Glenn Kasten [Thu, 2 Feb 2012 21:31:23 +0000 (13:31 -0800)]
Merge "Fix const sp<>& in parameter list and return value"

12 years agoMerge "Removed a loop for buffer lookup"
James Dong [Thu, 2 Feb 2012 21:30:49 +0000 (13:30 -0800)]
Merge "Removed a loop for buffer lookup"

12 years agoMerge "Unconditional delete"
Glenn Kasten [Thu, 2 Feb 2012 21:30:48 +0000 (13:30 -0800)]
Merge "Unconditional delete"

12 years agoMerge "More audio_stream_type_t"
Glenn Kasten [Thu, 2 Feb 2012 21:26:12 +0000 (13:26 -0800)]
Merge "More audio_stream_type_t"

12 years agoRemove dead mutex in AudioTrack/AudioRecord thread
Glenn Kasten [Thu, 2 Feb 2012 18:52:34 +0000 (10:52 -0800)]
Remove dead mutex in AudioTrack/AudioRecord thread

The client callback threads had mutexes called AudioTrackThread::mLock
and ClientRecordThread::mLock.  These mutexes were only used by start()
and stop(), and were unused by the thread itself.  But start() and
stop() already have their own protection provided by AudioTrack::mLock
and AudioRecord::mLock.  So the thread mutexes can be removed.

Change-Id: I098406d381645d77fba06a15511e179a327848ef

12 years agoRemoved a loop for buffer lookup
James Dong [Thu, 2 Feb 2012 03:28:54 +0000 (19:28 -0800)]
Removed a loop for buffer lookup

o used the nInputPortIndex and nOutputPortIndex from header instead
  of assuming that the # of ports == 2

Change-Id: I4b615912b088b4e2bac9c00e89986e811a5c58bb

12 years agoMerge "modifications for new webrtc modules"
Eric Laurent [Thu, 2 Feb 2012 17:39:28 +0000 (09:39 -0800)]
Merge "modifications for new webrtc modules"

12 years agoMerge "Remove the restriction that the width must be a multiple of 4 or 2 for thumbna...
James Dong [Thu, 2 Feb 2012 02:53:50 +0000 (18:53 -0800)]
Merge "Remove the restriction that the width must be a multiple of 4 or 2 for thumbnail generation"

12 years agoRemove the restriction that the width must be a multiple of 4 or 2 for thumbnail...
James Dong [Wed, 1 Feb 2012 21:33:33 +0000 (13:33 -0800)]
Remove the restriction that the width must be a multiple of 4 or 2 for thumbnail generation

o simply applying the change in commit a161af9d1e2baa3f23e32634ef7dd9b4bbce9a6c to other formats
o related-to-bug: 5947196

Change-Id: I95de10263b8ad9ec941f1d6b907fca6aeff0a90f

12 years agoThe AMR WB software encoder is now an OMX component.
Andreas Huber [Wed, 1 Feb 2012 19:47:54 +0000 (11:47 -0800)]
The AMR WB software encoder is now an OMX component.

Change-Id: Iebfb2696783a0163369ba4e8ef72e19ee76ad893

12 years agomodifications for new webrtc modules
Eric Laurent [Tue, 31 Jan 2012 20:35:20 +0000 (12:35 -0800)]
modifications for new webrtc modules

Adapt include files path in makefile and source to new
directory structure in webrtc library.

Also changed the default AGC parameters to less extreme settings.

Change-Id: Ia6a41c816dc44645f8f867937b722eede88c06ea

12 years agoThe software AMR NB encoder is now an OMX component.
Andreas Huber [Tue, 31 Jan 2012 22:34:59 +0000 (14:34 -0800)]
The software AMR NB encoder is now an OMX component.

Change-Id: I890eab052a7c36409b8b694c964884e28dd8d8fc

12 years agoMerge "add tests module tag to omx_tests"
James Dong [Wed, 1 Feb 2012 01:36:37 +0000 (17:36 -0800)]
Merge "add tests module tag to omx_tests"

12 years agoadd tests module tag to omx_tests
James Dong [Tue, 31 Jan 2012 20:24:17 +0000 (12:24 -0800)]
add tests module tag to omx_tests

Change-Id: I5dcb5187f09a7ae980ac5857b7ade81001d459f8

12 years agoMerge "First step of refactoring 'timedtext' code."
James Dong [Tue, 31 Jan 2012 20:06:43 +0000 (12:06 -0800)]
Merge "First step of refactoring 'timedtext' code."

12 years agoSlighly improve the API to discover if the OMX stack is running in the local
Andreas Huber [Tue, 31 Jan 2012 19:16:24 +0000 (11:16 -0800)]
Slighly improve the API to discover if the OMX stack is running in the local

process.

Change-Id: Idd3c2f0d4a9542af01a3fdfd3b0bfab90e083505

12 years agoMerge "Make sure our aac decoder handles this parameter."
Andreas Huber [Tue, 31 Jan 2012 16:31:03 +0000 (08:31 -0800)]
Merge "Make sure our aac decoder handles this parameter."

12 years agoFirst step of refactoring 'timedtext' code.
Insun Kang [Wed, 11 Jan 2012 13:03:12 +0000 (22:03 +0900)]
First step of refactoring 'timedtext' code.

Goal
1. Removed dependency of TimedTextPlayer on AwsomePlayer.
2. Generalized TimedTextParser to TimedTextSource and its subclasses.

Summary
1. Introduced TimedTextDriver, TimedTextPlayer (new implementation),
TimedTextSource (and its subclasses).
2. Removed TimedTextParser.

Remaining TODOs
1. Revise VideoVidew, Gallery3D app, AwesomePlayer to check if
'pause' and 'resume' works well.
2. Consider revising MediaPlayer APIs such as setParameter() ->
addTextSource(). Need more thoughts.

Change-Id: Ie0c4f15b9690102de755cef6940f8c31ccf78e27

12 years agoMake sure our aac decoder handles this parameter.
Andreas Huber [Tue, 31 Jan 2012 00:45:28 +0000 (16:45 -0800)]
Make sure our aac decoder handles this parameter.

Change-Id: I212ef0c21776b99d0c6662b118ca0ec16b1d3d5e

12 years agoam 827a597d: Reconcile with ics-mr1-release
The Android Open Source Project [Mon, 30 Jan 2012 21:38:52 +0000 (13:38 -0800)]
am 827a597d: Reconcile with ics-mr1-release

* commit '827a597df70a51c6cccad52fae2be2c469f58b77':
  Revert "Fix 5797764: Increase lock screen's thread priority"
  DO NOT MERGE Revert "AudioFlinger: mix track only when really ready (2)"

12 years agoReconcile with ics-mr1-release
The Android Open Source Project [Mon, 30 Jan 2012 21:36:34 +0000 (13:36 -0800)]
Reconcile with ics-mr1-release

Change-Id: Ic6454a4b7929bbf2f453022fd8ae596e215657c2

12 years agoFor performance, return large objects by reference
Glenn Kasten [Mon, 30 Jan 2012 20:56:03 +0000 (12:56 -0800)]
For performance, return large objects by reference

Change-Id: Ibf737018ef1d3c7d717584615dcb2d4ecdb50c99

12 years agoImprove performance for sp<> on stack
Glenn Kasten [Mon, 30 Jan 2012 21:00:02 +0000 (13:00 -0800)]
Improve performance for sp<> on stack

Combine default constructor for sp<> immediately followed by assignment,
as the reference-counting is relatively slow.  Also return sp<> directly
rather than via local variable, for the same reason.

Change-Id: If55931f1e407994f6591ddde41b53db72fb4fc40

12 years agoInstantiate software codecs locally whenever possible.
Andreas Huber [Mon, 30 Jan 2012 19:48:02 +0000 (11:48 -0800)]
Instantiate software codecs locally whenever possible.

Change-Id: I58d5c30b9de9873a239e91955a8900e2c72518a7

12 years agoFix const sp<>& in parameter list and return value
Glenn Kasten [Mon, 30 Jan 2012 18:15:48 +0000 (10:15 -0800)]
Fix const sp<>& in parameter list and return value

EffectModule::addHandle and Client::heap() were declared incorrectly.

As a parameter, an sp<> should be & for efficiency, and for input
parameters it should also be const to protect the caller's value.

But as a return value, an sp<> should have neither const or &.  The "e"
in "return e;" might be located on the stack, and if there is "&" then
the caller would see the address of a variable which no longer exists.
Also, an & would make it hard to do "return 0;".
A "const" without & is meaningless in the return type.
(In this particular case, the "e" is a member field, so it was safe.)

Change-Id: I3df5f294214eb15a9d4d596c6d5ef29de97b5c27

12 years agoUnconditional delete
Glenn Kasten [Sat, 28 Jan 2012 02:08:45 +0000 (18:08 -0800)]
Unconditional delete

Don't check that pointer is non-NULL before delete.

Don't leave deleted member fields non-NULL, except in a destructor,
since it could be misleading in a dump or debugger. (mRsmpOutBuffer)

Change-Id: Ic0492a6b752f74a67f4c96dfb89ca2de4e69eecf

12 years agoMore audio_stream_type_t
Glenn Kasten [Fri, 20 Jan 2012 17:37:45 +0000 (09:37 -0800)]
More audio_stream_type_t

Change-Id: I1260259efe0aa3fc1ef13de69758aaa592e1f815

12 years agoNo need to check a wp<> for 0 before promote()
Glenn Kasten [Wed, 25 Jan 2012 22:27:41 +0000 (14:27 -0800)]
No need to check a wp<> for 0 before promote()

Also remove unnecessary wp<> local variable.

Change-Id: I620e67b5d559d28616f8e00609a525cfe19c5ddc

12 years agoUse ToneGenerator::tone_type consistently
Glenn Kasten [Fri, 27 Jan 2012 23:25:25 +0000 (15:25 -0800)]
Use ToneGenerator::tone_type consistently

Also remove defaults in startToneCommand(), they're not needed and the
default for tone type was nonsense.

Change-Id: I70fa8cee4f3dbb8c66ceb3719c8d3d2f447f05b9

12 years agoUse audio_in_acoustics_t consistently
Glenn Kasten [Fri, 27 Jan 2012 20:32:34 +0000 (12:32 -0800)]
Use audio_in_acoustics_t consistently

Change-Id: I0a9dd668fb2e57b1c3ece3190588194974b99062

12 years agoMerge "Use audio_source_t consistently"
Glenn Kasten [Fri, 27 Jan 2012 17:06:00 +0000 (09:06 -0800)]
Merge "Use audio_source_t consistently"

12 years agoMerge "AudioStreamIn and AudioStreamOut"
Glenn Kasten [Fri, 27 Jan 2012 15:22:30 +0000 (07:22 -0800)]
Merge "AudioStreamIn and AudioStreamOut"

12 years agoMerge "Declare methods in binder opcode order"
Glenn Kasten [Fri, 27 Jan 2012 15:19:53 +0000 (07:19 -0800)]
Merge "Declare methods in binder opcode order"

12 years agoMerge "Use enum effect_state consistently"
Glenn Kasten [Fri, 27 Jan 2012 15:19:10 +0000 (07:19 -0800)]
Merge "Use enum effect_state consistently"

12 years agoMerge "Use enum track_state consistently"
Glenn Kasten [Fri, 27 Jan 2012 15:18:23 +0000 (07:18 -0800)]
Merge "Use enum track_state consistently"

12 years agoMerge "Use enum mixer_state consistently"
Glenn Kasten [Fri, 27 Jan 2012 15:16:46 +0000 (07:16 -0800)]
Merge "Use enum mixer_state consistently"

12 years agoUse audio_source_t consistently
Glenn Kasten [Mon, 23 Jan 2012 21:58:49 +0000 (13:58 -0800)]
Use audio_source_t consistently

Was a mix of audio_source_t, uint8_t, and int.

Related fixes:
 - fix comments in MediaRecorder.java
 - AudioPolicyService server side was not checking source parameter at
   all, so if the client wrapper was bypassed, invalid values could be
   passed into audio HAL
 - JNI android_media_AudioRecord_setup was checking source for positive
   values, but not negative values. This test is redundant, since already
   checked at Java and now checked by AudioPolicyService also, but might
   as well make it correct.

Change-Id: Ie5e25d646dcd59a86d7985aa46cfcb4a1ba64a4a

12 years agoMerge "The software AAC encoder is now an OMX component."
Andreas Huber [Fri, 27 Jan 2012 00:00:38 +0000 (16:00 -0800)]
Merge "The software AAC encoder is now an OMX component."

12 years agoAudioStreamIn and AudioStreamOut
Glenn Kasten [Thu, 26 Jan 2012 17:46:34 +0000 (09:46 -0800)]
AudioStreamIn and AudioStreamOut

These are immutable, so make the fields const.
getOutput() and getInput() methods are now const.

Change-Id: I128246ebd56ea50b3e542be43f2aa1bcb55f1373

12 years agoThe software AAC encoder is now an OMX component.
Andreas Huber [Thu, 26 Jan 2012 22:39:50 +0000 (14:39 -0800)]
The software AAC encoder is now an OMX component.

Yay.

Change-Id: I74938a20b4e0a622836ea5184d3761180eb0f5de

12 years agoCleanup thread types
Glenn Kasten [Thu, 26 Jan 2012 18:38:26 +0000 (10:38 -0800)]
Cleanup thread types

Use type_t instead of int for thread types.
Initialize ThreadBase::mType in constructor and make it const.

Change-Id: I43d141388b9639e4783c30b97dbda5688bf7555f

12 years agoDeclare methods in binder opcode order
Glenn Kasten [Thu, 26 Jan 2012 21:40:12 +0000 (13:40 -0800)]
Declare methods in binder opcode order

This makes it easier to compare interface and implementation.

Change-Id: Ie060e43dec348902abcf40f5a610cec639d6d0d3

12 years agoUse enum mixer_state consistently
Glenn Kasten [Thu, 26 Jan 2012 21:37:52 +0000 (13:37 -0800)]
Use enum mixer_state consistently

Change-Id: I5b71ed20f939dfc4b98143334b7aa064d282f584

12 years agoUse enum effect_state consistently
Glenn Kasten [Thu, 26 Jan 2012 21:43:46 +0000 (13:43 -0800)]
Use enum effect_state consistently

Also fix indentation

Change-Id: I393ef9e37ffceed5ad4a78df439726ae1fe139df

12 years agoUse enum track_state consistently
Glenn Kasten [Thu, 26 Jan 2012 21:39:18 +0000 (13:39 -0800)]
Use enum track_state consistently

Change-Id: Ie5ebb7befa092e1de1e4df9c6e2d51e6bcfd176a

12 years agoPreserve NEON callee-save registers d8-d15.
Ben Cheng [Wed, 25 Jan 2012 21:42:42 +0000 (13:42 -0800)]
Preserve NEON callee-save registers d8-d15.

BUG:5918479
Change-Id: I574bd5845d9e7e007c637aabbf170d03217d10ad

12 years agoam 95306d85: am 64085d03: am 12045502: Merge "stagefright aacenc: Fix type definition...
Jean-Baptiste Queru [Wed, 25 Jan 2012 18:33:42 +0000 (10:33 -0800)]
am 95306d85: am 64085d03: am 12045502: Merge "stagefright aacenc: Fix type definitions for 64 bit platforms"

* commit '95306d85005f165e10a25ad45cadadec6e1f6a31':
  stagefright aacenc: Fix type definitions for 64 bit platforms

12 years agoam 64085d03: am 12045502: Merge "stagefright aacenc: Fix type definitions for 64...
Jean-Baptiste Queru [Wed, 25 Jan 2012 18:30:31 +0000 (10:30 -0800)]
am 64085d03: am 12045502: Merge "stagefright aacenc: Fix type definitions for 64 bit platforms"

* commit '64085d03efcd2abf3b8cd64f4e6344a1ece014c3':
  stagefright aacenc: Fix type definitions for 64 bit platforms

12 years agoam 12045502: Merge "stagefright aacenc: Fix type definitions for 64 bit platforms"
Jean-Baptiste Queru [Wed, 25 Jan 2012 18:27:41 +0000 (10:27 -0800)]
am 12045502: Merge "stagefright aacenc: Fix type definitions for 64 bit platforms"

* commit '12045502c49bfee40d7d3cb806d9b70d3b5d3ed3':
  stagefright aacenc: Fix type definitions for 64 bit platforms

12 years agoMerge "stagefright aacenc: Fix type definitions for 64 bit platforms"
Jean-Baptiste Queru [Wed, 25 Jan 2012 17:17:45 +0000 (09:17 -0800)]
Merge "stagefright aacenc: Fix type definitions for 64 bit platforms"

12 years agoam e2594b85: am 4cd2b343: am 98539a4e: Merge "stagefright amrwbenc: Make the mem_mall...
Jean-Baptiste Queru [Wed, 25 Jan 2012 16:57:36 +0000 (08:57 -0800)]
am e2594b85: am 4cd2b343: am 98539a4e: Merge "stagefright amrwbenc: Make the mem_malloc function 64-bit safe"

* commit 'e2594b85093441e03150c7cfc540abc626884eaf':
  stagefright amrwbenc: Make the mem_malloc function 64-bit safe

12 years agoam a3946104: am fc2ea24b: am d80cbb8c: Merge "stagefright aacenc: Add missing newline...
Jean-Baptiste Queru [Wed, 25 Jan 2012 16:57:32 +0000 (08:57 -0800)]
am a3946104: am fc2ea24b: am d80cbb8c: Merge "stagefright aacenc: Add missing newlines at the end of files"

* commit 'a39461047eae6022e2c0ef873ffa2c7241b393a9':
  stagefright aacenc: Add missing newlines at the end of files

12 years agoam 535b0264: am 7eeaf3f0: Merge "AudioFlinger: refine mixer sleep time logic" into...
Eric Laurent [Wed, 25 Jan 2012 16:57:23 +0000 (08:57 -0800)]
am 535b0264: am 7eeaf3f0: Merge "AudioFlinger: refine mixer sleep time logic" into ics-mr1

* commit '535b0264a4cfa790e549bd9cd09980788f1375f4':
  AudioFlinger: refine mixer sleep time logic

12 years agoam dd347d51: (-s ours) am 41773d46: Merge "DO NOT MERGE Revert "Revert "AudioFlinger...
Eric Laurent [Wed, 25 Jan 2012 16:57:20 +0000 (08:57 -0800)]
am dd347d51: (-s ours) am 41773d46: Merge "DO NOT MERGE Revert "Revert "AudioFlinger: mix track only when really ready (2)""" into ics-mr1

* commit 'dd347d515d4a4e42b2aba89a8e471392026b7824':
  DO NOT MERGE Revert "Revert "AudioFlinger: mix track only when really ready (2)""

12 years agoam 4cd2b343: am 98539a4e: Merge "stagefright amrwbenc: Make the mem_malloc function...
Jean-Baptiste Queru [Tue, 24 Jan 2012 23:15:36 +0000 (15:15 -0800)]
am 4cd2b343: am 98539a4e: Merge "stagefright amrwbenc: Make the mem_malloc function 64-bit safe"

* commit '4cd2b343272429adae1fc68e0ec0628c69e88e6d':
  stagefright amrwbenc: Make the mem_malloc function 64-bit safe

12 years agoam fc2ea24b: am d80cbb8c: Merge "stagefright aacenc: Add missing newlines at the...
Jean-Baptiste Queru [Tue, 24 Jan 2012 23:15:33 +0000 (15:15 -0800)]
am fc2ea24b: am d80cbb8c: Merge "stagefright aacenc: Add missing newlines at the end of files"

* commit 'fc2ea24b024f7a015d67054adb7bb09ff3a0a97b':
  stagefright aacenc: Add missing newlines at the end of files

12 years agoam 98539a4e: Merge "stagefright amrwbenc: Make the mem_malloc function 64-bit safe"
Jean-Baptiste Queru [Tue, 24 Jan 2012 23:13:45 +0000 (15:13 -0800)]
am 98539a4e: Merge "stagefright amrwbenc: Make the mem_malloc function 64-bit safe"

* commit '98539a4eba7677c0b1ddc12f800c7075851ef302':
  stagefright amrwbenc: Make the mem_malloc function 64-bit safe

12 years agoam d80cbb8c: Merge "stagefright aacenc: Add missing newlines at the end of files"
Jean-Baptiste Queru [Tue, 24 Jan 2012 23:13:42 +0000 (15:13 -0800)]
am d80cbb8c: Merge "stagefright aacenc: Add missing newlines at the end of files"

* commit 'd80cbb8c7b13203704c00e2f5941588811c15665':
  stagefright aacenc: Add missing newlines at the end of files

12 years agoMerge "stagefright amrwbenc: Make the mem_malloc function 64-bit safe"
Jean-Baptiste Queru [Tue, 24 Jan 2012 22:26:30 +0000 (14:26 -0800)]
Merge "stagefright amrwbenc: Make the mem_malloc function 64-bit safe"

12 years agomerge in ics-mr1-release history after reset to ics-mr1
The Android Automerger [Tue, 24 Jan 2012 21:33:01 +0000 (13:33 -0800)]
merge in ics-mr1-release history after reset to ics-mr1

12 years agoam 7eeaf3f0: Merge "AudioFlinger: refine mixer sleep time logic" into ics-mr1
Eric Laurent [Tue, 24 Jan 2012 19:51:50 +0000 (11:51 -0800)]
am 7eeaf3f0: Merge "AudioFlinger: refine mixer sleep time logic" into ics-mr1

* commit '7eeaf3f07aa6fb10639d9f96c1367eb98c3e8839':
  AudioFlinger: refine mixer sleep time logic

12 years agoam 41773d46: Merge "DO NOT MERGE Revert "Revert "AudioFlinger: mix track only when...
Eric Laurent [Tue, 24 Jan 2012 19:51:47 +0000 (11:51 -0800)]
am 41773d46: Merge "DO NOT MERGE Revert "Revert "AudioFlinger: mix track only when really ready (2)""" into ics-mr1

* commit '41773d46556aa47d4322ff89fdaf7d1345c2d1f2':
  DO NOT MERGE Revert "Revert "AudioFlinger: mix track only when really ready (2)""

12 years agoMerge "stagefright aacenc: Add missing newlines at the end of files"
Jean-Baptiste Queru [Tue, 24 Jan 2012 18:34:24 +0000 (10:34 -0800)]
Merge "stagefright aacenc: Add missing newlines at the end of files"

12 years agoam 10c4641c: am 2ae5805c: am 164a9842: Merge "stagefright aacenc: Properly terminate...
Jean-Baptiste Queru [Tue, 24 Jan 2012 16:46:26 +0000 (08:46 -0800)]
am 10c4641c: am 2ae5805c: am 164a9842: Merge "stagefright aacenc: Properly terminate a comment"

* commit '10c4641c2e2f747f8fef7d217d550b7f9a6de588':
  stagefright aacenc: Properly terminate a comment