OSDN Git Service

android-x86/frameworks-av.git
11 years agoMerge "h264 streaming: make profile-level-id optional"
Jean-Baptiste Queru [Tue, 21 Aug 2012 16:18:31 +0000 (09:18 -0700)]
Merge "h264 streaming: make profile-level-id optional"

11 years agoMerge "avcenc: Fix the gcc inline assembly constraints"
Jean-Baptiste Queru [Tue, 21 Aug 2012 15:37:26 +0000 (08:37 -0700)]
Merge "avcenc: Fix the gcc inline assembly constraints"

11 years agoMerge "avcenc: Properly indent assembly blocks"
Jean-Baptiste Queru [Tue, 21 Aug 2012 15:36:58 +0000 (08:36 -0700)]
Merge "avcenc: Properly indent assembly blocks"

11 years agoavcenc: Fix the gcc inline assembly constraints
Martin Storsjo [Fri, 13 Apr 2012 11:17:30 +0000 (14:17 +0300)]
avcenc: Fix the gcc inline assembly constraints

Use +r for registers whose value is used as input to the block.

Also switch input registers to output registers with +r, for
registers that are modified by the assembly block (SUB_SAD in
sad_inline.h).

Without this, the inline assembly fails on modern gcc versions
(tested on gcc 4.6), where the intended input value never is
loaded into the registers, causing segfaults at runtime.

Currently this assembly isn't used, but this change makes it
work properly if enabled.

Change-Id: I37c247cd5c8aff213445e31ab3bd33d8f0f37e05

11 years agoavcenc: Properly indent assembly blocks
Martin Storsjo [Fri, 13 Apr 2012 11:16:54 +0000 (14:16 +0300)]
avcenc: Properly indent assembly blocks

Also line break multiline assembly blocks - previously they were
virtually unreadable.

Change-Id: Icb269909b78746e26b28ab7dcb6979c4655a0b0c

11 years agoMerge "hardening: eliminate text relocations from h264"
Nick Kralevich [Mon, 20 Aug 2012 18:55:17 +0000 (11:55 -0700)]
Merge "hardening: eliminate text relocations from h264"

11 years agoMerge "MPEG2TSWriter: Fix an off-by-one in the condition for padding"
Jean-Baptiste Queru [Mon, 20 Aug 2012 15:54:11 +0000 (08:54 -0700)]
Merge "MPEG2TSWriter: Fix an off-by-one in the condition for padding"

11 years agoMerge "avcenc: Remove a leftover define"
Jean-Baptiste Queru [Mon, 20 Aug 2012 15:53:54 +0000 (08:53 -0700)]
Merge "avcenc: Remove a leftover define"

11 years agoMPEG2TSWriter: Fix an off-by-one in the condition for padding
Martin Storsjo [Mon, 20 Aug 2012 13:34:14 +0000 (14:34 +0100)]
MPEG2TSWriter: Fix an off-by-one in the condition for padding

This makes a remuxed file play without any decoding warnings,
while there still were some occasional warnings with the previous
version.

Change-Id: I31d73446ded1eeb3da10783c5ef841270563ca12

11 years agoh264 streaming: make profile-level-id optional
Patrik2 Carlsson [Mon, 20 Aug 2012 07:53:09 +0000 (09:53 +0200)]
h264 streaming: make profile-level-id optional

profile-level-id is made optional according to rfc3984:
"If no profile-level-id is present, the Baseline Profile without
additional constraints at Level 1 MUST be implied."

Change-Id: If868468a48917ceccb963b8ac15767583da29723

11 years agoavcenc: Remove a leftover define
Martin Storsjo [Fri, 17 Aug 2012 08:28:09 +0000 (11:28 +0300)]
avcenc: Remove a leftover define

This should have been removed as part of f5af6314db, but
the second __arm__ define was introduced after the original commit
was written, and not removing the second one was an oversight in
rebasing the patch on top of the new code.

Change-Id: Iccf08ffdef130446dd6e97dc1e15e6a1d4f9386f

11 years agoMerge "stagefright amrnb: Remove unused leftover source files"
Jean-Baptiste Queru [Thu, 16 Aug 2012 23:42:06 +0000 (16:42 -0700)]
Merge "stagefright amrnb: Remove unused leftover source files"

11 years agoMerge "stagefright amrnb: Move the bitstream_format enum into frame_type_3gpp.h"
Jean-Baptiste Queru [Thu, 16 Aug 2012 23:40:19 +0000 (16:40 -0700)]
Merge "stagefright amrnb: Move the bitstream_format enum into frame_type_3gpp.h"

11 years agoMerge "hardening: eliminate text relocations from mp3"
Jean-Baptiste Queru [Thu, 16 Aug 2012 21:45:30 +0000 (14:45 -0700)]
Merge "hardening: eliminate text relocations from mp3"

11 years agoMerge "avcenc: Clarify the ifdefs surrounding inline assembly"
Jean-Baptiste Queru [Thu, 16 Aug 2012 21:44:54 +0000 (14:44 -0700)]
Merge "avcenc: Clarify the ifdefs surrounding inline assembly"

11 years agoMerge "hardening: eliminate text relocations from amrwb"
Jean-Baptiste Queru [Thu, 16 Aug 2012 21:44:39 +0000 (14:44 -0700)]
Merge "hardening: eliminate text relocations from amrwb"

11 years agoMerge "aacenc/amrwbenc: Remove the now unused LINUX define"
Jean-Baptiste Queru [Thu, 16 Aug 2012 21:42:35 +0000 (14:42 -0700)]
Merge "aacenc/amrwbenc: Remove the now unused LINUX define"

11 years agoMerge "aacenc/amrwbenc: Remove unused and dubious typedefs"
Jean-Baptiste Queru [Thu, 16 Aug 2012 21:42:16 +0000 (14:42 -0700)]
Merge "aacenc/amrwbenc: Remove unused and dubious typedefs"

11 years agoaacenc/amrwbenc: Remove the now unused LINUX define
Martin Storsjo [Wed, 7 Mar 2012 11:17:52 +0000 (13:17 +0200)]
aacenc/amrwbenc: Remove the now unused LINUX define

This is no longer used by the codec code (except for the
amrwbenc sample).

Change-Id: I78c7f62f257c490bf44477217975a93abcc7779e

11 years agoaacenc/amrwbenc: Remove unused and dubious typedefs
Martin Storsjo [Thu, 23 Jun 2011 07:53:55 +0000 (10:53 +0300)]
aacenc/amrwbenc: Remove unused and dubious typedefs

These typedefs aren't used anywhere, and their description
doesn't really match their actual definition. The VO_TCHAR
type is even defined as pointer for one configuration and
as non-pointer for another configuration.

This gets rid of the last use of the LINUX define in the
VO codecs (except for the amrwbenc sample application).
Within the codecs, the use of this define is unclear - does it
indicate code specific to android but not normal linux, linux
but not other unixes, unix but not windows, or gcc but not
other compilers, etc.

Originally, prior to cleanup, the define has meant many
different of these - these uses has now been cleaned up to
use the proper feature detection macros (e.g. looking for
defines automatically set by the compiler, instead of
relying on manually setting this define).

Change-Id: I1a675d653d6d305f558aa7f74c7255858dc62758

11 years agoMerge "MPEG2TSWriter: Correctly add padding using the adaptation field"
Jean-Baptiste Queru [Wed, 15 Aug 2012 17:14:29 +0000 (10:14 -0700)]
Merge "MPEG2TSWriter: Correctly add padding using the adaptation field"

11 years agoMPEG2TSWriter: Correctly add padding using the adaptation field
Paul Fariello [Mon, 23 Jan 2012 16:32:06 +0000 (17:32 +0100)]
MPEG2TSWriter: Correctly add padding using the adaptation field

Change-Id: I268cce36ada910ba168c4c9dcedab0b20f6e6ebb

11 years agoMerge "amrnb: Restart the p_xn pointer when retraversing the array on overflow"
Jean-Baptiste Queru [Mon, 13 Aug 2012 21:37:57 +0000 (14:37 -0700)]
Merge "amrnb: Restart the p_xn pointer when retraversing the array on overflow"

11 years agohardening: eliminate text relocations from h264
Ard Biesheuvel [Mon, 13 Aug 2012 20:55:51 +0000 (22:55 +0200)]
hardening: eliminate text relocations from h264

Change-Id: I8bdf9746f28f2f7bed6c1109c0f4badbab765a7c

11 years agohardening: eliminate text relocations from mp3
Ard Biesheuvel [Mon, 13 Aug 2012 20:47:33 +0000 (22:47 +0200)]
hardening: eliminate text relocations from mp3

Change-Id: I671ed4c8c7bb15172a8b64d513e2844319495102

11 years agohardening: eliminate text relocations from amrwb
Ard Biesheuvel [Mon, 13 Aug 2012 20:44:43 +0000 (22:44 +0200)]
hardening: eliminate text relocations from amrwb

Change-Id: I47235fc4afe6b916da1e999401fb3c8b12c94e62

11 years agoMerge "amrnb: Fix audio distortion in an AMR-NB clip"
Jean-Baptiste Queru [Mon, 13 Aug 2012 17:37:03 +0000 (10:37 -0700)]
Merge "amrnb: Fix audio distortion in an AMR-NB clip"

11 years agoMerge "MPEG2TSWriter: Write a proper CRC in PAT and PMT"
Jean-Baptiste Queru [Fri, 10 Aug 2012 22:26:43 +0000 (15:26 -0700)]
Merge "MPEG2TSWriter: Write a proper CRC in PAT and PMT"

11 years agoMerge "MPEG2TSWriter: Write proper continuity counters for PAT and PMT"
Jean-Baptiste Queru [Fri, 10 Aug 2012 16:03:54 +0000 (09:03 -0700)]
Merge "MPEG2TSWriter: Write proper continuity counters for PAT and PMT"

11 years agoMerge "Fix initialization of audio mixer track resources for MIPS."
Jean-Baptiste Queru [Fri, 10 Aug 2012 16:03:15 +0000 (09:03 -0700)]
Merge "Fix initialization of audio mixer track resources for MIPS."

11 years agoMerge "Store correct time stamps in recorded mp4 files"
Jean-Baptiste Queru [Wed, 8 Aug 2012 20:40:33 +0000 (13:40 -0700)]
Merge "Store correct time stamps in recorded mp4 files"

11 years agoMerge "Correct ID3::StringSize calculation for UCS-2 data."
Jean-Baptiste Queru [Wed, 8 Aug 2012 19:58:42 +0000 (12:58 -0700)]
Merge "Correct ID3::StringSize calculation for UCS-2 data."

11 years agoMerge "avcenc: Remove redundant zero-initialization of allocated memory"
Jean-Baptiste Queru [Wed, 8 Aug 2012 19:09:01 +0000 (12:09 -0700)]
Merge "avcenc: Remove redundant zero-initialization of allocated memory"

11 years agoMerge "avcenc: Initialize all memory allocated by the CBAVC_Malloc callback function"
Jean-Baptiste Queru [Wed, 8 Aug 2012 19:08:44 +0000 (12:08 -0700)]
Merge "avcenc: Initialize all memory allocated by the CBAVC_Malloc callback function"

11 years agoMerge "avcenc: Remove useless casts"
Jean-Baptiste Queru [Wed, 8 Aug 2012 19:08:19 +0000 (12:08 -0700)]
Merge "avcenc: Remove useless casts"

11 years agoMerge "Initialize pool's buffer pointers to null."
Jean-Baptiste Queru [Wed, 8 Aug 2012 19:08:04 +0000 (12:08 -0700)]
Merge "Initialize pool's buffer pointers to null."

11 years agoMerge "avcenc: Fix indentation"
Jean-Baptiste Queru [Wed, 8 Aug 2012 19:07:47 +0000 (12:07 -0700)]
Merge "avcenc: Fix indentation"

11 years agoMerge "avcenc: Switch malloc/free callbacks to use pointers instead of ints"
Jean-Baptiste Queru [Wed, 8 Aug 2012 19:07:27 +0000 (12:07 -0700)]
Merge "avcenc: Switch malloc/free callbacks to use pointers instead of ints"

11 years agoMerge "mediacodec: Return an error on getOutputFormat if there is no output format...
Jean-Baptiste Queru [Wed, 8 Aug 2012 19:06:48 +0000 (12:06 -0700)]
Merge "mediacodec: Return an error on getOutputFormat if there is no output format yet"

12 years agoFix initialization of audio mixer track resources for MIPS.
Paul Lind [Thu, 2 Aug 2012 01:49:49 +0000 (18:49 -0700)]
Fix initialization of audio mixer track resources for MIPS.

The value 1 << 32 (maxNumTracks defaults to 32) is surprisingly not defined
in C, and differs on MIPS than the other arch.

Therefore the track resources were not initialized properly, resulting
in failure to play any audio.

The fix allows the mConfiguredNames bitmask to be correctly set to all
1's in the 32-track case, for all arch.

Change-Id: Ied3e1305952e9567602e2cd76c5ef3acb0809ee0
Signed-off-by: Paul Lind <plind@mips.com>
12 years agoMerge "Shutter sound is heard at the end of the recorded video"
Eino-Ville Talvala [Wed, 1 Aug 2012 23:13:50 +0000 (16:13 -0700)]
Merge "Shutter sound is heard at the end of the recorded video"

12 years agoStore correct time stamps in recorded mp4 files
Johannes Carlsson [Thu, 8 Mar 2012 10:06:13 +0000 (11:06 +0100)]
Store correct time stamps in recorded mp4 files

According to the specification (ISO_IEC_14496-12_2008 page 17, 19 and
21) the time is "in seconds since midnight, Jan. 1, 1904, in UTC time"
which is not the same as Unix epoch time. This is already correctly
handled in MPEG4Extractor.cpp when reading files.

The specification also supports 64 bit timestamps which would make
it a bit more future proof, but unfortunately it seems like support
for this in players are poor (tested both Windows media player and
Quicktime player).

Change-Id: I102d728f098b8b0a21013956f017b0c4f2d69112

12 years agomediacodec: Return an error on getOutputFormat if there is no output format yet
Martin Storsjo [Fri, 13 Jul 2012 09:48:19 +0000 (12:48 +0300)]
mediacodec: Return an error on getOutputFormat if there is no output format yet

This avoids crashing if calling the public java method
MediaCodec.getOutputFormat immediately after MediaCodec.start
(as the documentation example shows that one should do!).

Change-Id: Icd3dd7ab9ae9e8f2a67cef171ff176735de95f58

12 years agoavcenc: Remove redundant zero-initialization of allocated memory
Martin Storsjo [Mon, 16 Apr 2012 17:51:34 +0000 (20:51 +0300)]
avcenc: Remove redundant zero-initialization of allocated memory

The allocation callback is now supposed to always return initialized
memory.

Change-Id: I7a4d036e498c9f87c730dda3aab5057eb764a1e7

12 years agoavcenc: Initialize all memory allocated by the CBAVC_Malloc callback function
Martin Storsjo [Mon, 16 Apr 2012 17:50:37 +0000 (20:50 +0300)]
avcenc: Initialize all memory allocated by the CBAVC_Malloc callback function

Valgrind reported use of uninitialized memory in AVCEncodeSlice,
which this fixes.

Change-Id: Ia09ad3e50d05b2b0487a4d588d9b00fc0828c816

12 years agoInitialize pool's buffer pointers to null.
Truls Bengtsson [Wed, 11 Jul 2012 13:50:44 +0000 (15:50 +0200)]
Initialize pool's buffer pointers to null.

If these pointers are not properly initialized to null and
allocation of one of the buffers subsequently fail, the freePool
mechanism might erroneously free uninitialized pointers.

Change-Id: I2101bdb2cc576e5f50b5a1440e6fc0c54dc4e81b

12 years agoCorrect ID3::StringSize calculation for UCS-2 data.
Jesper Tragardh [Wed, 11 Jul 2012 12:13:47 +0000 (14:13 +0200)]
Correct ID3::StringSize calculation for UCS-2 data.

Album art was not displayed when albums were transferred via MediaGo
because the length of the description in the APIC tag was not
correctly calculated.

The method StringSize used to calculate the length of the tag
does not add the size of the null termination in UCS-2 case.

Change-Id: I5409d42f5e87d315e6e03c4d7e6cbd8a378a160d

12 years agoavcenc: Fix indentation
Martin Storsjo [Fri, 13 Apr 2012 11:17:14 +0000 (14:17 +0300)]
avcenc: Fix indentation

Change-Id: I4fcdf619e3b75387a7909272f3164b8d608b6cc2

12 years agoShutter sound is heard at the end of the recorded video
Patric Frederiksen [Tue, 10 Jul 2012 11:38:35 +0000 (13:38 +0200)]
Shutter sound is heard at the end of the recorded video

On fast hardware, which is more common, the shutter sound is
played and recorded before video recording is stopped.

Before we call play shutter sound before calling stop recording.
With this fix we call play shutter sound after calling stop recording.

Change-Id: I06a3e65fd2700f1d2457fea5ff0f64500eac436f

12 years agoavcenc: Clarify the ifdefs surrounding inline assembly
Martin Storsjo [Fri, 13 Apr 2012 11:23:26 +0000 (14:23 +0300)]
avcenc: Clarify the ifdefs surrounding inline assembly

Previously, the code was structured like this:

#if defined(__GNUC__) && defined(__arm__)
    /* Generic C implementation */
#elif defined(__CC_ARM)
    /* RVCT version */
#elif defined(__GNUC__) && defined(__arm__)
    /* GCC arm assembly version */
#endif

This had two implications - the gcc arm version never was used
in practice, and the code required -D__arm__ to build at all for
other architectures.

The inline assembly constraints are buggy (requires changes so as
to not crash), and if fixed, the optimizations still aren't faster
than the generic C code on modern gcc versions. Therefore, just
ignore them and use the generic C code.

Due to the peculiar code block ordering and preprocessor conditions,
the generic C code is what has been used before anyway, but it
did require -D__arm__ to build, which can cause problems if building
for other architectures, if the system headers have ifdefs for this
define. This change makes it build without that define.

Change-Id: Ib3fcf5651ee720310b7dff10fd04279b5f90759c

12 years agoavcenc: Remove useless casts
Martin Storsjo [Fri, 13 Apr 2012 11:15:36 +0000 (14:15 +0300)]
avcenc: Remove useless casts

Change-Id: I73a6f74268bcf6ca22d489d13936be8fd1b3fc92

12 years agoavcenc: Switch malloc/free callbacks to use pointers instead of ints
Martin Storsjo [Fri, 13 Apr 2012 11:15:08 +0000 (14:15 +0300)]
avcenc: Switch malloc/free callbacks to use pointers instead of ints

There is no reason for casting the pointers to ints. This fixes
building the code on platforms where pointers are larger than ints,
e.g. 64 bit platforms.

Change-Id: I910cd207d0908287931c9a96eb270139967e029b

12 years agostagefright amrnb: Remove unused leftover source files
Martin Storsjo [Thu, 23 Jun 2011 14:56:25 +0000 (17:56 +0300)]
stagefright amrnb: Remove unused leftover source files

This allows getting rid of adding the plain common directory
(not common/include) to the include paths in projects that use
this code, too.

Change-Id: Ifb1fe4fc1f5c6f1264cb2b64f7b78e42b1e7833a

12 years agostagefright amrnb: Move the bitstream_format enum into frame_type_3gpp.h
Martin Storsjo [Thu, 1 Mar 2012 13:03:29 +0000 (15:03 +0200)]
stagefright amrnb: Move the bitstream_format enum into frame_type_3gpp.h

Change-Id: I6d6f9dc5799750394a853325466ffa2ebec44625

12 years agoamrnb: Fix audio distortion in an AMR-NB clip
PacketVideo CM [Wed, 18 May 2011 09:31:42 +0000 (12:31 +0300)]
amrnb: Fix audio distortion in an AMR-NB clip

This fix had the PacketVideo internal tracking number
RIO-9190.

This was committed on the master branch of the opencore
repository in commit 7d6ccc907d3a15ca2ff74b9c76f122ab87d0ffbe.

Change-Id: Icbd7d6fe94e02cae360946cfb6f3b8dc90c8b518

12 years agoamrnb: Restart the p_xn pointer when retraversing the array on overflow
Martin Storsjo [Wed, 18 May 2011 09:29:28 +0000 (12:29 +0300)]
amrnb: Restart the p_xn pointer when retraversing the array on overflow

This is a safeguard to avoid issues in a rare codepath.

This was committed on the master branch of the opencore
repository in commit f532d145194e474cb30d8644b8eee83873ad24cc,
based on AOSP contribution 10905.

Change-Id: I9688b009bcbe63d88ee5040e883a021aefa30664

12 years agoMPEG2TSWriter: Write a proper CRC in PAT and PMT
Paul Fariello [Mon, 23 Jan 2012 16:21:33 +0000 (17:21 +0100)]
MPEG2TSWriter: Write a proper CRC in PAT and PMT

Change-Id: I6ec8039b176878da6f02691194d96b0a1e894055

12 years agoMPEG2TSWriter: Write proper continuity counters for PAT and PMT
Paul Fariello [Mon, 23 Jan 2012 16:24:35 +0000 (17:24 +0100)]
MPEG2TSWriter: Write proper continuity counters for PAT and PMT

Change-Id: If94e31e51c563867db6d2376ad771f9751416084

12 years agoReconcile with jb-release
The Android Open Source Project [Tue, 3 Jul 2012 19:11:47 +0000 (12:11 -0700)]
Reconcile with jb-release

Change-Id: I7c52b8e82465f28b32e8ebf6df56c258458f2cd3

12 years agomerge in jb-release history after reset to jb-dev
The Android Automerger [Mon, 2 Jul 2012 19:54:56 +0000 (12:54 -0700)]
merge in jb-release history after reset to jb-dev

12 years agoam 109347d4: audioflinger: fix regression in attachAuxEffect().
Eric Laurent [Mon, 2 Jul 2012 19:43:14 +0000 (12:43 -0700)]
am 109347d4: audioflinger: fix regression in attachAuxEffect().

* commit '109347d421413303eb1678dd9e2aa9d40acf89d2':
  audioflinger: fix regression in attachAuxEffect().

12 years agoaudioflinger: fix regression in attachAuxEffect().
Eric Laurent [Mon, 2 Jul 2012 19:31:03 +0000 (12:31 -0700)]
audioflinger: fix regression in attachAuxEffect().

Commit 717e1286 introduced a regression in PlaybackThread::Track::attachAuxEffect()
when called with an effect ID of 0 to detach the auxiliary effect.

It is normal in this case that AudioFlinger::getEffectThread_l() returns 0.

Bug 6768757.

Change-Id: I7430bd1aad2f68da38f7c3e4794e7ad657bfc6be

12 years agoam 717e1286: audioflinger: fix auxiliary effect attachment
Eric Laurent [Sun, 1 Jul 2012 16:33:59 +0000 (09:33 -0700)]
am 717e1286: audioflinger: fix auxiliary effect attachment

* commit '717e128691f083a9469a1d0e363ac6ecd5c65d58':
  audioflinger: fix auxiliary effect attachment

12 years agoaudioflinger: fix auxiliary effect attachment
Eric Laurent [Fri, 29 Jun 2012 23:36:52 +0000 (16:36 -0700)]
audioflinger: fix auxiliary effect attachment

Auxiliary effects (Reverb) are global effects and as such follow
the default rule which is to attach them to the output thread that
handles music streams by default. This causes a problem when several
threads are eligible to handle music streams as tracks can be attached
to either thread based on criteria unknown when teh effect is created.

The fix consists in moving the auxiliary effect if necessary when an
AudioTrack is attached to it and this track is not on the same
output thread.

Bug 6608561.

Change-Id: Ib32c3cabc731b2046aba728be1771982999c6069

12 years agoReconcile with jb-release
The Android Open Source Project [Mon, 25 Jun 2012 14:57:15 +0000 (07:57 -0700)]
Reconcile with jb-release

Change-Id: I1ce46fa6332f684c76c276117c7b406bdd5e9336

12 years agomerge in jb-release history after reset to jb-dev
The Android Automerger [Sat, 23 Jun 2012 01:21:17 +0000 (18:21 -0700)]
merge in jb-release history after reset to jb-dev

12 years agoam 33e28dd3: Merge "SurfaceMediaSource: keep refs to current buffers" into jb-dev
Jamie Gennis [Fri, 22 Jun 2012 22:20:48 +0000 (15:20 -0700)]
am 33e28dd3: Merge "SurfaceMediaSource: keep refs to current buffers" into jb-dev

* commit '33e28dd3c033641f4db9d1b99c7c88d6d08c61f7':
  SurfaceMediaSource: keep refs to current buffers

12 years agoMerge "SurfaceMediaSource: keep refs to current buffers" into jb-dev
Jamie Gennis [Fri, 22 Jun 2012 22:18:26 +0000 (15:18 -0700)]
Merge "SurfaceMediaSource: keep refs to current buffers" into jb-dev

12 years agoSurfaceMediaSource: keep refs to current buffers
Jamie Gennis [Fri, 22 Jun 2012 21:42:00 +0000 (14:42 -0700)]
SurfaceMediaSource: keep refs to current buffers

This change fixes a bug in SurfaceMediaSource where it would not keep a
reference to all of the Gralloc buffers that the video encoder is currently
using.

Bug: 6655597
Change-Id: Ifd99976cc7ae57ed724bbf2bbcff357a39497391

12 years agoReconcile with jb-release
The Android Open Source Project [Fri, 22 Jun 2012 15:16:51 +0000 (08:16 -0700)]
Reconcile with jb-release

Change-Id: Iea91c1026e7fc3a59815e01e4e4168f875527e7c

12 years agoRevert input buffer count for the SoftAAC* decoders back to 4.
Andreas Huber [Wed, 20 Jun 2012 19:46:38 +0000 (12:46 -0700)]
Revert input buffer count for the SoftAAC* decoders back to 4.

This improves the netflix bandwidth change responsiveness.
With the old count of 32 the audio decoder running at a higher
priority would starve the video decoder for about 1 sec while
draining the input data.

I haven't been able to break any stream playback using MediaCodec
using the lower count, even though I'm sure I had a good reason
to increase it in the first place (see bug 6478823)

Change-Id: I7fd3e71794b0cc1570ea9d27d1d869ab1e1750c4
related-to-bug: 6685031

12 years agoam 39a31135: Revert input buffer count for the SoftAAC* decoders back to 4.
Andreas Huber [Wed, 20 Jun 2012 21:31:30 +0000 (14:31 -0700)]
am 39a31135: Revert input buffer count for the SoftAAC* decoders back to 4.

* commit '39a3113539f9bcc86a8efe562780c23ac8731098':
  Revert input buffer count for the SoftAAC* decoders back to 4.

12 years agoRevert input buffer count for the SoftAAC* decoders back to 4.
Andreas Huber [Wed, 20 Jun 2012 19:46:38 +0000 (12:46 -0700)]
Revert input buffer count for the SoftAAC* decoders back to 4.

This improves the netflix bandwidth change responsiveness.
With the old count of 32 the audio decoder running at a higher
priority would starve the video decoder for about 1 sec while
draining the input data.

I haven't been able to break any stream playback using MediaCodec
using the lower count, even though I'm sure I had a good reason
to increase it in the first place (see bug 6478823)

Change-Id: I7fd3e71794b0cc1570ea9d27d1d869ab1e1750c4
related-to-bug: 6685031

12 years agoReconcile with jb-release
The Android Open Source Project [Tue, 19 Jun 2012 13:13:22 +0000 (06:13 -0700)]
Reconcile with jb-release

Change-Id: Ic4989b359b302c5d72751cdebae3433064770a29

12 years agoam eb15716b: Change definition of warmup period
Glenn Kasten [Mon, 18 Jun 2012 17:32:49 +0000 (10:32 -0700)]
am eb15716b: Change definition of warmup period

* commit 'eb15716b59020f342df62bce5b293f0603b94861':
  Change definition of warmup period

12 years agoam 972af221: FastMixer compensates for SRC jitter in audio HAL
Glenn Kasten [Mon, 18 Jun 2012 17:32:13 +0000 (10:32 -0700)]
am 972af221: FastMixer compensates for SRC jitter in audio HAL

* commit '972af221ae7253e406e0e1ea853e56a3010ed6b1':
  FastMixer compensates for SRC jitter in audio HAL

12 years agomerge in jb-release history after reset to jb-dev
The Android Automerger [Sat, 16 Jun 2012 14:01:55 +0000 (07:01 -0700)]
merge in jb-release history after reset to jb-dev

12 years agoChange definition of warmup period
Glenn Kasten [Wed, 13 Jun 2012 21:59:07 +0000 (14:59 -0700)]
Change definition of warmup period

Previously, warmup was considered done as soon as any write() took
more than 0.5 nominal cycle time.  In practice, this was always the
first write() that turned on power to the output path, and it didn't
accurately account for filling the full kernel buffer queue, or for
buffering in the HAL sample rate conversion.

Now warmup is considered done when a write() _after_ the first write
takes more than 0.5 nominal cycle time.

This will throttle the initial pull rate after coming out of standby.
When combined with another change to throttle the pull rate for
devices with HAL sample rate conversion, it may help reduce some of the
notification glitches.  The only downside is that it will increase the
warmup time a bit.

Bug: 650831766508466607056
Change-Id: I39f324c5195578170a55308e9601d3a1b41db3e0

12 years agoFastMixer compensates for SRC jitter in audio HAL
Glenn Kasten [Thu, 14 Jun 2012 00:14:03 +0000 (17:14 -0700)]
FastMixer compensates for SRC jitter in audio HAL

Force a sleep if audio HAL write() returns too quickly.

Bug: 6661275
Change-Id: Ie810f5a087b89fb24dab229a0e500af1dee53571

12 years agoam fc42e16d: Merge "Fix track selection code in AwesomePlayer" into jb-dev
James Dong [Fri, 15 Jun 2012 17:16:34 +0000 (10:16 -0700)]
am fc42e16d: Merge "Fix track selection code in AwesomePlayer" into jb-dev

* commit 'fc42e16d96b7e1b18b573f40df1319c56f8ab933':
  Fix track selection code in AwesomePlayer

12 years agoMerge "Fix track selection code in AwesomePlayer" into jb-dev
James Dong [Fri, 15 Jun 2012 17:14:04 +0000 (10:14 -0700)]
Merge "Fix track selection code in AwesomePlayer" into jb-dev

12 years agoam 0a14c4ce: Make CPU frequency statistics optional
Glenn Kasten [Fri, 15 Jun 2012 15:23:44 +0000 (08:23 -0700)]
am 0a14c4ce: Make CPU frequency statistics optional

* commit '0a14c4ce1a41bc09eb7855fa531a3af629a69139':
  Make CPU frequency statistics optional

12 years agoFix track selection code in AwesomePlayer
Insun Kang [Fri, 15 Jun 2012 04:02:45 +0000 (13:02 +0900)]
Fix track selection code in AwesomePlayer

Change-Id: I1c5a2f13f562a0ba253a93b4346cbe2c60099ddd
related-to-bug: 6671977

12 years agoMake CPU frequency statistics optional
Glenn Kasten [Wed, 13 Jun 2012 21:58:49 +0000 (14:58 -0700)]
Make CPU frequency statistics optional

Certain CPUs with dynamic cluster swapping and hotplug
don't report CPU frequency accurately.  The file descriptors
used to read the frequency become stale and report bogus data.
So make this feature a build time option for debugging only.
This will also improve performance of the fast mixer loop.

Change-Id: I602f81ec3281a37992769208be08084ed1469e8c

12 years agoam 362ebcbf: DO NOT MERGE Remove log spam for fast track denied
Glenn Kasten [Fri, 15 Jun 2012 00:26:44 +0000 (17:26 -0700)]
am 362ebcbf: DO NOT MERGE Remove log spam for fast track denied

* commit '362ebcbf100f7fccd37551c77e67c4faa7241b63':
  DO NOT MERGE Remove log spam for fast track denied

12 years agoDO NOT MERGE Remove log spam for fast track denied
Glenn Kasten [Thu, 24 May 2012 15:44:00 +0000 (08:44 -0700)]
DO NOT MERGE Remove log spam for fast track denied

Bug: 6531054
Change-Id: Iedf58e810a157aae88b5900da27c81054c437058

12 years agoam 7b4de101: Merge "stagefright: set scaling mode for blank frames" into jb-dev
Jamie Gennis [Thu, 14 Jun 2012 20:10:42 +0000 (13:10 -0700)]
am 7b4de101: Merge "stagefright: set scaling mode for blank frames" into jb-dev

* commit '7b4de10199f3a1803350e9a65baa671874d9c2f1':
  stagefright: set scaling mode for blank frames

12 years agoMerge "stagefright: set scaling mode for blank frames" into jb-dev
Jamie Gennis [Thu, 14 Jun 2012 20:08:08 +0000 (13:08 -0700)]
Merge "stagefright: set scaling mode for blank frames" into jb-dev

12 years agoam 192cbbad: Fix audio track pause.
Eric Laurent [Thu, 14 Jun 2012 19:12:52 +0000 (12:12 -0700)]
am 192cbbad: Fix audio track pause.

* commit '192cbbad773979a6fe3b5a0c223356de4fc3309c':
  Fix audio track pause.

12 years agostagefright: set scaling mode for blank frames
Jamie Gennis [Thu, 14 Jun 2012 02:13:16 +0000 (19:13 -0700)]
stagefright: set scaling mode for blank frames

This change sets the ANativeWindow scaling mode before pushing the blank frames
during decoder tear down.

Bug: 6603254
Change-Id: Ic64011645e2d3671b4a8d302ac7f39e6fd3affcd

12 years agoFix audio track pause.
Eric Laurent [Wed, 13 Jun 2012 15:38:36 +0000 (08:38 -0700)]
Fix audio track pause.

AudioTrack::pause() should signal the control block condition
to release threads waiting for available buffers in obtainBuffer().
Otherwise the behavior relies on the timout on the condition
or the fact that audioflinger will mix a new audio buffer while executing
the pause.

Bug 6653769.

Change-Id: I5f8f73c471fe306070f30b814f32fd4b4dc1d575

12 years agoam 82233b4b: Merge "Better handle invalid duration in Xing/VBRI tags" into jb-dev
Marco Nelissen [Thu, 14 Jun 2012 16:54:09 +0000 (09:54 -0700)]
am 82233b4b: Merge "Better handle invalid duration in Xing/VBRI tags" into jb-dev

* commit '82233b4b5bee06f11618a558ae465abbfdedfd6e':
  Better handle invalid duration in Xing/VBRI tags

12 years agoMerge "Better handle invalid duration in Xing/VBRI tags" into jb-dev
Marco Nelissen [Thu, 14 Jun 2012 16:51:49 +0000 (09:51 -0700)]
Merge "Better handle invalid duration in Xing/VBRI tags" into jb-dev

12 years agoam e6adde40: Merge "Shorten dumpsys media.audio_flinger output" into jb-dev
Glenn Kasten [Thu, 14 Jun 2012 15:43:58 +0000 (08:43 -0700)]
am e6adde40: Merge "Shorten dumpsys media.audio_flinger output" into jb-dev

* commit 'e6adde403480333b613452ae0df442df9da4a5b2':
  Shorten dumpsys media.audio_flinger output

12 years agoam 49dd5cf3: Merge "Log track name on obtain/releaseBuffer warnings" into jb-dev
Glenn Kasten [Thu, 14 Jun 2012 15:43:56 +0000 (08:43 -0700)]
am 49dd5cf3: Merge "Log track name on obtain/releaseBuffer warnings" into jb-dev

* commit '49dd5cf3469cd755321f8ec10013ad4fcfaf723d':
  Log track name on obtain/releaseBuffer warnings

12 years agoMerge "Shorten dumpsys media.audio_flinger output" into jb-dev
Glenn Kasten [Thu, 14 Jun 2012 15:41:54 +0000 (08:41 -0700)]
Merge "Shorten dumpsys media.audio_flinger output" into jb-dev

12 years agoMerge "Log track name on obtain/releaseBuffer warnings" into jb-dev
Glenn Kasten [Thu, 14 Jun 2012 15:41:38 +0000 (08:41 -0700)]
Merge "Log track name on obtain/releaseBuffer warnings" into jb-dev

12 years agomerge in jb-release history after reset to jb-dev
The Android Automerger [Thu, 14 Jun 2012 14:01:45 +0000 (07:01 -0700)]
merge in jb-release history after reset to jb-dev

12 years agoam e20435de: Merge "Add NOTICE and MODULE_LICENSE_APACH2 to libs build under /framewo...
James Dong [Thu, 14 Jun 2012 03:51:33 +0000 (20:51 -0700)]
am e20435de: Merge "Add NOTICE and MODULE_LICENSE_APACH2 to libs build under /frameworks/av/" into jb-dev

* commit 'e20435de10f63af812da9d4f801b61a8817733a0':
  Add NOTICE and MODULE_LICENSE_APACH2 to libs build under /frameworks/av/

12 years agoMerge "Add NOTICE and MODULE_LICENSE_APACH2 to libs build under /frameworks/av/"...
James Dong [Thu, 14 Jun 2012 03:49:34 +0000 (20:49 -0700)]
Merge "Add NOTICE and MODULE_LICENSE_APACH2 to libs build under /frameworks/av/" into jb-dev