| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* This patch introduces a new API which allows applications to
query the state of the audio effects system, and receive
callbacks with the necessary information to attach effects
to any stream.
* In the future, this may come as part of the AudioPort system,
but since that's an active area of development by Google, we
will dodge it for now.
* The policy now simply keeps a refcounted list of objects which
hold various bits of stream metadata. Callbacks are sent on
stream open/close to applications which might be listening
for them.
Change-Id: I2d554d36e1378f4eb7b276010a3bfe8345c22ecd
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* This has been rearchitected in a better way, as this
feature turns out to be more difficult than it seems.
* Reverting all of this stuff and rolling it into a single commit.
This reverts commit c27a16c33c78a36482336a16199b1b8be794cea4.
This reverts commit 32ef0556ae58ff6b7c6fe6fb0a17d3ff7f01de31.
This reverts commit 489c9fb62f02e1d23d6d6c89b22f7d19c596e65e.
This reverts commit a4123803d0a0e9e0c69faa4207d357cc74a65d58.
This reverts commit e13b58b988ab642d4ae5ca6d0a89013510714956.
This reverts commit 47f8c7303c9e2054f1492b02b6c7472385c52dc9.
This reverts commit 0479d7c79a7fd6f112e8dc7e45c009cf6602dbaa.
Change-Id: Iaed9f198d806aa414c95960713e8187c98db248b
|
|
|
|
| |
Change-Id: Ifba4025532debc50c7931d08a9bcd1593d9b1721
|
|\
| |
| |
| |
| | |
Ticket: RM-234
Android 6.0.1 release 30
|
| |
| |
| |
| |
| |
| |
| | |
to prevent a cleared service binder from being used.
Bug: 26040840
Change-Id: Ifb5483c55b172d3553deb80dbe27f2204b86ecdb
|
| |
| |
| |
| |
| |
| |
| | |
Previous fix accidentally only fixed the fragmented read case.
Bug: 27208621
Change-Id: Ie16f1920b84c8aba613842659238fcd5925694ad
|
| |
| |
| |
| |
| | |
Bug: 26914474
Change-Id: Ie1a86e86d78058d041149fe599a4996e7f8185cf
|
| |
| |
| |
| |
| |
| | |
We moved the safteynet logging into storeMetaDataInBuffers_l
Bug: 26324358
Change-Id: I2171742b53192ebb71d546bcb8970bb3c68f366f
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Uninitialized MetadataBufferType pointer in case:
SET_INPUT_SURFACE
STORE_META_DATA_IN_BUFFERS
CREATE_INPUT_SURFACE
Fix: initialize them to kMetadataBufferTypeInvalid
Bug: 26324358
Change-Id: Ib3bcac9bedc98bc65efa39f67cdbffbd90b374c9
|
| |
| |
| |
| |
| | |
bug: 26323455
Change-Id: I25bb30d3666ab38d5150496375ed2f55ecb23ba8
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This can lead to local code execution in media server.
Fix initializes the pointer and checks the error conditions before
returning
Bug: 26403627
Change-Id: I7fa90682060148448dba01d6acbe3471d1ddb500
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* Since we're not using the policy directly to auto-attach,
let's clean it up and remove the unused stuff.
* Also fixes notifications for session release.
This reverts commit 47f8c7303c9e2054f1492b02b6c7472385c52dc9.
Change-Id: Ibe65f427773c6ef012dde4f289d10e4089c094ea
|
| |
| |
| |
| | |
Change-Id: Id6f24195721a48dc5a0d3a7ff7fbee2bd8614128
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Change Id0af9c8aed1a983095275ac03f7f59abc31594cc
using boot time for video recording start time
as default, but it is specific on devices using
qcom camera HAL3 and it breaks all devices
using camera HAL1.
Fix by changing the default behaviour when
property media.camera.ts.monotonic is not set.
And newer devices such as 8996 might need to setprop
media.camera.ts.monotonic 0 in order to work
properly when using camera HAL3.
Change-Id: Iee07ed1147713a6397fff43e3bbd4f4ed08b56b7
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
If track creation fails on server side, mState/mActive is left
unitialized. After this a crash can happen when track stop()
is called because of a reference to a NULL object.
Initialize mState/mActive to specific value in initialized list
of track constructor.
Change-Id: If8c3611e3229c0c1b14b81285e07b9357fec7658
CRs-Fixed: 992608
|
| |
| |
| |
| |
| |
| | |
Fix invalid type in error log's format string.
Change-Id: I26aabfcdc1a03c0a7d9cd10fb3b39a660a9be03a
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
- If flush is called on both A/V tracks and video
decoder is not yet created, In offload case the
first audio buffer will get flush and audio playback
starts from timestamp of second buffer, where as
video playback starts from zero leads to AV sync
issue in start and cause framedrops.
- Call seek to current position to align the Audio
and Video tracks. Seek to zero if playback has not
yet started.
Change-Id: I4f75d0aa908714ad292f348fa8bf7aa5f837f8d7
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Camera HAL3 uses boot time for buffer timestamp, rather
than system monotonic time. This leads to issues as
framework uses system monotonic time as reference start
time for timestamp adjustment.
Add change to use boot time for reference start time.
CRs-Fixed: 946735
Change-Id: Id0af9c8aed1a983095275ac03f7f59abc31594cc
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
- AudioSink is reopened on AudioTearDown event as well as on
decoder output format change. These events are from different
contexts, independent of each other and can occur concurrently,
doing set of thread-unsafe operations on audioSink, which
sometimes may result in fatal error.
- Any request to openSink, after teardown event is received but
before its completed, is redundant since once teardown is
completed sink is anyway going to be opened again. Also
ignoring sink reopen while in teardown would avoid concurrent
processing of thread-unsafe operations on audioSink.
- Update nuplayer to prevent reopening of sink while teardown is
still going on.
CRs-Fixed: 978013
Change-Id: Ieb3825f133c2099e07d6e92ef101dd5156af8a20
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
In non-offload usecase, mAnchorTmeMediaUs is updated
for every audio FillBufferDone. Added the same logic
for offload case, otherwise video updates max media
duration as ZERO, for clips which have small audio
data, and all the buffers are sent to DSP before first
video buffer is processed at NuPlayerRenderer level
Change-Id: Ib6e0f251207c87ad1525041d07187cef91f4dda5
|
| |
| |
| |
| |
| |
| | |
Fix compilation issues that appear when enabling LOG_NDEBUG.
Change-Id: I87e9e5ac66157759dd6f521fab0dd346089a011a
|
| |
| |
| |
| |
| |
| |
| |
| | |
LOG_NDEBUG is inherited indirectly from both MyHandler.h & Log.h if
no LOG_NDEBUG is defined locally within AVMediaServiceUtils. Add a
local definition to keep debug usage for this file straightforward.
Change-Id: I878cdc747e8349e937b13dd27c738d45f99dee0a
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Without this patch, writing B-frame to MediaMuxer will be blocked and
cannot return a result. This may cause an endless progress bar shown in
gallery when triming or muting some videos.
Change-Id: Ib9f902a57da470c0357ec68aab00a35fbb26fb77
CRs-Fixed: 521005
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
ZSL mode is not working on HAL3 due to camera
backend failures. Run HAL3 forcefully in Non zsl
mode till the problem in camera backend fixed.
This change will be reverted once the backend issue
is fixed.
CRs-Fixed: 956182
Change-Id: I1dfcda8eb0fe5cd2d9be6cb2e103f5fd84955ecb
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Configure QTI resampler as 32 bit input format to avoid
losing precision.
audioflinger: add channel count check to use QTI resampler
Change-Id: I8f76dd82b72a0dd8b77343e77e0d0545e1be2114
CRs-Fixed: 940287
Change-Id: I7e1b8582cd6bb106ab0bd25f9bc1bd9e4092318b
|
| |
| |
| |
| |
| |
| | |
- Remove Virtual for SetDataSource()
Change-Id: Ie8502297376f6663a38f95195ff8329b7a42dcb9
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Make RTSPSource to accept some video damaged access units to enhance
fault tolerance.
Disable this change as default. add the property to enable it such as:
"adb shell setprop rtsp.video.keep-damaged-au video/3gpp" for H263
"adb shell setprop rtsp.video.keep-damaged-au video/avc" for H264
Change-Id: I3b7fb4b098aba5daf149cf36dab7e9380c6d2f69
|
| |
| |
| |
| |
| |
| | |
Defer MPEG4 muxer creation to AVFactory.
Change-Id: If0918be77ab7f8d82c78203f371df789e3cc29b8
|
| |
| |
| |
| |
| |
| |
| |
| | |
Current makefile compiles the resampler only for 64 bit environment.
Allow compilation for 32 bit environment as well.
CRs-Fixed: 735776
Change-Id: I626725551af73fc5ea98a7dbf87cacea9dbbc0ef
|
| |
| |
| |
| |
| |
| | |
Add support for enhancements in QSSP instrumentation code.
Change-Id: I00aa0b2bee5044d91eeeddaa43d3d28518a5ba93
|
| |
| |
| |
| |
| |
| |
| | |
- Extend onResume in-order to implement the Audio Voice Concurrency
changes.
Change-Id: I2b02f644df23f35b6d37d07a99d74ad56470b142
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Adds support for the HAL_PIXEL_FORMAT_RGBX_8888 format used by
mediatek's hardware.
Fixes several CTS tests where interaction between hardware and software
encoders and decoders is tested.
Change-Id: Ie3267c7d34706c49ad415ebc04ad05f2f2ef3506
|
| |
| |
| |
| |
| |
| |
| | |
Return on MTK hardware, the HAL does this internally and duplicating it
here causes an infinite loop.
Change-Id: I5bdb925ddb49980747b58dfae4543f812cef4c7d
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* The edge cases, ZOMG!
* Instead of relying on effects to be automatically attached, let's
just always notify userspace and send a bit more information. This
lets the application decide if effects should be attached rather
than relying on a hard-coded configuration file.
* Perform the setup in getOutputForAttr, but do it on the command
thread so we don't block the client.
OPO-593
Change-Id: I3900b349f2e895d51fa3a3dcc2de0c4bdf6dbc08
|
| |
| |
| |
| |
| |
| |
| | |
* Because we get a hellish firestorm during offload fallback, and
our effects library really dislikes this.
Change-Id: I22adb55d04502eedf03ce22f35f8b9f61de1c7b0
|
|\ \
| | |
| | |
| | |
| | | |
Ticket: CYNGNOS-2213
Android 6.0.1 release 24
|
| |\ \ |
|
| | |\ \
| | |/ /
| |/| | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
to prevent a cleared service binder from being used.
Bug: 26040840
Change-Id: Ifb5483c55b172d3553deb80dbe27f2204b86ecdb
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Previous fix accidentally only fixed the fragmented read case.
Bug: 27208621
Change-Id: Ie16f1920b84c8aba613842659238fcd5925694ad
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Bug: 26914474
Change-Id: Ie1a86e86d78058d041149fe599a4996e7f8185cf
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We moved the safteynet logging into storeMetaDataInBuffers_l
Bug: 26324358
Change-Id: I2171742b53192ebb71d546bcb8970bb3c68f366f
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Uninitialized MetadataBufferType pointer in case:
SET_INPUT_SURFACE
STORE_META_DATA_IN_BUFFERS
CREATE_INPUT_SURFACE
Fix: initialize them to kMetadataBufferTypeInvalid
Bug: 26324358
Change-Id: Ib3bcac9bedc98bc65efa39f67cdbffbd90b374c9
|
| | | |
| | | |
| | | |
| | | |
| | | | |
bug: 26323455
Change-Id: I25bb30d3666ab38d5150496375ed2f55ecb23ba8
|
| |/ /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This can lead to local code execution in media server.
Fix initializes the pointer and checks the error conditions before
returning
Bug: 26403627
Change-Id: I7fa90682060148448dba01d6acbe3471d1ddb500
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
- During offload (WH) to non-offload fallback (BT),
anchor time is getting updated incorrectly with
offload renderer position when it tries to start
sink using offload mode only.
- update anchor time post resume only when sink is
started successfully.
Change-Id: I2c2992afa90506e08389226964b0bd5a5be07529
CRs-Fixed: 978596
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Default standby delay for legacy direct is too aggressive for track
offloaded use case. Using offload standby delay instead, which allows
sufficient time for track transition before standby the output.
CRs-Fixed: 982822
Change-Id: Ifd0e39c5b3fe4a4af9f444dbbad21ef1e1a5edfa
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Delay standby for offload track use case is not suitable for direct
track in PCM format, because player torn down doesn't necessarily happen.
To avoid AP looping infinitely, allow standby right after direct track
in PCM format is paused.
CRs-Fixed: 986784
Change-Id: Id0026ac9891a12acc4affa446b2bc841559fcc28
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Clear the mMixerBuffer if frame is not ready before the FastMixer
starts processing. Otherwise it might result in large glitches on
output device due to junk data present in the mixbuffer.
Change-Id: I9f025234d4ac100f85d2ec67c06a8df4195bea98
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Break CC SEI parsing when payload size exceeds buffer size
to avoid a CHECK that have been seen in MTBF statistics.
Change-Id: Ifd97648678a935ac815dd616301d46f9bf583838
|