| caaa00f5ede96461e11de866ac15f5224442d1eb |
|
04-Aug-2015 |
Pavlin Radoslavov <pavlin@google.com> |
Fix a crash for a race condition during Bluetooth shutdown
This is same race condition observed in btif_media_task_aa_tx_flush_req()
because btif_media_task_aa_tx_flush_req() and btif_media_task_stop_aa_req()
are called back-to-bach.
This race condition is triggered when A2DP audio is streaming on shutdown:
"btif_a2dp_on_stopped() -> btif_media_task_stop_aa_req()" is called
to stop the particular audio stream, and this happens right after
the "cleanup() -> btif_a2dp_stop_media_task()" processing during
the shutdown of the Bluetooth stack.
Bug: 22700411
Change-Id: Ia0c98d44a108cf0f57731ac8129e4d76c9934542
/system/bt/btif/src/btif_media_task.c
|
| 196413f5ca088ef97866092f02bb7571d2a44390 |
|
22-Jul-2015 |
Andre Eisenbach <eisenbach@google.com> |
Fix TX queue overflow detection
- Prevent possible endless loop if |nb_frame| is too high
- Remove off-by-one calculation before sending frames
- Remove log spam and add better debug info
Bug: 22658329
Change-Id: I374ee980aec48763beb49b4f6f8b076124cadf40
/system/bt/btif/src/btif_media_task.c
|
| 197ee6856b43d8eea137a7f37798ee2a012ce518 |
|
22-Jul-2015 |
Pavlin Radoslavov <pavlin@google.com> |
Fix a crash for a race condition during Bluetooth shutdown
This race condition is triggered when A2DP audio is streaming on shutdown:
"btif_a2dp_on_stopped() -> btif_media_task_aa_tx_flush_req()" is called
to stop the particular audio stream, and this happens right after
the "cleanup() -> btif_a2dp_stop_media_task()" processing during
the shutdown of the Bluetooth stack.
Bug: 22602117
Change-Id: I5de6a8f15b6a2771dde2e299a5b60554063696a2
/system/bt/btif/src/btif_media_task.c
|
| 9c5d0727dd4cd09e44c032506c28ab4120ad3b88 |
|
27-Jul-2013 |
Anubhav Gupta <anubhavg@codeaurora.org> |
Override remote suspend flag on local play state changes
- Remote device may initiate AVDTP Suspend within 3s of AVRCP pause,
which sets remote suspend flag. Later on resumption of local playback
A2DP packets can not be processed further as Remote suspend flag
remains set. current change overrides remote suspend flag if local
suspend is triggered later.
- This change covers the usecase where play is pressed from DUT UI
before local suspend triggers in such scenarios current change takes
care of resetitng remote suspend flag on AVRCP play state change event.
- Moreover this change will drop A2DP packets when the setparameter
is called with A2dpSuspended=true and resumes when the mentioned parameter
is reset.
This change along with corresponding app change helps mitigating
issues reported due to A2Dp, HFP synchronizations and local and remote
device initiated AVDTP suspend handling.
Bug: 22028876
Change-Id: I3bb334373414170254744d9233cbd2e66219bfa0
/system/bt/btif/src/btif_media_task.c
|
| 1f293a50abf0bd37d149604c674ae5a6dbab7750 |
|
01-May-2015 |
Nitin Arora <niarora@codeaurora.org> |
Stop media task alarm when task is exiting
Bug: 20717705
Change-Id: I76e76cf33d7c9f4366f921daf2230e4f681098f7
/system/bt/btif/src/btif_media_task.c
|
| aea27ef29bd0be1d9faa5dcf94588107e9b6a51d |
|
03-Apr-2015 |
Rakesh Iyer <rni@google.com> |
Fix build in BT AV_SINK.
Change-Id: Ie20165ee69597e3287283322d14773ac3699c323
/system/bt/btif/src/btif_media_task.c
|
| e1a9e52ff59d36bd1bb5b7b3a02fafba6394edfe |
|
31-Mar-2015 |
Etan Cohen <etancohen@google.com> |
Merge commit 'f7f839985b3931682363d2ef3b7c5cae55a842ee' into merge
Change-Id: Iaaec1ea0bf3009b7e32a9a60f697631a3f56e889
|
| f8027005333c88a2f097cfd70d15c3d54c7764ae |
|
12-Mar-2015 |
Chris Manton <cmanton@google.com> |
Demote, cleanup and extend observed logging
/system/bt/btif/src/btif_media_task.c
|
| 0f9b91e150e153229235c163861198e23600e636 |
|
12-Mar-2015 |
Sharvil Nanavati <sharvil@google.com> |
Use fully qualified path for OSI includes.
/system/bt/btif/src/btif_media_task.c
|
| 71864f490d4dc31857df0b31924cd62a337e9f1a |
|
11-Mar-2015 |
Zach Johnson <zachoverflow@google.com> |
Add ability to set periodic alarms
Adds alarm_set_periodic so the alarm code can have more
contextual information when rescheduling alarms.
Problem: A2DP would stream for a few seconds and then
stop working.
Cause: The Java garbage collector. Bluedroid reaches out
to javaland to acquire and release the wake lock. Alarm was
always reaching out to get the wake lock when it scheduled a
short timeout. If GC kicked in during that call out to make
sure we have the wake lock, it could take more than 100ms to
get back us. That would screw over the alarm implementation
particularly for small 20ms timers.
So now if the wake lock was already acquired, we don't try to
reacquire it.
Cool. But we still have thrashing. Why? Because the alarm code
doesn't know the alarm is actually being used in a periodic way.
Here's what used to happen:
alarm expires
alarm is removed
reschedule
alarm callback is called
alarm callback sets the alarm again
alarm is added
reschedule
The problem is the first reschedule will get rid of the wake lock
if the next alarm is too far out or doesn't exist, meaning the next
reschedule needs to get the wake lock again.
With the extra periodicity information we can eliminate the
unnecessary intermediate reschedule, meaning no thrashing on the
wake lock. yay!
/system/bt/btif/src/btif_media_task.c
|
| e13fbfafe2105d264357ac41133f0ac0e054429c |
|
28-Dec-2014 |
Sharvil Nanavati <sharvil@google.com> |
Remove more GKI code.
This change removes all remaining GKI task remnants as well as
a few timer and buffer bits.
/system/bt/btif/src/btif_media_task.c
|
| 23e561a1f374557d6030b497337a1c96e9bb6aa4 |
|
03-Mar-2015 |
Sharvil Nanavati <sharvil@google.com> |
Remove dependency on GKI tasks from BTIF media task.
The previous code started a thread and posted a long-running function
to the thread's event loop. In essence, it created a thread and ran
a custom thread main that did its own event handling using GKI_wait.
The new code doesn't use GKI_wait or GKI_send_event at all. Instead,
it uses thread_post and the thread's reactor to enqueue and dispatch
events.
/system/bt/btif/src/btif_media_task.c
|
| c0a87fe60f79ddf41595a30e75a77f43c19cd323 |
|
02-Mar-2015 |
Sharvil Nanavati <sharvil@google.com> |
Remove GKI timers
/system/bt/btif/src/btif_media_task.c
|
| 753173dee0b92efdbdc0eb4d44338cd1a84dbdb2 |
|
26-Sep-2014 |
Chris Manton <cmanton@google.com> |
Replace bt media GKI task with thread_t
/system/bt/btif/src/btif_media_task.c
|
| 8d9d02862b1e00d27681846e3de9c4cf4f6634fc |
|
07-Sep-2014 |
Sharvil Nanavati <sharvil@google.com> |
GKI task entry points should have a consistent signature.
Previously, GKI task entry points had inconsistent signatures. For
example, we had:
int btif_media_task(void *p)
void btif_task(UINT32 params)
void btu_task (UINT32 param)
The single argument was universally ignored and the caller always
set it to 0. This change consolidates all of that and defines the
entry point as having 0 arity and no return value.
/system/bt/btif/src/btif_media_task.c
|
| 0d5de78354b5f7c819950fc5680a6cfd0052f78a |
|
07-Sep-2014 |
Sharvil Nanavati <sharvil@google.com> |
Eliminate unused 'stack' concept for GKI tasks.
This change also fixes the type for task names: should be a const
string, not a mutable INT8*.
/system/bt/btif/src/btif_media_task.c
|
| 860a9af89f8dbf8ef65785d5db960c1fa488d2ba |
|
27-Aug-2014 |
Chris Manton <cmanton@google.com> |
Replace mailboxes with fixed queues.
Also refactored btu_task a bit.
/system/bt/btif/src/btif_media_task.c
|
| fe7216ca12f91baae733e7c93063db73121af308 |
|
06-May-2014 |
Chris Manton <cmanton@google.com> |
Enforce GKI API buffer usage
Also add another API GKI_queue_length(BUFFER_Q *)
/system/bt/btif/src/btif_media_task.c
|
| fb17c122b09171492d99b208c1fdd6b76d2c431c |
|
16-Jan-2015 |
Andre Eisenbach <eisenbach@google.com> |
am 335bc32b: Merge "Relax max PCM frames per tick a bit" into lmp-mr1-dev
* commit '335bc32bc7dae893358bde20557927e6ed6ad58c':
Relax max PCM frames per tick a bit
|
| 26e80aad86fae4e468a330d64ee02291f3a1980e |
|
07-Jan-2015 |
Andre Eisenbach <eisenbach@google.com> |
Relax max PCM frames per tick a bit
Also made the value overwrite-able in bt_target.h.
Bug: 18913091
Change-Id: I64bbe453b35fdce6f127b3c23c0552b14d4ac49b
/system/bt/btif/src/btif_media_task.c
|
| 11cf5ccd5cd6dd1568bf69e4b6cbc27c95f5026d |
|
11-Dec-2014 |
Chih-Hung Hsieh <chh@google.com> |
resolved conflicts for merge of ed0bdd42 to lmp-mr1-dev-plus-aosp
Change-Id: Ife8b72d3688ed78a95ccb8d5e0b60f40e7fec9e9
|
| 104e3f238f500cacdb8afaccdc672dfc34eae7ee |
|
12-Nov-2014 |
Bernhard Rosenkränzer <Bernhard.Rosenkranzer@linaro.org> |
Cleanups needed to make clang happy
This patch removes a slew of unused variables and some unused functions,
fixes some incomplete initializations of structs and an sprintf
parameter mismatching the format string.
Change-Id: I35f22fdccc8350f885cd357d3685b869224fa433
Signed-off-by: Bernhard Rosenkränzer <Bernhard.Rosenkranzer@linaro.org>
/system/bt/btif/src/btif_media_task.c
|
| c5916e97384f07ec5a2e061e6af24f70107c74a2 |
|
08-Nov-2014 |
Andre Eisenbach <eisenbach@google.com> |
Remove overflow rate limiting when reading A2DP frames
A2DP now supports adjusting the number of A2DP frames sent per timer
tick to adjust for timer drift and missed media task ticks. When the
signal to the headset/speakers becomes interrupted, the number of
packets to be read can queue up and rate limiting will not allow for the
queue to be cleared.
The overflow mechanism introduced in commit 4aebca4 will cause the media
task to stop sending packets completely, which can underflow the jutter
buffer on the remote device and lead to audio drop-outs.
This patch removes the overflow mechanism and also adds code do discard
audio frames that could not be sent to the remote device (weak signal
etc) to allow the device to stay in sync and not build up audio delays.
Also added additional debug logging and changed the UIPC flush mechanism
to address an issue where reading byte by byte causes an endless flush
loop if remote UIPC producer writes data faster than the flush loop
consumes it.
Bug: 18326405
Change-Id: I9a424984806bb2a464877399804b3355b2c439c3
/system/bt/btif/src/btif_media_task.c
|
| 01d102b3b187ab0d24050c1e12f96e6ee1ce50be |
|
17-Oct-2014 |
Andre Eisenbach <eisenbach@google.com> |
A2DP: Lowering maximum PCM frame limit to avoid audio breaks
Some car kits (notably Renesas car kits used in Nissan/Infiniti models)
will throttle incoming A2DP frames if too many frames are sent at once.
This patch lowers the maximum number of frames sent per tick to not
overwhelm the remote device.
Bug: 17880920
Change-Id: I7bfbba625405c96159d1c94eebd8dc498f100daa
/system/bt/btif/src/btif_media_task.c
|
| f061e74b81863771b174d1717a6165034dbad8e5 |
|
09-Oct-2014 |
Mattias Agren <magren@broadcom.com> |
Added missing A2DP priority on java alarm service
Running GKI timer engine via JAVA alarm service was causing severe
timer preemption at times which affects A2DP media task timer. This in
turn was stressing the compensation logic towards UIPC resulting in
very bursty reads of up to 40 UIPC frames at a time. This led to
instabilities on the UIPC pipe and trickled upwards in audio framework.
Inclusion of java alarm thread in A2DP high priority thread list results
in a much smoother data flow for entire A2DP data path and improved
audio quality.
Bug: 17520043
Change-Id: I7cea6ec64aa5b973b5fff767389a8827e681619e
/system/bt/btif/src/btif_media_task.c
|
| ae13ec986e188d66eaad3c69bb916ea18c142fef |
|
03-Sep-2014 |
Andre Eisenbach <eisenbach@google.com> |
Avoid race condition in SBC encoder initialization
In the A2DP media task, btif_a2dp_encoder_update() relies on the fact
that btif_a2dp_encoder_init() is called beforehand. Otherwise division
by zero errors will occur.
Since _init() called after the audio stream is opened, and _update() is
called in resonse the the audio stream being opened, a race condition
could ensue where _update() was called before open.
This patch changes the A2DP stream start sequence to ensure the SBC
encoder is initialized before the updated parameters are calculated.
Bug: 17339248
Change-Id: I6ed36a7fff0e4aa672a8866727030681eb4b3dfc
/system/bt/btif/src/btif_media_task.c
|
| a391bf1d4f4250e27b6f157a1a8d2493c36934dc |
|
09-Aug-2014 |
Andre Eisenbach <eisenbach@google.com> |
A2DP: Protect against division-by-zero error in bitpool calculation
If the encoder parameters have not been configured correctly before the
SBC encoder is initialized, a division-by-zero can occur resulting in a
SIGFPE.
Bug: 16902125
Change-Id: Ie4c6715b6b374a2d62fcfa0a1a99d40c689d0e2a
/system/bt/btif/src/btif_media_task.c
|
| 25067c0afc4d3ed95181b91801d3d638d9aa3f8f |
|
22-Jul-2014 |
Andre Eisenbach <eisenbach@google.com> |
Fix integer overflow in GKI timer calculation
Fix an integer overflow in the GKI timer elapsed time calculation that
would cause Bluetooth Stereo Audio (A2DP) playback to stop suddenly
after a period of INT32_MAX microseconds (about 35 minutes) of
continuous uptime of the phone/tablet.
Bug: 16412377
Change-Id: I28cd3155a23444e41b26b3f3b8424cc2a3c5bf91
/system/bt/btif/src/btif_media_task.c
|
| f13db8a534adf5281ee915c4b73b3141957ed17b |
|
12-Jul-2014 |
Andre Eisenbach <eisenbach@google.com> |
Calculate A2DP frames to send based on elapsed time
This patch is aimed at improving A2DP audio quality by sending the
correct number of frames based on the actual time elapes between "ticks"
in order to account for timer drift.
Change-Id: If5b7cde90f5eb10b926ad247f51cff0eb60bbff7
/system/bt/btif/src/btif_media_task.c
|
| 0317a401142e927be82a0259f54cfa10a686e0fe |
|
09-Jul-2014 |
Andre Eisenbach <eisenbach@google.com> |
Merge "Prevent A2DP bit pool calculation endless loops."
|
| 4e3593a9a1c596904a844286fbe704e969e7230e |
|
08-Jul-2014 |
Andre Eisenbach <eisenbach@google.com> |
Prevent A2DP bit pool calculation endless loops.
This prevents log spam and endless loops in the A2DP bit pool
calculation. Protecting against the endless loop with prevent battery
draining issues and phone unresponsiveness.
This patch addresses the symptom. A separate patch will be submitted
once root cause is determined to address the problem.
Change-Id: I045b724c1c1638551b9333249f39eee109fa9ba1
/system/bt/btif/src/btif_media_task.c
|
| 6718c6d9fcfe373a679a529aca003d32062f81dc |
|
02-Jul-2014 |
lungtsai_lin <lungtsai_lin@htc.com> |
Bluetooth: Rename BTA_AVK_INCLUDED flag to BTA_AV_SINK_INCLUDED (bt_target.h)
Now OEM could enable/disable A2DP SINK role
Change-Id: I62cf17b0397a86c5f886a2c1ca724ca77a7f9bd4
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
/system/bt/btif/src/btif_media_task.c
|
| afa6e1abbedaad8fe854b0f43999b8aeb801af91 |
|
28-Jun-2014 |
Matthew Xie <mattx@google.com> |
resolved conflicts for merge of e8c3d75b to master
Change-Id: I78ef69c4d54a36243620ae14296d3507e3339567
|
| 8e50f047de2999e89b6a19ec9bf513017fcde3f9 |
|
28-Jun-2014 |
Matthew Xie <mattx@google.com> |
resolved conflicts for merge of 6449e497 to master
Change-Id: I01ef7ea6c4ce49c9d775b355fb5b082b25ffbc39
|
| e8c3d75b75493911ebf0f99c83676359657178f7 |
|
04-May-2014 |
Sharvil Nanavati <sharvil@google.com> |
Logging cleanup: BTIF and APPL.
Change-Id: I5b1214642bbb4b9aecc0fd2c899a6ec2c9793286
/system/bt/btif/src/btif_media_task.c
|
| 3e8a242fcbeebea2857fa964ca48624d8433333e |
|
23-May-2014 |
Mike Lockwood <lockwood@google.com> |
More work on A2DP Sink:
Output audio data through A2DP audio HAL rather than playing directly
to native AudioTrack API.
Add separate HAL interface for A2DP sink
Change-Id: I6c6cb6088c350e104b4a7354f328b29c7178e295
/system/bt/btif/src/btif_media_task.c
|
| ce24765fe7620c34e8d88ed4f826c8a6917582b2 |
|
20-Mar-2014 |
Hemant Gupta <hemantg@codeaurora.org> |
SBC Decoder for A2DP SINK
Change-Id: Ic8d1dd81d1788324ec140fe0d1379dfff33deb54
/system/bt/btif/src/btif_media_task.c
|
| f7dd9f5779680da37dc89e5df2b26d436487818c |
|
24-Oct-2013 |
Hemant Gupta <hemantg@codeaurora.org> |
Add A2DP Sink Support
Change-Id: I9affefdd2d00597545e49c593ef3bddb110d4c9c
/system/bt/btif/src/btif_media_task.c
|
| 1a3006f317b1b493dde1b80c4debbcf1bf5f6727 |
|
04-May-2014 |
Sharvil Nanavati <sharvil@google.com> |
Eliminate dead code (helped by compiler warnings).
I've removed all functions that the compiler decided were unused.
Change-Id: I0cbddf874ebd1586abe29b80e8d6390680679542
/system/bt/btif/src/btif_media_task.c
|
| e7c4f99f1ca3da0ac46aa7425a37b383177c699f |
|
05-Sep-2013 |
Hemant Gupta <hemantg@codeaurora.org> |
Bluetooth: On Suspend ignore media task timer event
On Recieving Suspend confirmation from remote device,
Media task timer reload and tx_timer is made false,
hence media task will not be stopped at that instant.
Media task timer expires causing buffer to be allocated.
Changes done to ensure buffer is allocated only if tx_timer
is true.
Change-Id: I971a82cccc64ba033475ad532b096862968de942
/system/bt/btif/src/btif_media_task.c
|
| 309243751678f395e6f7323e45433e5476dc17ef |
|
23-May-2014 |
Mike Lockwood <lockwood@google.com> |
More work on A2DP Sink:
Output audio data through A2DP audio HAL rather than playing directly
to native AudioTrack API.
Add separate HAL interface for A2DP sink
Change-Id: I6c6cb6088c350e104b4a7354f328b29c7178e295
/system/bt/btif/src/btif_media_task.c
|
| 8e08ad12eae66877f9725ca2d66fc69fe6762340 |
|
20-Mar-2014 |
Hemant Gupta <hemantg@codeaurora.org> |
SBC Decoder for A2DP SINK
Change-Id: Ic8d1dd81d1788324ec140fe0d1379dfff33deb54
/system/bt/btif/src/btif_media_task.c
|
| cc35559160745c60b42ae182002a59337d844d09 |
|
24-Oct-2013 |
Hemant Gupta <hemantg@codeaurora.org> |
Add A2DP Sink Support
Change-Id: I9affefdd2d00597545e49c593ef3bddb110d4c9c
/system/bt/btif/src/btif_media_task.c
|
| 6449e497dfe4d500810c3e7b4d7d056027c33a0e |
|
06-Jun-2014 |
Sharvil Nanavati <sharvil@google.com> |
Eliminate more dead code in GKI.
Change-Id: I4c0000445535e5300b83a18f5f6ca45f1ef797da
/system/bt/btif/src/btif_media_task.c
|
| 740def526e757f5ea8f30d9d28d1e7a668dadc42 |
|
04-May-2014 |
Sharvil Nanavati <sharvil@google.com> |
Eliminate dead code (helped by compiler warnings).
I've removed all functions that the compiler decided were unused.
Change-Id: I0cbddf874ebd1586abe29b80e8d6390680679542
/system/bt/btif/src/btif_media_task.c
|
| 7e4c60ce7757d935901f53d1f21ecc696aa89979 |
|
05-Sep-2013 |
Hemant Gupta <hemantg@codeaurora.org> |
Bluetooth: On Suspend ignore media task timer event
On Recieving Suspend confirmation from remote device,
Media task timer reload and tx_timer is made false,
hence media task will not be stopped at that instant.
Media task timer expires causing buffer to be allocated.
Changes done to ensure buffer is allocated only if tx_timer
is true.
Change-Id: I971a82cccc64ba033475ad532b096862968de942
/system/bt/btif/src/btif_media_task.c
|
| 54db2c249efa91082c5a167a924a527ec73ff38d |
|
13-Apr-2014 |
Sharvil Nanavati <sharvil@google.com> |
Remove executable bit on source files.
Change-Id: Id456478aea626f57fed3479346e8fb4dbd0fafa6
/system/bt/btif/src/btif_media_task.c
|
| d2ccbbb73c7851d2fa28dc212d2fffc0ad4e5d50 |
|
13-Apr-2014 |
Sharvil Nanavati <sharvil@google.com> |
Remove executable bit on source files.
Change-Id: Id456478aea626f57fed3479346e8fb4dbd0fafa6
/system/bt/btif/src/btif_media_task.c
|
| 5cd8bff2dd0337cb52bf48f312e3d2d55a8882fb |
|
01-Feb-2014 |
Mike J. Chen <mjchen@google.com> |
Major warnings cleanup
Mostly fixing unused parameter warnings.
A few other warnings also fixed like possible use of
unitialized variables (no real issue found, just compiler couldn't
follow the path), signed vs unsigned warning.
Also fixed some typos, indent issues, removal of dead code, etc.
Change-Id: I95eb887aefc4d559d7921f71a0af5f3bfb01ac01
Signed-off-by: Mike J. Chen <mjchen@google.com>
/system/bt/btif/src/btif_media_task.c
|
| 4aebca4662c280fdc32e0ad3705cae064bfa068d |
|
19-Sep-2013 |
Zhihai Xu <zhihaixu@google.com> |
implement flow control to avoid dropping packets in btif_media_task.
implment flow conrol to save CPU usage and avoid dropping
packet in btif_media_task when overflow and cpu preemption happen.
when the packet count in the queue is above high watermark, we will
stopping adding packets, when the packet count in the queue is below
low watermark, we will start adding packets again.
bug:8252054
Change-Id: I3b7795dffbe5f2ae455e239a0234533f6cbb4403
/system/bt/btif/src/btif_media_task.c
|
| 01c686ce0051433d75605ff9ca4c3b7388ce3b6b |
|
16-Sep-2013 |
Zhihai Xu <zhihaixu@google.com> |
should return correct length and use shorter poll time for ad2p data
the default poll timeout(100ms) is too long for a2dp data read,
because the a2dp timer pertiod to read data is 20ms. The poll timeout
should be far less than a2dp timer pertiod. also when poll timeout happen
for read, we should return the data length which we already read instead of 0.
bug:8252054
Change-Id: Ifa8587c6b7d127cbbbff76d9e20b630303b6379f
/system/bt/btif/src/btif_media_task.c
|
| 379743beded881127703334e1dda2cd327ec651d |
|
29-Sep-2013 |
Zhihai Xu <zhihaixu@google.com> |
Failure to start playback on A2DP sink after connection
This is what happen:
after Headset is connected, we call start_audio_datapath which will send AVDTP_Start command to Headset,
Headset reject it with bad state. Bluedroid stack will ack failure to start_audio_datapath.
The next time we write audio data to bluetooth, we will call start_audio_datapath again to send AVDTP_Start command to Headset
Headset reject it with bad state again. Bluedroid stack will ack failure to start_audio_datapath.
When the third time we call start_audio_datapath, right at that time we receive AVDTP_Start command from Headset.
Handle AVDTP_Start command and Handle start_audio_datapath are in two different threads.
Handle AVDTP_Start command is in btu_task thread.
Handle start_audio_datapath() is in btif_task thread.
We have race condition in this case
Because when btif_task processed BTIF_AV_START_STREAM_REQ_EVT(triggered by start_audio_datapath),
it don't know we receive AVDTP_Start command which is processed in btu_task.
btif_task will send a message BTA_AV_API_START_EVT to btu_task, which will be handled by bta_av_do_start.
AVDTP_start command from headset is handled by bta_av_start_ok.
bta_av_start_ok will send BTA_AV_START_EVT with suspending true to btif_task and send AVDTP_Suspend command
to headset to suspend the AVDTP for reconfiguration purpose.
in bta_av_do_start, we will check whether the AVDTP is already started, we will know the AVDTP is already start at this time because
bta_av_do_start is also running in btu_task. We will send BTA_AV_START_EVT with success to btif_task.
In the btif_task, BTA_AV_START_EVT will be processed by btif_av_state_opened_handler:
For the first BTA_AV_START_EVT with suspending true sent by bta_av_start_ok, it will ignore it:
if ((p_av->start.status == BTA_SUCCESS) && (p_av->start.suspending == TRUE))
return TRUE;
For the second BTA_AV_START_EVT with success sent by bta_av_do_start , it will ack success to start_audio_datapath, and change to
BTIF_AV_STATE_STARTED/BTAV_AUDIO_STATE_STARTED, after receive success ack from bluedroid stack, we will start send Audio data to bluetooth.
At last we received AVDTP_Suspend response accept from Headset, we will send BTA_AV_SUSPEND_EVT to btif_task, which will be
handled by btif_av_state_started_handler. It will call btif_a2dp_on_suspended and call audio_state_cb with new audio state
BTAV_AUDIO_STATE_STOPPED.
so The state between bluedroid stack and audio data path is out of sync.
The fix is to send failure message if we know we suspend AVDTP in bta_av_do_start,
also make sure we won't miss acknowledgement for pending start if we exit opened state,
to avoid audio data path dead lock.
bug:10953908
Change-Id: I1704839977324b7c4e234eb843cddf3719e10d2c
/system/bt/btif/src/btif_media_task.c
|
| b8ceaa484a09be14263d5f10d022a272f9602378 |
|
05-Apr-2013 |
Mattias Agren <magren@broadcom.com> |
Set default a2dp sbc encoding quality to high
* Increased bitpool to high quality setting (53). * Added rate quality
reduction to medium quality for basic rate connections * Enhanced
a2dp frame calculation ratio to smooth out frame distrubution * Removed
iop scaling patch.
Bug 8252054
Change-Id: I20725d0decfe5f820e1c03407889b6272e830aca
/system/bt/btif/src/btif_media_task.c
|
| 5fd74f06caab4324d0ba69df86ad5170c072ec64 |
|
05-Apr-2013 |
Mattias Agren <magren@broadcom.com> |
Stability fixes for a2dp hal control path
Ensure av statemachine is preventing a2dp hal to restart the stream
when not ready. This prevents ending up in a scenario where a2dp hal
locks up. Code cleanup.
bug 7002859
Change-Id: I3f3eeaab4ca185733b8b0042ed2cfa701f76a203
/system/bt/btif/src/btif_media_task.c
|
| 45f5f9077c773b9d4533fef13045b5ff77e32e46 |
|
08-Feb-2013 |
Ganesh Ganapathi Batta <ganeshg@broadcom.com> |
Resolved partial uipc channel read issue
When partial data is read the previously read size wasn't
compensated for. Ensure we don't transmit empty a2dp frames in case
of underrun
Bug 8026551
Change-Id: I4195ed798008378b93343bf7cd266c6e2f11b094
/system/bt/btif/src/btif_media_task.c
|
| 5738f83aeb59361a0a2eda2460113f6dc9194271 |
|
13-Dec-2012 |
The Android Open Source Project <initial-contribution@android.com> |
Snapshot cdeccf6fdd8c2d494ea2867cb37a025bf8879baf
Change-Id: Ia2de32ccb97a9641462c72363b0a8c4288f4f36d
/system/bt/btif/src/btif_media_task.c
|