-Sound trigger HIDL interface supports loading of more
than 1MB sound models from 2.1 version. Update to use HIDL
version 2.1.
Change-Id: I564f598f41f20a01ba9e8dd69be125097a134f1c
Update microphone characteristics to closely match
the default microphone routings for different
sound devices.
Change-Id: I6716a9cd68515deb25957d6fe8290da08d088029
CRs-Fixed: 2264242
Fix the BT SampleRate from '16000' to 'KHZ_16'
in "mmap-playback bt-sco-wb".
Fix the path name from "mmap-buffer-playback hdmi"
to "mmap-playback hdmi".
CRs-Fixed: 2278484
Change-Id: I44ae1d02cb214555967e3f1f42b0cd022e5c4bfe
H/w binder reserves 1 MB by default for memory operations.
Customization is to only reserve usable memory
Change-Id: Ide08e72e4437da270d498ff30b3cc1640d39d963
Populate microphone characteristics for supported microphones on
platform. Map input sound devices to specific microphone ids.
Change-Id: Iddb1d5e2a7b77985290df7d84c0342aaf80278e1
FM Playback Audio routes to A2DP path when BT Headset is connected.
The property vendor.fm.a2dp.conc.disabled is not used in FM
app as it is using app loopback for speaker and WHS(with BT connected)
FM app takes care of BT concurrency of not enabling FM over A2DP ,
Hence removing FM and a2dp concurrency changes from audio
modules.
CRs-Fixed: 2272694
Change-Id: I551ca9b0a9e060e72a6523ce7677234f2da2874f
No Audio heard in FM Speaker when FM record and FM speaker are on simultaneously
With default values for maxActiveCount and maxOpenCount,
FM app is unable to record and play FM audio on speaker simultaneously
since it uses application loop back(AudioRecord->AudioTrack)
to route FM audio to non-default devices(Speaker)
Add attributes maxActiveCount="2" and maxOpenCount="2" to mixPort "primary input"
CRs-fixed: 2249758
Change-Id: I82a1cb0d5596486d531d5324b91f932f52b07f9e
VTS audio effects test cases are failing due
non availability of audio_effects.xml file.
Added new audio_effects.xml as per existing
audio_effects.conf
CRs-Fixed: 2225774
Change-Id: Ief5dc78bbefc3dd7d6eab65b731806aa975dbf1e
Keep_alive can now handle silence playback on multiple concurrent
devices such as primary output devices as well as HDMI
Change-Id: Icdf22c9d8d1264d463b23c8f54020e792840b73c
Signed-off-by: Md Mansoor Ahmed <mansoor@codeaurora.org>
Add support for USB_HEADSET device in audio policy configuration
files of all targets which support USB Tunnel Mode.
CRs-Fixed: 2150144
Change-Id: I14c124e1a5407d004143c977348637fd373a4126
-Remove default setting of sampling rate to 48K and stereo channel
for bt a2dp as support for multiple sampling rates and channels
over a2dp is added and being set from hal.
Change-Id: I8a40182af0d97b31b136a1928cbbe6f728eb950f
Remove the path of audio_dynamic_log.xml file as proprietary
file cannot be added in open source repo.
Change-Id: I72b0638e8d3a11036b1183c62d8c3d28758a583f
This reverts commit 1c57390b961d2eaee6817d91e9453832780261bf.
As per surround sound recording design reverting this
commit to remove built-in mic from primary inputs.
Change-Id: Ic0727a4c59d437045a2ef38cfe4ca2ea66ae21d2
-DTSHD passthrough is not working when drift correction is enabled.
-Drift correction enablement calls out_write(0) to start compress
session. However DTSHD iec61937 transmission info parsing fails due
to size zero write buffer.
-To fix this added
-Changes to call dts parser when first valid buffer is enabled.
-Changes to trigger select_device() to update backend configuration.
-Also added mixer path change to configure default HDMI backend
configuration
Change-Id: I68b2c79ff008a7d081efa475c8b9782072f5ce03
Playback of high sample rate multichannel clips fails when
repeat mode is set to repeat one. This is because the default
heap size allocated for AudioFlinger clients is 4MB which is insufficient.
Set the system property ro.af.client_heap_size_kbyte to 7KB. This property
is read in AudioFlinger and multiplied by 1024, so effectively the allocated
heap size becomes 7MB which is sufficient to play such clips in repeat mode.
CRs-Fixed: 2108978
Change-Id: I72f4bb9892465afb311bb269545a5fe926696b02
To address voice latency during device switch,
update IIR coefficients at one time at bootup
and store in kernel driver to handle it during
WCD power collapse and resume.
CRs-Fixed: 2052352
Change-Id: Ib49e03c139767aa548e5194b821e5875dab72471
Update default audio_policy_configuration.xml to remove configurations
that are not compatible with aosp system image and are causing vts failures.
Add custom audio policy configurations at /vendor/etc/audio/ to be picked
up by qti customized system image.
Change-Id: Iae1f4c72090ef5a0c6472083dce8b0065beb6851
VtsHalAudioV2_0Target fails to execute completely and results in
failure of target boot up after VTS execution as system server waits
indefinitely for audioserver to come up. This is due to fd leaks while
running the test cases in the module.
Disable dolby flag on all relevant targets and close open fds in
deinit of sound monitor to resolve fd leaks. This ensures successful
execution of VTS and proper target reboot after execution of VTS.
CRs-Fixed: 2067095
Change-Id: Id4c9b680017f591dec8df385280888cc893dca8a
Built-in Mic is the fallback input source if A2dp,SCO,wired
headset,usb headset/device are not connected. During the transient
period between an output device and input device connection, if
primary input profile is selected and none of the above devices
are available, APM selects built-in mic as the new device.
Current configuration does not support built-in mic in mix routes
for primary input resulting in premature input close. This leads
to AudioRecord track recreation. Timing issues have been observed
due to incorrect input attributes being set leading to start record
failures. One such case is device switch (handset to headset)during
incall record. Add built-in mic to avoid timing issues relating to
restoreRecord_l call flow.
CRs-Fixed: 2066905
Change-Id: I97cc5370a843ddfd14f3b43ba4435c2930b8e19a
Add the switches to enable fluence and select fluence type in the
platform xml. Also add the new mixer controls for WDSP QMIC and
TMIC, along with ADSP TMIC configurations.
Change-Id: I8b0c4e1be6ce79b9bb59c8b591b0491d874c7a2d
-During device switch in a voip call, some apps dont close the session
and instead change the mode to MODE_IN_NORMAL.
-This leads to the voip session to get routed to A2DP-Handset mic pair,
which is not a standard or supported pair.
-To avoid routing errors and only to handle this transitory state,
add compress-voip a2dp path.
Change-Id: I2fea6be63cfc67f527fa7ac506861d6b580cb6b4
Add fractional/native sample rates to applicable profiles
to enable configuration of ADM at those rates.
Without this end to end playback of native clips with
sample rate >= 88.2 kHz is not truly native.
Change-Id: I399d725c57769c730db6acdb03b4626a7aa1ce6f
Vendor created android properties are required to have
vendor/qti prefix. Update vendor specific audio system
property configs.
CRs-Fixed: 2034869
Change-Id: Ie7283cbe00eb0a0e80ba0b478ecf904ceb2f9280
With this addition to the xml, sound trigger hal will add the
custom config header to the opaque data received from the app when
it is running on ADSP.
Change-Id: I1234eeb55b555ca566bd4a2ed6cb0b69f7bbd522
Enable SoundMonitor in audio hal to handle ADSP SSR events,
do required cleanup and restore any active sessions gracefully.
Change-Id: I83270b7c12095e56a6677f41efe0f5618a27e0e5
- All qcom proprietary modules should use
/vendor path instead /system for
treble compatibility
- Modify configs and android makefiles accordingly
Change-Id: I1a4db8c39b5e7585d1a4d60584d5b1672c9599ac
Ramjee Singh
Linux Build Service Account
yidongh
Manisha Agarwal
Samyak Jain
Lakshman Chaluvaraju
Bharat Pawar
Md Mansoor Ahmed
Satya Krishna Pindiproli
Preetam Singh Ranawat
Linux Build Service Account
Revathi Uddaraju
Haynes Mathew George
Aniket Kumar Lata
Manish Dewangan
kunleiz
Laxminath Kasam
Dhananjay Kumar
Weiyin Jiang
Quinn Male
Ashish Jain
Naresh Tanniru