| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
Bug: 30229821
CYNGNOS-3312
Test: fixing CL. Existing unit tests still pass.
Change-Id: I73d3fe413dd7590470c5b3b19569d6a7969c4dd8
(cherry picked from commit 0c39abc483ba980f7c417e57a0aa39139136c97a)
(cherry picked from commit fb2df5b44201c9af67f963c0495c23a8d640bbc6)
|
|
|
|
|
|
|
|
|
|
| |
RM-290
CYNGNOS-3303
Bug: 30229821
Change-Id: Iea1c4a21735e893aeded95b980044ec0861a7ea8
(cherry picked from commit 2fa52194ef64843c2908c69527384c6c2fcdbafa)
(cherry picked from commit 57ac66340ad488a17fc285b6fc2635cb7375d72b)
|
|
|
|
|
|
|
| |
Bug: 28173666
Ticket: CYNGNOS-3020
Change-Id: I055af37a721b20c5da0f1ec4b02f630dcd5aee02
(cherry picked from commit 57fd9637536d40ec8c40a6bed76a71471dab0f64)
|
|
|
|
|
| |
Change-Id: I60075568d1a7d96fc2bf753a8cfa8ac6cddeeac1
(cherry picked from commit 27c02b0f0f1b247e4f5ac0895e5839a8c90bdac6)
|
|
|
|
|
|
|
| |
* Fix surround sound recording build
* Fix build when compress voip is disabled
Change-Id: Ia8f7598809ae9b8ae3767cae3cefdb6d49a04029
|
|
|
|
|
|
|
|
|
|
| |
AudioPolicyManagerCustom is outputting garbage
for the following log and potentially crashing it.
AudioPolicyManagerCustom: offload disabled by av.streaming.offload.enable = ^[;¤¶
It should not use prop_value.
Change-Id: I5dc388d273c12a7c736faa50aeeab684f9600125
(cherry picked from commit 17067a92061757c88faeb3f2874e567501653673)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If the wired headset/headphone/line devices are handled by
a different backend than speaker/earpiece devices, the combo
devices such as speaker+headphones can be split into individual
devices and enabled/disabled independently.
Bug: 21581860.
Change-Id: Icdd962a4dc1da536fe89c4de2202e7383275603d
hal: fix backend check routine for combo device
If there is path change between combo device and general device that
shares backend, previous device should be disabled before the new device
is enabled. Otherwise, by the race conditio, the new device would be
disabled by previous combo device
For example, when the previous device is 'speaker-and-headphone' and the new
devcie is 'voice-headphones', there is possibility that
'voice-headphones' get mute by disabling 'speaker-and-headphone'.
BUG=23916345
Change-Id: I765b3c0b6576f22c4aca290846d58dd1c6dfb8c5
hal msm8974: fix logic in platform_check_backends_match().
Fix logic error in if() condition in commit 9af4b91
"hal: fix backend check routine for combo device"
BUG: 23916345
Change-Id: I587dd157ba34cad5f539d88e8161750f3e2b32c7
|
|
|
|
| |
Change-Id: I036a5bc8811dbee698ffe7e49fcf1452e5b74309
|
|
|
|
|
|
|
| |
This broke it
https://github.com/CyanogenMod/android_hardware_qcom_audio/commit/f17c0998351f248051645ab3ba21ba9ee57a264d
Change-Id: Idcce338f8491977ec1d8da8cb52c8709cbe5789e
|
|
|
|
|
|
| |
* This was mistakenly removed in a previous patchset. Fixit.
Change-Id: I34a415a1875ba5343072228ad5e172c2138c03e4
|
|
|
|
| |
Change-Id: I0664812a58022045cce6836c65e651fb6eb6445f
|
|
|
|
| |
Change-Id: If7e46636e4512ab4eb482ce729297d1f04396bff
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- setPhoneState checks call-state in audio hal to filter out
unworthy usage of this API.
- With M upgrade, change to update call-state with audio hal is
removed, instead permission checks are introduced to prevent
random usage of such APIs. But setPhoneState is still checking
for this call-state causing improper concurrency handling.
- Fix the issue by removing state check from setPhoneState.
- Remove set and get for parameter "in_call" from voice_extn.
Change-Id: I92fe10c617d6158976a0e07eed3275acb0aadd15
|
|
|
|
|
|
|
| |
Enable Wall and Werror and fix warnings in
HAL and policy manager
Change-Id: Ie93c8f933943875ac68e8318d777602ed864f531
|
|
|
|
|
|
|
| |
Add performance lock during device path bringup in
playback usecase to reduce cold start output latency.
Change-Id: I2ff217beedb270f057fd4ea577a7a317ee1f20e6
|
|
|
|
|
|
|
|
|
|
| |
Add support for AUDIO_DEVICE_OUT_LINE device. Treat it same
as Wired Headphone device w.r.t. routing and acdb device.
If the acdb or backend is different it can be updated using
audio_platform_info.xml.
Change-Id: I203f38b1629a837478da1ae304a67bb016b06fa7
CRs-Fixed: 929500
|
|
|
|
|
|
|
|
|
| |
Call handleIncallSonification in startSource and stopSource without
checking mRefCount to make sure ringtone mute/unmute is called for
multiple active stream.
Change-Id: I52290f9ae09dff3619ac3a36c8969da6c07853ec
CRs-Fixed: 972582
|
|
|
|
|
|
|
|
| |
Ensure that the audio device's active input is valid before
attempting to access the input source.
CRs-Fixed: 959009
Change-Id: If66cc1e3d5467bf7d03137fdbbffd724a57ecaf6
|
|\
| |
| |
| |
| |
| |
| | |
"LA.BF64.1.2.2-03340-8x94.0"
Issue-Id: SAMBAR-1290
Change-Id: I3a8cb0e2785a2a98a86871cfc5fe7d575ea86305
|
| |
| |
| |
| |
| |
| | |
Fix potential memory leaks in HAL
Change-Id: Ifd491cc185a7d1afa647f03639e57d1685d18803
|
|\|
| |
| |
| |
| |
| |
| | |
"LA.BF64.1.2.2-03240-8x94.0"
Issue-Id: SAMBAR-1280
Change-Id: Ia1481c2e8726c24c721e78df680ad8851f74f02b
|
| |\ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
- Increase primary output ref count before getting new device
- This prevents to use NULL device for FM playback.
Change-Id: Idd2b36dab1990cdc1c1f603c5bdfd53b53c5c311
CRs-Fixed: 957807
|
| |\| |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Delay update of FM volume to start of FM playback.
This avoids unnecessary set_parameter calls on primary output
whenever volume changes on primary output. Unnecessary calls
on primary output can cause breaks in audio since FastMixer
will be put to HOT_IDLE more often
CRs-Fixed: 948361
Change-Id: Id614d32df7f35ac93dd029a08b10e19f64ac631e
|
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| | |
- Issue: Volume burst is heard during device switch
(headphone and speaker) for FM playback
- Root cause: delayMs passed to CheckAndSetVolume is
not applied for FM playback
- Fix: Double delayMs and apply FM volume with this delay
Change-Id: Ifa0b214d16527973277a5c84b8a4f3fbd0760d07
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
APM calls a set_parameter on primary output with key fm_volume
whenever a checkAndSetVolume is called. This is done irrespective
of whether the volume values are same of different.
A side effect of doing so is that the FastMixer associated with
the MixerThread will be put into HOT_IDLE unnecessarily which in turn
can cause glitches/breaks in audio.
CRs-Fixed: 948361
Change-Id: I7fd8b7340e6c78ee6d00b41ccf679582338c011a
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
-For stopping the alarm tone played during in call,
handleIncallSonification must be called with second
parameter i.e. starting as false from setPhoneState
when the call is disconnected
-isInCall method retuns false,hence handleIncallSonification
is not getting called on disconnecting the call
Change-Id: Ia6398eb1b0a3cb117a16f9fd4cb3637f5ce17be8
|
|\|
| |
| |
| | |
"LA.BF64.1.2.2-02240-8x94.0"
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The current logic for identifying the track offload
session is using strlcat in wrong way, third parameter
is given the string length to be appended, but in fact
it should be the size of the destination buffer.
Due to this the string returned is missing the last
character, in case of "true", it contains "tru" due this
the session is not considered as a track offload by the
calling module. To fix this use sizeof of the destination
buffer instead of string length of string to be appended.
Change-Id: I3d456aad2ca49095a0dfe5883f4374e22a437256
CRs-Fixed: 946670
|
| |\ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
- adding both deepbuffer and direct pcm is causing issue where wrong
config is set to ADM
- add deepbuffer flag when fail to get direct output if deepbuffer
was forced
Change-Id: I3bd65002c7a8895d461df257111faf22101cbe30
|
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Ringtone mute is observed in LINE app MT call over BT headset.
VoIP direct output isn't returned due to incorrect flags update.
Primary output is returned even though the FLAG_DIRECT is set,
this in turn causing issues when renderer queries for getPosition(),
which is resulting in mute. Fix this by updating flags, so that
track gets VoIP direct output in this scenario.
Change-Id: I6d570ae6ed9e0a7506cb29be5b41687417b7d221
CRs-Fixed: 893558
|
| |
| |
| |
| |
| |
| |
| | |
Change-Id: If6fcfdef0b051f9015c127bdd7d28e3668e47435
(cherry picked from commit 4879ebb6d14558695044be0a22a89fc8b7ac24ec)
Change-Id: I5ecf6c792bf1d5c481879a38231acfdb2a3ccc3e
|
| |
| |
| |
| |
| |
| |
| |
| | |
If AUDIO_FEATURE_ENABLED_FLAC_OFFLOAD rather
than AUDIO_FEATURE_ENABLED_EXTN_FLAC_OFFLOAD is true,
we should enable COMPRESS_METADATA_NEEDED also.
Change-Id: I8f3b0cea2f59a9967d845133716dd7db5f3facb0
|
|\ \ |
|
| |\ \
| | |/
| |/|
| | | |
Change-Id: I2abc7812ea07ca2302c7c29cbf4e9cfd81997363
|
| |\ \ |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Fix condition for offload small buffer use by masking format with
AUDIO_FORMAT_MAIN_MASK before comparing with AUDIO_FORMAT_PCM_OFFLOAD
Change-Id: Ia81fdc6061fbbf8c203bdfb3074730c6436fdf73
CRs-Fixed: 938198
|
| |/ /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Support getparameter to retrieve if a
particular output is associated with a track
offloaded session
Change-Id: Ibcc986514f802e59c514fdaa8b2d7f53841047a2
CRs-Fixed: 933740
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
- in case of direct PCM, Direct flag is added explicitly which
is causing deep buffer flag not to be added for Music case.
- absence of deep buffer flag is causing low latency path to be
selected by default if direct pcm fails. Fastmixer path has
higher power footprint.
- to avoid low latency and allow stream music to use deep buffer path
check if direct flag is added explicitly, if so allow adding
deep buffer flag. This will allow selecting deep buffer output.
Change-Id: Icc1de9f4ac3ff7981d2325fbda4891d9b8476df7
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* PCM_OFFLOAD, EXTN_FORMATS, PROXY_DEVICE are all enabled by default in
the HAL/post_proc, make them enabled by default for the policy HAL as
well.
Change-Id: I88a90ec138854cc5153fc9ed72dd629a92895ad3
|
|\ \ \
| | |/
| |/|
| | | |
Change-Id: Idee04b3a7c4685a852b853b101ecfc14b4bd74f7
|
| |\ \
| | |/
| |/|
| | | |
Change-Id: I109171f75ce0d76cbe657651c4eecec3ce29c8d4
|
| |\ \ |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Voice call setup latency value is exceeding 500 msec. It is due to
primary output selecting handset instead of voice handset causing
the extra delay. Fix this by starting voice usecase first before
changing device for primary output.
Change-Id: I39ac8186b62d86b2047bf73860e7027d15df7fd2
|
| |\ \ \ |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
- Update FM device references to match latest
definitions to fix compilation of FM extention.
- Update FM feature flag for power optimization.
Change-Id: I17a9e240e552c4089d3e9c7a25e437fb5b5433f9
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Pops heard when MO call is initiated while music playback is in progress.
This is due to a new device is set up for the output even before the
leftover buffers were drained out on the previous device.
This change is to make sure to add proper delay while device routing,
which is equal to twice of Output descriptor latency.
CRs-Fixed: 906398
Change-Id: I3b8059b8966251e0e0cc979c968205dffebc14f0
|
| |/ / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Use device connection event against AUDIO_DEVICE_OUT_FM
to configure FM device loopback.
Update primary output to get routing and volume events.
Change-Id: I5ddce36581568d377a08ea5afe470bdff4349ec7
|