Age | Commit message (Collapse) | Author |
|
Conflict files:
ion_physical_heap.h
SBMerger: 603054162
Change-Id: I13b4dd2b4336af7ce7dc7b337b87600ada97feac
Signed-off-by: benmcgee <benmcgee@google.com>
|
|
This change ensures that the struct dma_heap that is passed in
the file descriptor via ioctl is created/maintained by aoc.
Bug: 313424730
Test: researcher's PoC doesn't print handle
Change-Id: I408872a66964bd4d03d3b09794b2eb626028b60b
Signed-off-by: Ismael Castro <iicastro@google.com>
(cherry picked from commit 7672a4aaa5bb92bc69337942e5fa047e4964ae62)
|
|
This change ensures that the struct dma_heap that is passed in
the file descriptor via ioctl is created/maintained by aoc.
Bug: 313424730
Test: researcher's PoC doesn't print handle
Change-Id: I408872a66964bd4d03d3b09794b2eb626028b60b
Signed-off-by: Ismael Castro <iicastro@google.com>
(cherry picked from commit 7672a4aaa5bb92bc69337942e5fa047e4964ae62)
|
|
Test: on device
Bug: 207711097
Change-Id: I69ce237be000ffb351b61ed33a846f6252971303
Signed-off-by: Alex Iacobucci <alexiacobucci@google.com>
|
|
Conflict files:
alsa/aoc_alsa.h
SBMerger: 571992243
Change-Id: Ie9d3c086b902757e1b57dab966cc99f1c4e4dd4b
Signed-off-by: benmcgee <benmcgee@google.com>
|
|
Support dynamic change PDM mask for normal recording
and voice uplink. In AoC with differnt projects.
PDM and microphone may not map with same order, so it
need to adjust mask order. Add control to adjust it.
Bug: 299035199
Change-Id: I6c425126f9151b7433034718abe231e084ecd555
Signed-off-by: yixuanjiang <yixuanjiang@google.com>
|
|
into android14-gs-pixel-5.15-24Q1
|
|
SBMerger: 571992243
Change-Id: I0a1bd23a196efab73a6f3f42238c240e8c812591
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
Test: on device
Bug: 309950738
Change-Id: I60c14d8d89847c0681ff285b1f4d26d37858be18
Signed-off-by: Alex Iacobucci <alexiacobucci@google.com>
|
|
SBMerger: 571992243
Change-Id: Id9aa6ac74b81d23ecd84a14be171b8e13ecce224
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
The NULL termination was wrong when triggered from ALSA, due to trigger_watchdog
passing strlen (without NULL) as the count. The explicit NULL termination here would
overwrite the last character, but strscpy is supposed to use the full size of the
destination buffer and handle NULL termination itself.
Bug: 309902598
Change-Id: I9cf91660a5fd85fdcc2589dafa5f3c8af4e0fdb8
Signed-off-by: Craig Dooley <dooleyc@google.com>
|
|
This CL adds the necessary entries in the aoc_chan driver to handle the
new channel that will be used by BT.
Bug: 308452948
Test: Validated locally.
Change-Id: I4227029dbdbaa56858d6a45b1970dc65e9599db6
Signed-off-by: Luis Delgado de Mendoza <luisddmg@google.com>
|
|
move spatial eanble control to mixer control from HAL
Bug: 278650628
Change-Id: Ic8b186235ce7ffe884b82a958faddd013fba7930
Signed-off-by: yixuanjiang <yixuanjiang@google.com>
|
|
Test: on device
Bug: 307383094
Change-Id: Id474c099fa8392e3379b7a2a8861bf508951352a
Signed-off-by: Alex Iacobucci <alexiacobucci@google.com>
|
|
SBMerger: 571992243
Change-Id: Ic9f339f81dd56440ad8e1cb834aca4aa2088607e
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
into android14-gs-pixel-5.15-24Q1
|
|
add build dependency for disaplay port audio, and fix buildifiler error
Glob pattern "Kbuild" has no wildcard ('*'). Constant patterns can be
error-prone, move the file outside the glob.
http://go/buildifier-warnings#constant-glob
Bug: 309871378
Test: build and no warning is output
Change-Id: Iee8dff7b7e0d38eb1e299186a7d599e570705c68
Signed-off-by: Robert Lee <lerobert@google.com>
|
|
It cause next probe fail after DEFER
Bug: 303681990
Change-Id: I40bc1808f3ad3ed5367af24747ead6a6ee851f0c
Signed-off-by: yixuanjiang <yixuanjiang@google.com>
|
|
SBMerger: 571992243
Change-Id: I518a30f4090c8398f1f29564423173b15d854e2a
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
SBMerger: 571992243
Change-Id: Ib7ad360a9e953f586b6d2993f653f3272d8c7d07
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
Bug: 294495501
Change-Id: Ie1929a75017cad29b7a09ba5ee0afb391fc885a2
Signed-off-by: vincenttew <vincenttew@google.com>
|
|
Test: on device
Bug: 307383094
Change-Id: I9a1949810a1cf8352d6708a1759053e36e1fb921
Signed-off-by: Alex Iacobucci <alexiacobucci@google.com>
|
|
into android14-gs-pixel-5.15-24Q1
|
|
Change the struct of aoc_moldule_parameters so that we can access the
up-to-date value all the times.
Bug: 303187479
Test: trigger AoC SSR multiple times by aocutil reset
Change-Id: I6f5d8cc27560049273cf32d08ce42964349eac5b
Signed-off-by: Guan-Yu Lin <guanyulin@google.com>
(cherry picked from commit 99149e030fd89ed64f6c610a951a0ded2fd7798e)
|
|
The original recovery flow didn't cover the case when usb connection
changes during AoC SSR. Instead of changing id to turn on/off host mode,
this fix uses fsm_reset so we can monitor id information during aoc
restart. (Additional patch for aligning implementation in kernel 6.1)
Bug: 297480242
Test: The following 6 tests passed:
1. AoC SSR during USB_ROLE_NONE.
2. AoC SSR during USB_ROLE_HOST.
3. AoC SSR during USB_ROLE_DEVICE.
4. Role changed during AoC SSR.
5. Restart pixel with USB_ROLE_HOST engaged.
6. Restart pixel with USB_ROLE_DEVICE engaged.
Change-Id: I9b0475a0d75181cbf09cea7dc2ae34511c0a0bc1
Signed-off-by: Guan-Yu Lin <guanyulin@google.com>
|
|
Conflict files:
Makefile
aoc.c
aoc.h
aoc_channel_dev.c
aoc_v1.c
SBMerger: 571992243
Signed-off-by: benmcgee <benmcgee@google.com>
Change-Id: Ib59c71c642e0887e16cec9367bb5ec4b0308db48
|
|
when coredump is corrupt, the value of num_sections
from ramdump_header may larger than the number of
sections. Out of range access will cause the KP.
Bug: 302929196
Change-Id: I423886cb8223617d84cf1b21231a914c70a30dd2
Signed-off-by: yixuanjiang <yixuanjiang@google.com>
(cherry picked from commit 6c07bf70500e2163fb6654a16780eabe63000d12)
|
|
Change the struct of aoc_moldule_parameters so that we can access the
up-to-date value all the times.
Bug: 303187479
Test: trigger AoC SSR multiple times by aocutil reset
Change-Id: I6f5d8cc27560049273cf32d08ce42964349eac5b
Signed-off-by: Guan-Yu Lin <guanyulin@google.com>
(cherry picked from commit 01fbb47b203b19d46890270475196320b4b30ca3)
|
|
Bug: 294495501
Change-Id: Ie1929a75017cad29b7a09ba5ee0afb391fc885a2
Signed-off-by: vincenttew <vincenttew@google.com>
|
|
Bug: 300854197
SBMerger: 571992243
Change-Id: I892f6aa147f6cb4e7f81e2ce4dcb1f0504b2e1d4
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
when coredump is corrupt, the value of num_sections
from ramdump_header may larger than the number of
sections. Out of range access will cause the KP.
Bug: 302929196
Change-Id: I423886cb8223617d84cf1b21231a914c70a30dd2
Signed-off-by: yixuanjiang <yixuanjiang@google.com>
|
|
Conflict files:
aoc.c
Bug: 300854197
SBMerger: 571992243
Change-Id: I4f8cbdab0ca7783ac37cbfc5fb326de12703d481
Signed-off-by: Pindar Yang <pindaryang@google.com>
|
|
Test: on device
Bug: 297923758
Change-Id: I9cf221a8673fd87884043d4c6472c07ef90e11d5
Signed-off-by: Alex Iacobucci <alexiacobucci@google.com>
|
|
It has chance hit KP when GSA unload FW and AoC SSR
then write coredump to dram without permission.
Also sync irq control flow with watchdog_irq.
Bug: 297627980
Change-Id: Idbfb71e2b729719ef5e93f9b68d6022396fa580a
Signed-off-by: yixuanjiang <yixuanjiang@google.com>
|
|
Test: on device
Bug: 297923758
Signed-off-by: Alex Iacobucci <alexiacobucci@google.com>
Change-Id: If520847c4a9516890f3a4f5bacc0b0cfa7e89b19
(cherry picked from commit b30d9385da96c052a80e49b1efd22c8d4d648b7a)
(cherry picked from commit 745d0f348984694ed0c62f997e0e448afe597a90)
|
|
SBMerger: 558810260
Change-Id: I73718aee57037b8ca8fdb905f58b28bae1b8211b
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
Bug: 300854197
SBMerger: 558810260
Change-Id: Ia0c91cabf4df69e46de4f00dda6b062ec38a5e40
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
Bug: 269221858
Change-Id: I4454054a6d559b203539043f1f4dfdfc3f15a29e
Signed-off-by: vincenttew <vincenttew@google.com>
|
|
SBMerger: 558810260
Change-Id: Ic8ff059fd210f1053191c11ac106e7e290c1ec15
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
add alsa control to config AoC mel value
Bug: 269221858
Change-Id: I15a236ba2471457f3ce06c42c8cc3c71ce655617
Signed-off-by: vincenttew <vincenttew@google.com>
|
|
Bug: 299407821
Test:
1. Verified VoIP record(incall_cap0) only scenario on T6 Pro
2. Verified the live caption(incall_cap0) for voice-call on C10
3. atest VtsHalAudioV7_1TargetTest
Change-Id: Ib1c3e9760390d8a733b1027210ff1561d55256b4
Signed-off-by: Carter Hsu <carterhsu@google.com>
|
|
The ALSA pcm lib driver assumes one period data was consumed in the
snd_pcm_update_hw_ptr0 function when updating the hw_ptr.
Update the hw_ptr if the consumed data size is equal to or
bigger than one period size.
Bug: 295546940
Test: Local test
Change-Id: I668c4272dbccb08fe7c46105b3c74b3a64bdd2d5
Signed-off-by: millerliang <millerliang@google.com>
|
|
The ALSA pcm lib driver assumes one period data was consumed in the
snd_pcm_update_hw_ptr0 function when updating the hw_ptr.
Update the hw_ptr if the consumed data size is equal to or
bigger than one period size.
Bug: 295546940
Test: Local test
Change-Id: I668c4272dbccb08fe7c46105b3c74b3a64bdd2d5
Signed-off-by: millerliang <millerliang@google.com>
|
|
SBMerger: 558810260
Change-Id: Ib82908cf71dca3f1fb802a928a8fe9dddafffdae
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
The original recovery flow didn't cover the case when usb connection
changes during AoC SSR. Instead of changing id to turn on/off host mode,
this fix uses fsm_reset so we can monitor id information during aoc
restart.
Bug: 297480242
Test: The following 5 tests passed:
1. AoC SSR during playback/capture.
2. Headset plugged-in/plugged-out during AoC SSR.
3. AoC SSR during gadget mode (ADB connection).
4. Restart pixel with host mode engaged.
5. Restart pixel with gadget mode engaged.
Change-Id: I9b0475a0d75181cbf09cea7dc2ae34511c0a0bc1
Signed-off-by: Guan-Yu Lin <guanyulin@google.com>
|
|
Set the stream type for pcm playback to mmaped
if runtime->access is SNDRV_PCM_ACCESS_MMAP_INTERLEAVED.
Bug: 283402753
Test: Playback and Record
Change-Id: Ib62931a42364089b44b32289b661a4798c0e23f4
Signed-off-by: millerliang <millerliang@google.com>
(cherry picked from commit d86f91a0e010d078d499184135fc460d3ecb30ab)
|
|
Start reading dp ringbuffer after available read size is
more then start threshold to prevent hit not enough data
to read from dp ringbuffer too easily.
Bug: 264508873
Test: local test
Change-Id: Ifaeeff80ed3a795318abd8c21d8d23c33bccef2b
Signed-off-by: Robert Lee <lerobert@google.com>
|
|
If AoC crashes while mic is muted, the mute command isn't sent again due
to a gate checking a wrong local cached value.
Bug: 296518692
Test: Tested on device
Change-Id: I46e3bc7b8127e610fad6d1fec4be90fbb2e652f4
Signed-Off-By: Greg Konzett <konzett@google.com>
|
|
CTS Sensor/DeviceSuspendTests was failing because the device fails
to enter the suspend state.
Bug: 288985122
Fix: 288985122
Test: Validated CTS passes.
Signed-off-by: Luis Delgado De Mendoza <luisddmg@google.com>
(cherry picked from commit 222ba497c1337c6f25ec0d4bcb4cff2782a52665)
(cherry picked from https://partner-android-review.googlesource.com/q/commit:ac2c776190fd01ea729513315ac6024c14929f70)
Merged-In: I1cfdae577e8e3636795afc9b900b50585db6da5a
Change-Id: I1cfdae577e8e3636795afc9b900b50585db6da5a
Change needed in kernel 5.15
|
|
SBMerger: 558810260
Change-Id: Ibbb9bd6068957e1957df8fa86c8b783ec3fd7219
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|