| Age | Commit message (Collapse) | Author |
|---|
|
Some devices don't respond to GATT Read by UUID requests for LE Appearance characteristics. This causes ATT transaction blockage and disconnection due to ATT transaction timeout.
Test: mmm packages/modules/Bluetooth
Flag: EXEMPT IOP fix
Bug: 402213061
Change-Id: Ib8cc3fbf32ccd0c1291314b65e56abe523acfdd6
|
|
Bug: 369381361
Test: m com.android.bt
Flag: EXEMPT, no logical change
Change-Id: I88e569e1e0b6c04716fa48345c1ede816fb405a3
|
|
|
|
Acquire lock before validating config to avoid race conditions
where config being reset with device_iot_config_clear and other thread
trying to access config instance
Bug: 365139919
Flag: Exempt, minor change
Test: mmm packages/modules/Bluetooth
Change-Id: I37f602648e4336b4a8740ef903b394d9dc940e7b
|
|
This reverts commit 704f84e152e253bbc5efe69d2a8f200b42203572.
Reason for revert: Does not fix the issue.
Change-Id: Ie7e45b65500d4f0ff3c9c84090a49da6bc875406
|
|
Once an AddressedPlayerChange notification has been sent,
the target device should reject PLAYBACK_STATUS_CHANGED,
TRACK_CHANGED, PLAYBACK_POS_CHANGED and
NOW_PLAYING_CONTENT_CHANGED notifications for completion.
Some carkits don't handle it properly and will loop in
setAddressedPlayer as though it was rejected. we need to
have an interop for those.
Note: this change is not directly flagged but should be
considered as such because this part of the code isn't
reachable without a flag enabled.
Bug: 393059901
Bug: 379039786
Bug: 381719439
Bug: 352279986
Flag: com.android.bluetooth.flags.set_addressed_player
Test: atest BluetoothInstrumentationTests
Change-Id: I075abd8d319b53bb8d87d42005ce5547885476f8
|
|
main
|
|
This reverts commit c2c7b9a9c7abbac4236bd8e7887c1640d5129685.
Reason for revert: One interop address was wrong
Change-Id: Icb18f868164902fa46b2c1f6bb4c0139c09e1a3c
|
|
|
|
Once an AddressedPlayerChange notification has been sent,
the target device should reject PLAYBACK_STATUS_CHANGED,
TRACK_CHANGED, PLAYBACK_POS_CHANGED and
NOW_PLAYING_CONTENT_CHANGED notifications for completion.
Some carkits don't handle it properly and will loop in
setAddressedPlayer as though it was rejected. we need to
have an interop for those.
Note: this change is not directly flagged but should be
considered as such because this part of the code isn't
reachable without a flag enabled.
Bug: 393059901
Bug: 379039786
Bug: 381719439
Bug: 352279986
Flag: com.android.bluetooth.flags.set_addressed_player
Test: atest BluetoothInstrumentationTests
Change-Id: If5e5febaa42a19d0312670549e2d2e6348e1aa12
|
|
Steps to reproduce: |-
1. Turn on BT
2. DUT pair/connect with remote buds
Repro rate: 10/10 times
Expected result: Both earbuds paired and connected
Observed result: only one earbud getting connected
Root cuase: For the second earbud the SIRK verification was not happening successfully
as remote not able to handle GATT_READ_BY_TYPE
Fix: Allow IOT devices to use GATT_READ_CHAR_VALUE instead of GATT_READ_BY_TYPE
Test: m -j
bug: 383695743
Flag: EXEMPT interop fix
Change-Id: I81dae7b91f4ff691957765759d023c98e82a7024
|
|
Bug: 290844229
Fix: 290844229
Flag: com.android.bluetooth.flags.device_iot_config_logging
Test: atest bluetooth_test_gd
Change-Id: I02c0b58cbafaaaa4c00973fe8ded997aea5fb420
|
|
Flag: EXEMPT adding Google module to build is not flagged
Test: presubmit
Test: m then flash device then check Bluetooth version & functionality
Bug: 383863941
Change-Id: If3ec0ce5bd6243721680510e2c7078ad61c2f0a3
Merged-In: If3ec0ce5bd6243721680510e2c7078ad61c2f0a3
(cherry picked from commit 14ba754f98634ab7ce42e020bf1b6e8736fd32c0)
|
|
JBL Flip 5/6 does not support HFP by specs and so is it in practice, but
it advertises the profile and that could cause trouble for Floss media
policy.
This CL adds an interop to disregard the HFP profile when advertised by
these headsets.
Bug: 381009230
Tag: #floss
Test: mmm packages/modules/Bluetooth
Flag: EXEMPT, enum adding and Floss-only logic change
Change-Id: I9079b0a877a3070cd1cbe7db9df202495e72f594
|
|
Bug: 305066880
Test: m com.android.btservices
Flag: EXEMPT, log change
Change-Id: If1dbab99250a4f62626206fc9f73d27e71ebab56
|
|
This function is inconsistently implemented by libc implementations
as it is not a standard ISO-C method. Rename the filler implementation
and make sure it is always used in place of the BSD version.
Bug: 379338269
Test: m com.android.btservices
Flag: EXEMPT, no logical change
Change-Id: I3c78d6e8ebf55da8137fbcced5aa9e97dbcf0b44
|
|
Bug: 364967694
Test: m .
Flag: Exempt refactor
Change-Id: Ic26c9b0305b76f66a461a4449845a8ca10726c36
|
|
The logging macros within have long been removed,
this is just cleanup work to inline the included headers
where necessary
Bug: 331817295
Test: m com.android.btservices
Flag: EXEMPT, no logical change
Change-Id: I2147ce4bdcea977b92318b2cb96e7020a8958161
|
|
Some HOGP keyboards expect a MTU exchange flow as part of the initialization to function properly. Trigger a MTU exchange for a device in the interop list.
Bug: 332761737
Tag: #floss
Test: mmm packages/modules/Bluetooth
Test: Manual test on Brya, Brydge keyboard working
Flag: EXEMPT, interop change and apply only to one and currently non-functioning device
Change-Id: I2efb78cae6072fb8fdd74a87628fb71804869164
|
|
The HID specification allows report maps to exceed 512 bytes. However, the Bluetooth specification requires all GATT attributes to be 512 bytes or less. This creates an interoperability issue when a device manufacturer implements a HID report map as a single GATT characteristic. This change handles this edge case by extending the report map after receiving the initial 512 bytes. This ensures compatibility with both HID and Bluetooth standards.
As this workaround is currently implemented for one specific device, the extra bytes of the report map exist as static const arrays; if more devices need such workarounds, it is suggested to store the report map in a separate file which can be loaded from the file system.
Bug: 332761737
Tag: #floss
Test: mmm packages/modules/Bluetooth
Test: Manual test on Brya, Brydge keyboard working
Flag: EXEMPT, interop change and apply only to one and currently non-functioning device
Change-Id: If9eb2e5e2828d0dea17f486207421525ffcccffc
|
|
This lint check will raise an error if a function is implemented
without a corresponding prototype declaration included in scope.
The main use case will be to ensure mock functions do not become out of
sync causing the original function to be linked instead.
Because of the large amount of errors, this change suppresses
errors in all locations, to be cleaned up gradually.
Bug: 369381361
Test: m com.android.btservices
Flag: EXEMPT, lint fixes
Change-Id: I888b5e284e237d8a6e2864bdde55c0a7fc0c348d
|
|
For the devices only use software path, the esco parameters should not
be changed to the DSP codec.
Bug: 358375314
Test: mmm packages/modules/Bluetooth
Test: manually test swb
Change-Id: Ief666dd7a2a965bd4034e6a206c6ec6f46875993
|
|
Restore SpacesBeforeTrailingComments to 2 to match current style
Then run clang-format -i **/**.{c,h,cc,cpp,hpp}
Then fix a bunch of typo (reported by gerrit)
Then fix unnecessary multiline string
Then fix whitespace for disabled clang format (reported by cpplint)
This is no-op
Bug: 311772251
Test: mmm packages/modules/Bluetooth
Flag: Exempt Format only
Change-Id: If135447803a40a2a07d4630ba2195e08ef8d250c
|
|
|
|
This flag was added in 24Q2 to allow the DSP to be used for LC3 when the BT controller does not support it. This is now the default behavior, so the flag is no longer needed.
Bug: 347316578
Change-Id: I581d217fa07472152cded36f160b964168735328
Flag: exempt, nextfood flag removal change
Test: atest net_test_bta
|
|
Updating connection parameter with non-zero latency before finishing
HOGP connection causes some HOGP devices reponse slowly during
profile connection. To avoid slowing down the connection, use interop
list to set zero latency for those devices. They can request proper
latency later based on its power state.
Bug: 346692549
Test: m com.android.btservices
Test pairing and reconnecting HOGP devices(Mad mouse) and confirm
the connection won't be slow down.
Change-Id: Ida03591bc4fa5137bd4b6960b4a92cf4f8ea3b15
|
|
Replacing TARGET_FLOSS with aflag.
The esco_parameters change was used to pass HFP 1.9 Qualification, but
since Android always uses HFP offload path, we made it Floss only. Added
behind aflag in this patch.
Bug: 345365532
Tag: #floss
Test: mmm packages/modules/Bluetooth
Flag: EXEMPT, Floss-only changes
Change-Id: I72fa6fe080f4ef0067806b00ca4606487de94dfb
|
|
Bug: 305066880
Flag: Exempt, trivial log change
Test: None
Change-Id: Icd642d73648c0a5401d910fda31a82ed21b42ecc
|
|
|
|
Bug: 337054479
Test: m .
Flag: Exempt, dead code removal
Change-Id: I957d5d98455df494693f542f0b5e397743fdc7b9
|
|
Bug: 261319328
Bug: 290844229
Bug: 261319328
Flag: com.android.bluetooth.flags.device_iot_config_logging
Test: atest bluetooth_test_gd
Change-Id: I80b2e09529da260183da5e20f0a166e35be03173
|
|
Bug: 307006275
Test: m com.android.btservices
Test: ./floss/build/build-in-container.py
Flag: EXEMPT no-op
Change-Id: Ia7c600f4ce1aa698b1a6c2739e7f6302e1e5f62f
|
|
Comment out unused parameters to have the compiler
check that they're unused.
Bug: 336310650
Test: mma -j32
Flag: EXEMPT, mechanical refactor
Change-Id: Id72f34af9f5a2cac076717e519c094d8ac2bf1de
|
|
|
|
Test: m com.android.btservices
Bug: 305066880
Flag: EXEMPT, mechanical refactor
Change-Id: I44e5106562fbbeb70a49c5640626e123ee7b568e
|
|
The added formatters for address types have all the same behavior.
ADDRESS_TO_LOGGABLE_STR still occurs in BTM_History, dumpsys,
log::assert_that calls
Test: m com.android.btservices
Bug: 305066880
Flag: EXEMPT, mechanical refactor
Change-Id: I950f76bd2f1d000638752759a3f08ba0be6f014c
|
|
Per HFP 1.9 spec 6.7.3 and 6.7.4, each codec has its own packet type.
After this CL, only the packet types defined in the spec would be sent
in the setup.
Bug: 324157546
Bug: 332650199
Test: m -j
Tag: #floss
Change-Id: Ie1a53fb743ef9bf7fbfaeff1b18c6f419aa95ce2
|
|
According to HFP 1.9 spec 6.7.1.2, AG should retry S1 as a
'safe setting' when the previous setup was rejected.
This patch is needed for the below tests
HFP/AG/ACC/BV-22-C
HFP/AG/ACC/BV-23-C
Bug: 324157546
Bug: 332650199
Test: Run above qualification tests
Test: m -j
Tag: #floss
Change-Id: If57eb871cd08616807a346fab0a51cb910425dc5
|
|
Some A2DP sink devices don't respond SDP request during A2DP
reconnection. Add an IOT device list for these devices to avoid
A2DP reconnection issue
Test: m com.android.btservices
Test: Manual. Trigger A2DP reconnection from peer device.
Change-Id: Id42f41dd008571fae771564e89f5c4f1c5cf4782
|
|
Used to include selectively <base/check.h> or <base/logging.h>
depending on the version of libchrome
Bug: 331817295
Test: m com.android.btservices
Flag: EXEMPT, dead code removal
Change-Id: I55c747973f1e354745cbb54f6c605ba6c07d9471
|
|
Convert remaining DLOG, LOG, CHECK_xx
occurences in the same pass
Test: m com.android.btservices
Bug: 305066880
Flag: EXEMPT, mechanical refactor
Change-Id: I7ae3e14c84b6ee5dc2ac0babc652c699c7424418
|
|
Test: m com.android.btservices
Bug: 305066880
Flag: EXEMPT, mechanical refactor
Change-Id: I40e01004b3cf81037ed2926397aa18904d7446fe
|
|
This issue is detected by
Checker: NO_EFFECT
Type: Enum compared against 0
Checker Wiki: https://swwiki.mediatek.inc/display/SCA/NO_EFFECT
Defect is in:
File: vendor/mediatek/proprietary/packages/modules/Bluetooth/system/device/src/esco_parameters.cc
Fucntion and lines:
esco_parameters_for_codec(esco_codec_t, bool):215
Bug: 328847127
Test: m com.android.btservices
Flag: EXEMPT, no logical change
Change-Id: I53325e00a1c6e77b50ee6902139a3ecd728be5ce
|
|
Test: m com.android.btservices
Bug: 305066880
Flag: EXEMPT, mechanical refactor
Change-Id: I134695be10eb0f8307753481027441e47bb0a803
|
|
is_ready can be replaced by GetController, since
the controller is ready if the shim returns it.
Bug: 321273521
Test: atest bluetooth_test_gd_unit
Flag: EXEMPT, no logical change
Change-Id: I8a2b7c700867674b9298e1433438d6f8941c38de
|
|
Due to the limitations of the current HID host implementation, it
assumes that the HID peripheral has only one HOG service, or if there
are multiple HOG services, the first one is always active. To address
this issue temporarily, an interop quirk has been introduced to allow
the HID host to connect to a non-first HID service for certain devices.
However, for a long-term solution, the HID stack needs to be refactored
to handle multiple HID services simultaneously.
Bug: 286413526
Bug: 328129633
Tag: #floss
Test: mmm packages/modules/Bluetooth
Test: Manual test on Brya, Logitech CASA working
Flag: EXEMPT, apply only to one and not-functioning device
Change-Id: I339908bc08bf33a977cf80b0f72ae63db1a61986
|
|
This reverts commit c8c845228643298fbbc19f0e1756be98a24fda3b.
Reason for revert: Logic error
Change-Id: I75f0791dbb2b6920c9b906508dd6b12d82141c95
|
|
Due to the limitations of the current HID host implementation, it
assumes that the HID peripheral has only one HOG service, or if there
are multiple HOG services, the first one is always active. To address
this issue temporarily, an interop quirk has been introduced to allow
the HID host to connect to a non-first HID service for certain devices.
However, for a long-term solution, the HID stack needs to be refactored
to handle multiple HID services simultaneously.
Bug: 286413526
Bug: 328129633
Tag: #floss
Test: mmm packages/modules/Bluetooth
Test: Manual test on Brya, Logitech CASA working
Flag: EXEMPT, apply only to one and not-functioning device
Change-Id: I0733f1597d9058c992e46b99c6a31748714d8dfa
|
|
Bug: 279512287
Test: mmm packages/modules/Bluetooth
Flag: Exempt, mechanical refactor
Change-Id: I962860d1e39a0b1b32bfa2816c12a79b557b8a62
|
|
Bug: 301661850
Test: mma -j32
Flag: EXEMPT, build change
Change-Id: I21075c2ceeb3c468a1c6e5c7aeac8d03054582e3
|
Ruina Liu
yamei du