Age | Commit message (Collapse) | Author |
|
"LA.QSSI.13.0.r1-10200-qssi.0"
Change-Id: I4db0203d2218efb18f82aa939e00f946f7361b84
|
|
Change-Id: I83b41c8f211ea0a4cf2f2e0c5be199231b7b9c40
|
|
"LA.QSSI.13.0.r1-10000-qssi.0"
Change-Id: I95f03a97f939a3f6309f06e33feffee509f5d557
|
|
Change-Id: Iebb92e5392ce1358cdb721f48d782cf9e123d4a4
|
|
When inquiry rnr cancelled ,if RNR complete received
with in rnr cancel timeout, then stack checking
search state as idle and sending
discovery complete event.
But if pairing discovery started immediately after RNR cancel,
the seach state is discovery active.hence stack not
sending discovery complete event.
hence process rnr complete irresctive of search state if
rnr cancel timer is active
CRs-Fixed: 3467200
Change-Id: Ie44e2bad8da17959041d3533c45e7a417a307c31
|
|
into bt-sys.lnx.13.0
|
|
Local and remote devices have exchanged capabilities and
support EATT during initial connection. When remote device
rejects the first EATT credit based connection request
from local device during reconnection, EATT supported is
set to false. But, pending GATT operations of Attribute
channel on EATT BCB are not moved to non-EATT structure.
This change moves pending GATT operations of Attribute
channel on EATT BCB to non-EATT structure and deallocates
EATT blocks for Attribute channel.
CRs-Fixed: 3469729
Change-Id: I6aa3eb526acca9559b1a9941e2ef212806faf170
|
|
When EATT connect requests from profiles are pending
and LE encryption has not yet completed, sometimes EATT
connect requests are never sent to L2CAP layer.
Hence after LE encryption is enabled, send pending
EATT connect requests to L2CAP layer.
CRs-Fixed: 3452384
Change-Id: I5700cf418a91769a1453fb1a4d932e0dd1cf2e52
|
|
Paired Device info removed from storage once moved into
saved_list at BOND_NONE state. Need to repair again once
paired device is active.
CRs-Fixed: 3463104
Change-Id: I0ec5d06df8b7e7d6aa7ced1a46cfaadff489bafd
|
|
- Correct function call to fetch frame duration correctly from XML
Change-Id: I13868ea713fa0269cea7bf33b8fa9630488b0955
|
|
Issue:
when remote rejects with invalid channel id, stack parsing rcid first.
Hence we are not able to indicate disc to l2cap channel.
FIX:
As per spec Local CID should be first and remote CID should be second to
be parsed in respective order.
CRs-Fixed: 3464334
Change-Id: Ic9cf82a38ad7da44acd5104cc3ac0e754a68fa47
|
|
bt-sys.lnx.13.0
|
|
bt-sys.lnx.13.0
|
|
CRs-Fixed: 3471100
Change-Id: I6ab56746eace8d4cd2d7c7d7431d49d6dfadf04b
|
|
|
|
bt-sys.lnx.13.0
|
|
|
|
|
|
the last byte from the packet, resulting one-byte out-of-bound read." into bt-sys.lnx.13.0
|
|
|
|
into bt-sys.lnx.13.0
|
|
bt-sys.lnx.13.0
|
|
This reverts commit 8fbe2279caeb001fbabd387761e472d12eb1dd1e.
Change-Id: Id65c00a496e11cc871623c2d7bf2e2ee6618a3d5
|
|
This reverts commit a8747585175c618674ae8d7b663ceea2f928b2cb.
Change-Id: I8df9ebae5672b41f1c6d1d2cfe90eb5a94d3192c
|
|
"LA.QSSI.13.0.r1-09800-qssi.0"
Change-Id: I950597ebb936f5cdd667a2c5ce99393047d200de
|
|
Change-Id: I2b7c7073494ea2a4c3c65f9d560b0a3198c068f8
|
|
CRs-Fixed: 3463971
Change-Id: I1ed1cde35d98a64eda4e327f3e8313f234875f77
|
|
BT stack returns TRUE in the API of set_active_device if A2DP is disconnected during the process of
A2DP SHO. It causes MM audio start streaming , but failed. The fix is to return FAILURE if A2DP active
device change is failed.
Change-Id: I1427f716b4437e8e18f5afaaec1e77ca59529bb7
CRs-Fixed: 3459418
|
|
Add A2DP_CTRL_ACK_SHORT_WAIT_ERR support for qc hal
Change-Id: I4fa020db5cd6659a9bc9476ee3151c3c30d1b747
CRs-Fixed: 3345248
|
|
Included transport link parameter to the BTA_JvIsEncrypted()
Change-Id: I8e7178ba8e06e49e4a19370475655737b69b6bd6
|
|
Specific vendor has AAC enabled for all remote device but
some specific remote devices doesn't work well for AAC.
Add logic to use AAC BL for such remote devices from
specific vendor.
CRs-Fixed: 3454753
Change-Id: I5adac4d5586ee7b4cdd36c9fb89dc0e732ddf279
|
|
Change-Id: I8783c701c29b8e981718dd785fccd811aa14f6f6
CRs-Fixed: 3435570
|
|
Pass codec info params also to avoid recalculation
CRs-Fixed: 3442651
Change-Id: Ib95b7117a13e5e331b0b3b0d092ec767076249bb
|
|
Reject Set Config Cmd if configurations are not negotiated
properly during capability exchange.
CRs-Fixed: 3429434
Change-Id: Ied0d481384e0e18a2067515c59a3b9bf70e57754
|
|
After LE connection is established, read of Client
supported features characteristic for Robust caching
support is performed by stack. Some remote devices
which do not suport Robust caching do not respond
to this read by UUID procedure and this leads to LE
disconnection due to GATT response timeout.
Hence for interoperability, read of Client supported
features characteristic is initiated by stack only
when remote version is BT 5.1 or above.
CRs-Fixed: 3430416
Change-Id: Ia647eb54670e9cd4ab279c976631e089a8f115a9
|
|
CRs-Fixed: 3451360
Change-Id: I06385cb8760eb3af02a885ba71fba74f626dc10e
|
|
Bug: 245916076
Test: manual
Ignore-AOSP-First: security
Change-Id: I901d973a736678d7f3cc816ddf0cbbcbbd1fe93f
(cherry picked from commit daa3efc5e53c8613f4b1a33e095ff6c6460b8d02)
CRs-Fixed: 3446938
|
|
to read the last byte from the packet, resulting one-byte
out-of-bound read.
Also avdt_scb_hdl_pkt_no_frag passes zero-lenth packets to
upper-layer, this patch adds code to detect such packets
and err out if detected
The regression test is I9c87e30ed58e7ad6a34ab7c96b0a8fb06324ad54
Bug: 258057241
Test: atest net_test_stack_avdtp
Ignore-AOSP-First: security
Merged-In: If0c7b25f2e6cb4531bbb6254e176e8ad1b5c5fb4
Change-Id: If0c7b25f2e6cb4531bbb6254e176e8ad1b5c5fb4
(cherry picked from commit 89255db501097bbec90e4fcfc48d634deb239cd6)
CRs-Fixed: 3446923
|
|
|
|
|
|
If the length of the packet is less than 4 its
OOB access.
CRs-Fixed: 3445336
Change-Id: I5f6109d9d470031943d21b5602a8110346708295
|
|
|
|
Change-Id: I575a5fdeb23ead9f091cee08fec35db328fc4d20
|
|
CRs-Fixed: 3431184
Change-Id: I2762ad3f06cf1fede731a2a3575196331e1512f7
|
|
Root cause:
overflow causing leak in slot fds.
As slot id 0 not valid, we are not able to release these fds later.
Fix:
Changes are made to avoid overflow while allocate rfc slots.
CRs-Fixed: 3417458
Change-Id: I5d7efa34bfb97a6dd8e9d68615d29120a0ae51f0
|
|
Some remote devices do not respond to Read of Client supported
features characteristic(for Robust caching support) from local
device after LE connection is established. This results in
disconnection of LE link.
Skip read of client supported features characteristic for
Robust caching support after LE connection is established to
interop with such remote devices.
CRs-Fixed: 3367570
Change-Id: I772bec2f0ca134ef37be64ce0800ef2c51b5edbd
|
|
bt-sys.lnx.13.0
|
|
We should not enqueue one buffer into the queue
with null pointer.
Test: mm
CRs-Fixed: 3410704
Change-Id: I2adad5fbec33d9dfe250a3b1ad7331459bb5bc32
Signed-off-by: wangfei <wangfei25@xiaomi.com>
(cherry picked from commit 72a647c829e432e559010163edc0ab737786737d)
|
|
Root cause:
SDP search is queued as expected. But, before SDP finish customer app triggered inquiry again.
Hence inquiry queued in state machine.
As queue mechanism here is state machine driven, its queue only latest event.
Hence inquiry queued finally and executing.
Fix:
We should not allow inquiry, until we complete the pairing.
Change-Id: I65f2569dcedc87b6f39f49b4cfab297f2658717b
|
|
Root cause:
Some remote devices do not respond to DUT when user trun off and turn ON the BT
Fix:
Add remote device into L2CAP_DISCONNECT_ACL_DIRECTLY blacklist.
Disconnect ACL with remote device during BT off in DUT.
CRs-Fixed: 3421296
Change-Id: I068fc1c2b339c79966c1258a11bf265f77803c55
|