OSDN Git Service
Nitin Srivastava [Thu, 13 Feb 2014 14:01:40 +0000 (19:31 +0530)]
Bluetooth: Clear Notification array on AVRCP disconnect
DUT does not clear previous AVRCP notification registration
on AVCTP disconnect hence DUT sends response even when remote
device hasn't registered.
Changes made to clear notification array and while sending
notification respose check if event is registered.
Change-Id: I610f8ae6381a63236378f179bb2a619ec0c6549e
Nitin Srivastava [Fri, 31 Jan 2014 10:21:24 +0000 (15:51 +0530)]
Set out stream to null if control socket fails to connect
Set out stream to null if control socket fails to connect.
This helps in avoiding wrong usage of audio device out stream
even after the same is freed as part of control socket connection
failure.
Change-Id: I8e2a7bbc71d581525b7eef1955e411c220d01627
Nitin Srivastav [Fri, 21 Feb 2014 13:03:14 +0000 (18:33 +0530)]
Bluetooth: retry recv in case of EINTR
- retry recv call in HAL, in case socket call
fails with EINTR
Change-Id: I3ab27752b0ce0a68f6776cf5da340580e7d1f345
Zhihai Xu [Wed, 27 Nov 2013 08:40:53 +0000 (14:10 +0530)]
Bluetooth: GAP: Remove bonded device info on Authentication fail
Remove bonded device information from the BT-IF layer storage
on authentication failure if already bonded with just works
model.
from QCOM
Change-Id: Ia3c1635c446f39b1dbd4cdd0122c2f70cbbc8dd7
Sharvil Nanavati [Thu, 17 Apr 2014 04:51:56 +0000 (21:51 -0700)]
Use standard C99 integral types. This offers better type safety.
Notice that I had to change a couple of function declarations
because the compiler caught the inconsistency with the new typedefs.
Change-Id: I202fd1dc086de41e1cfe083eb1a5f74a942d01e5
Zhihai Xu [Thu, 16 Jan 2014 05:49:32 +0000 (21:49 -0800)]
Increasing the SDP_MTU_SIZE from 256 to the default 672 bytes
This works better with IOT of devices
bug
10588724
Change-Id: I28fd74339d8f344b056b82bf2d71e7d70be7a0fa
Zhihai Xu [Fri, 21 Mar 2014 01:01:09 +0000 (18:01 -0700)]
a2dp audio HAL: use correct interface version
Do not use AUDIO_DEVICE_API_VERSION_CURRENT version
but AUDIO_DEVICE_API_VERSION_2_0 to reflect the version
actually implemeted.
Change-Id: I31407af7fc4769be1d259ee739422ce9f598be2e
Zhihai Xu [Sat, 22 Feb 2014 02:20:37 +0000 (18:20 -0800)]
Allow simultaneous LE Observe and BTM Inquiry/Scan
This patch allows BTM inquiry and LE observe to be launched
and overlap each others' occurance by redirecting results
and handling states accordingly.
CRs-Fixed: 579505
Conflicts:
stack/btm/btm_ble_gap.c
bug
12939764
Change-Id: Ifde73f17f8d40e9ebfff3b49e409591189574ee0
Zhihai Xu [Fri, 7 Feb 2014 19:47:26 +0000 (01:17 +0530)]
SDP: Allocate proper sized scratch buffer for SDP response
The buffer allocated for attribute response is being reused
for subsequent requests for an SDP connection. Check the
buffer size before using for subsequent requests and
reallocate if required.
Change-Id: Ib00aff2c4512192c29265cf0adb89e380adf0b35
Zhihai Xu [Mon, 7 Apr 2014 08:13:06 +0000 (16:13 +0800)]
flounder64: bluetooth: Adjust volume level on BT HS and DUT side
MW600 Bluetooth address is "58:17:0c:e8:b0:7a". It is not included in
AVRCP profile white list. So it can't support absolute volume. We need
to whitelist it by device name "MW600" instead device address method.
Signed-off-by: Thomas.TT_Lin <Thomas.TT_Lin@htc.com>
Change-Id: I41b27fbab13dcf9d0005e8d5ae25ffb95eb546d4
Zhihai Xu [Wed, 26 Mar 2014 01:36:27 +0000 (18:36 -0700)]
audio: fix set_parameters return value.
xxx_set_parameters functions were returning the status
returned by str_parms_create_str() which is incorrect.
These functions should return 0 when no error occurs.
Change-Id: I984d95457c985fbaf1a15baa522f8a84f99dd7c9
Hemant Gupta [Mon, 21 Apr 2014 18:53:21 +0000 (00:23 +0530)]
HID: Use dynamic memory while sending data
This patch uses dynamic memory allocation while sending HID output data to
remote device depending on size of data to be sent. Without this patch
fixed size static buffer of 200 bytes was being used for sending data to
remote device, which was resulting in crash in case data size to be sent
to remote device was greater than 200 bytes.
Change-Id: I5ed12422355e88d7ef0391668800d7619ca3b685
Hemant Gupta [Fri, 25 Oct 2013 16:30:07 +0000 (22:00 +0530)]
HID: Parse paired hid devices from bluez.
This patch adds support for parsing SDP records of paired HID devices
from bluez device storage path and store the parsed hid sdp data into
bluedroid storage path. Without this patch, it was not possible to
have paired HID devices during upgrade from BlueZ to Bluedroid Stack.
Change-Id: I7db4a801f3cf9ca46fc0eabfd929e8d7c16b5edb
Hemant Gupta [Thu, 28 Nov 2013 07:37:10 +0000 (13:07 +0530)]
GAP: Try to retrieve CoD from NVRAM if 0
This patch tries to retreive CoD from NVRAM if CoD is 0. This is
because in scenarios when airplane mode on/off, inquiry is not
performed again on devices shown in Available devices List of
UI, so CoD is 0 during SSP information as database of stack is
not properly updated since inquiry is not performed again. Without
this patch it was observed because of wrong CoD info device icon
for HID mouse was changed to Joystick and device is not shown in
Paired List. Also some extra debug logs are added in code for
better debugging.
Change-Id: I14858ccee04728cb4b9cd60d397b79fa6494b842
Hemant Gupta [Fri, 5 Jul 2013 18:05:58 +0000 (23:35 +0530)]
HID: Reset Disconnect Reason on Connection Establishment
This patch resets the disconnect reason to successful on HID Interrupt
channel establishment. Without this change, wrong reason is propogated
to BTA layer on disconnection initiated either from DUT or Remote end.
Change-Id: I42641b511e44a761574faf8eafb7b0826e24aeb3
Zhihai Xu [Fri, 13 Dec 2013 10:39:21 +0000 (16:09 +0530)]
HID: Remove device from nvram on auth failure
This patch removed hid pointing device from nvram on authentication failure
due to pairing timeout or when wrong passkey is entered. Without this patch
it was observed that hid pointing device was always shown in paired list
when DUT was restarted after authentication failure.
CL from qcom(Hemant Gupta)
Change-Id: I69c43d2ed982d5d5abfd5707db90ef4b5c65b34f
Zhihai Xu [Mon, 21 Apr 2014 19:05:42 +0000 (00:35 +0530)]
HID: Use dynamic memory while sending report
This patch uses dynamic memory allocation while sending HID output data to
remote device depending on size of data to be sent. Without this patch
fixed size static buffer of 200 bytes was being used for sending data to
remote device, which was resulting in crash in case data size to be sent
to remote device was greater than 200 bytes.
CL from qcom(Hemant Gupta)
Change-Id: Icc8cd4a4ecfd4bc30cbf848a7c865fcf9308ddf8
Hemant Gupta [Tue, 29 Oct 2013 06:11:16 +0000 (11:41 +0530)]
HID: Serialize L2CAP disconnection
This patch serializes host initiated hid disconnection by sending
control channel disconnect request only when disconnection response
for interrupt channel has been received.
Change-Id: Ie0c5c62593bb0dd32f83b76bf0d6c1d950f1a684
Hemant Gupta [Fri, 22 Nov 2013 15:23:53 +0000 (20:53 +0530)]
HID: Handle connection failure in disconnecting state
This patch handles connectio error of interrupt channel while in
disconnecting state. Without this patch it was seen that upper
layers would not be informed about failure of interrupt channel
connection in this corner case leading to UI stuck at connecting
forever.
Change-Id: I1f5f18e70470eff17b21f2ed921edaefd3c89f9c
Hemant Gupta [Mon, 9 Dec 2013 13:13:33 +0000 (18:43 +0530)]
HID: Reset Idle timeout to 0 while hid disconnection
This patch resets idle l2cap timeout to 0 while disconnecting HID L2CAP's.
without this patch it was observed that after hid l2cap disconnection,
ACL disconnection was taking 4 seconds.
Change-Id: I39f16d3981f4a8b5cdf28cd7865631900007d5b8
Zhihai Xu [Mon, 21 Apr 2014 19:08:50 +0000 (19:08 +0000)]
Merge "HID: Fix memory leak in HID Code" into klp-wireless-dev
Zhihai Xu [Sun, 24 Nov 2013 10:53:55 +0000 (16:23 +0530)]
HID: Fix memory leak in HID Code
This patch fixes memory leak in HID code when interrupt reports are
received from remote device before HID connection is setup.
CL from qcom(Hemant Gupta)
Change-Id: If4fe961921c7b1212035cf5f2aa60d2359e322e2
Zhihai Xu [Mon, 21 Apr 2014 19:04:04 +0000 (19:04 +0000)]
Merge "HID: Establish Interrupt Channel when Control channel is configured" into klp-wireless-dev
Hemant Gupta [Thu, 10 Apr 2014 11:00:55 +0000 (16:30 +0530)]
HID: Reduce poll time to 50 ms for polling from uhid driver
This patch reduces poll time for data received from uhid driver to 50 ms
from existing 500 ms. Without this change bta thread was getting blocked upto
max 500 ms when HID is disconnected, which caused delay in sending
of AVDTP Media packets during A2DP and HID concurrent test cases resulting
in audio break on BT headset during HID disconnection as closing of uhid
polling thread is done from BTA context.
Change-Id: I033e2d81e895831555bbfdff1f93fad5f2cc5e59
Hemant Gupta [Mon, 21 Apr 2014 18:14:13 +0000 (23:44 +0530)]
HID: Update SSR parameters during copy.
This patch correctly stores SSR parameters (max latency and min timeout)
retrieved from remote device SDP record in NVRAM.
Change-Id: I87d8b6181981090df21ab746deac7053a1d0e938
Zhihai Xu [Mon, 21 Apr 2014 18:15:54 +0000 (18:15 +0000)]
Merge "Fix GKI exception of calling free on an already freed buffer" into klp-wireless-dev
Zhihai Xu [Tue, 25 Feb 2014 02:07:14 +0000 (18:07 -0800)]
Fix GKI exception of calling free on an already freed buffer
Various parts of btif_hh.c were creating GKI buffers and
keeping references to them and freeing them in odd and
unnecessary ways. The buffer is freed by lower levels
of the stack once the buffer has been sent to the chip
at the l2c layer and shouldn't be freed by btif_hh itself
since it's possible to double free, and there could
also be race conditions with other threads already processing
the buffer while the reference is freed if the API calls
are invoked again before the previous invocation was completely
processed.
Also added a helper routine to simplify buffer creation and
initialization.
Change-Id: Ia6039983502e2670b2325d90310244edf843b692
Signed-off-by: Mike J. Chen <mjchen@google.com>
Zhihai Xu [Mon, 21 Apr 2014 18:04:57 +0000 (18:04 +0000)]
Merge "bluetooth: hid: store ssr_max_latency and ssr_min_tout in bonding db" into klp-wireless-dev
Zhihai Xu [Mon, 27 Jan 2014 22:05:17 +0000 (14:05 -0800)]
bluetooth: hid: store ssr_max_latency and ssr_min_tout in bonding db
Store the ssr_max_latency and ssr_min_tout parameters in the
bonding database. Previously these parameters were not stored for HID
devices. This caused an inconsistency in sniff behavior between
when the HID device was paired and following a reboot of the host.
When the HID device is paired the SSR parameters are retrieved
from the HID device and are stored in the HID Host's device list.
If the host is rebooted then the HID Host's device list is
constructed from the bonding database.
If the SSR max latency is non-zero then the SSR parameters are
sent to the firmware and SSR is used. In the event that the SSR
max latency is 0, as it is if the parameters are not found, then
SSR is disabled.
Bug:
12764547
Change-Id: I63637575fdfcf4cb4ca0ce8dc2e4ccda9cccaf66
Signed-off-by: Adam Hampson <ahampson@google.com>
Hemant Gupta [Wed, 2 Apr 2014 13:58:47 +0000 (19:28 +0530)]
HID: Establish Interrupt Channel when Control channel is configured
This patch starts establishment of Interrupt channel only when HID Control
channel has been properly configured.
Change-Id: I28f1100abeece9255fbdd5df312236c18baebe17
Sharvil Nanavati [Mon, 7 Apr 2014 20:52:37 +0000 (13:52 -0700)]
Add flow control for PAN.
This change reads from the TAP fd only when btpan_cb.flow == 1.
The value of this flag changes based on L2CAP congestion events.
http://b/
13246629
Change-Id: Ic28bf65737dbd74c41b68198fceee58e3bff08d4
Conflicts:
btif/src/btif_pan.c
Wei Wang [Fri, 18 Apr 2014 23:08:43 +0000 (16:08 -0700)]
Fixed a build warning.
Change-Id: I34885d6e588855d3ca8192db8a54b43ef348c794
Wei Wang [Fri, 18 Apr 2014 22:33:41 +0000 (22:33 +0000)]
Merge "LE: Add controller based advertising filter API (2/2)" into klp-wireless-dev
Andre Eisenbach [Thu, 21 Nov 2013 01:23:06 +0000 (17:23 -0800)]
LE: Add controller based advertising filter API (2/2)
Change-Id: I444654bf19a048d5fcef33f79edbf1cdda3d3b62
Sharvil Nanavati [Fri, 18 Apr 2014 05:32:05 +0000 (22:32 -0700)]
Eliminate warnings, add -Wall and -Werror to all projects.
This change enables all warnings and treats warnings as errors. As
a result, nobody should be able to submit new code to bluedroid
that contains compiler warnings. Instead, they should fix the warnings
before submitting.
This change has already caught a type mismatch bug in
stack/avrc/avrc_api.c where an integer was being stored in a boolean.
Change-Id: Ic41960754b39a1a6aa37f1c1095003195f466183
Sharvil Nanavati [Sat, 5 Apr 2014 06:57:10 +0000 (23:57 -0700)]
Rewrite PAN_Write in terms of PAN_WriteBuf.
The original code looked copy/pasted. The new implementation still
results in the same number of buffer copies but is much easier to
read and reduces redundancy.
Change-Id: I3ace395d2b124074e9687d13ae9a1b0b93a25a31
Sharvil Nanavati [Thu, 17 Apr 2014 17:50:13 +0000 (17:50 +0000)]
Merge "Remove unused file gki_inet.h and BIG_ENDIAN macro." into klp-wireless-dev
Matthew Xie [Thu, 17 Apr 2014 04:54:50 +0000 (04:54 +0000)]
Merge "Bluetooth: GAP: Send bond none in case of temp pairing" into klp-wireless-dev
Zhihai Xu [Wed, 16 Apr 2014 21:35:43 +0000 (21:35 +0000)]
Merge "LE: Add API to configure MTU for a given connection (2/4) This patch from broadcom(andre) bug:
13571470 Change-Id: I932905590ee9c40cb67e7d09715c0f8e59d2f662" into klp-wireless-dev
Zhihai Xu [Tue, 25 Mar 2014 13:31:17 +0000 (06:31 -0700)]
LE: Add API to configure MTU for a given connection (2/4)
This patch from broadcom(andre)
bug:
13571470
Change-Id: I932905590ee9c40cb67e7d09715c0f8e59d2f662
Zhihai Xu [Wed, 16 Apr 2014 19:05:22 +0000 (19:05 +0000)]
Merge "remove all advertised services in SDP btif_av_init is called twice. Don't need call it in enable. disable EDR ATT SDP record" into klp-wireless-dev
Zhihai Xu [Sat, 5 Apr 2014 07:25:20 +0000 (00:25 -0700)]
remove all advertised services in SDP
btif_av_init is called twice. Don't need call it in enable.
disable EDR ATT SDP record
bug:
13558005
Change-Id: Id834697eae150018f9a0382d9c1e6c972e885739
Zhihai Xu [Wed, 16 Apr 2014 19:03:38 +0000 (19:03 +0000)]
Merge "open GATT client didn't reserve the BLE link if open after connected." into klp-wireless-dev
Zhihai Xu [Mon, 3 Mar 2014 20:50:13 +0000 (12:50 -0800)]
open GATT client didn't reserve the BLE link if open after connected.
need to reserve the BLE link to avoid BLE disconnection.
BTA_GATTC_OPEN_ERROR won't resereve the BLE link usage.
use BTA_GATTC_OPEN instead of BTA_GATTC_OPEN_ERROR to reserve the
BLE link. other the stack disconnect the BLE link due to no Gatt interface
use this connection.
bug:
13228455
Change-Id: I1f9daa2171668f2ae535d998b401f1f0825d0451
Zhihai Xu [Wed, 16 Apr 2014 19:02:29 +0000 (19:02 +0000)]
Merge "Workaround IPHONE problem" into klp-wireless-dev
Zhihai Xu [Tue, 18 Mar 2014 22:27:05 +0000 (15:27 -0700)]
Workaround IPHONE problem
When we send LL_ENC_REQ to iphone early, iphone didn't send
LL_ENC_RESPONSE back to us, later the iphone will disconnect from us.
The workaround is to send LL_ENC_REQ after receive first ATT read request
from iphone, Iphone will be ready at that time by experiment.
bug:
13417692
Change-Id: Ic42dda36b927df260113771152cff67032b22e5f
Zhihai Xu [Wed, 16 Apr 2014 19:01:09 +0000 (19:01 +0000)]
Merge "treat BTA_BUSY as success to avoid unpair the BLE device" into klp-wireless-dev
Zhihai Xu [Thu, 13 Mar 2014 22:21:37 +0000 (15:21 -0700)]
treat BTA_BUSY as success to avoid unpair the BLE device
when switch from peripheral mode, both server and client will
call btif_gatt_check_encrypted_link, the first call will start the
encryption. the second call will get callback with BTA_BUSY status,
before encryption complete, which cause unpair with the remote device.
we shouldnot treat BTA_BUSY as error.
bug:
13417692
Change-Id: I065432dde2553e8ccee3a14a7b321b85bca8d3ba
Zhihai Xu [Wed, 16 Apr 2014 18:52:15 +0000 (18:52 +0000)]
Merge "print MTU size after get MTU request from iphone." into klp-wireless-dev
Zhihai Xu [Fri, 21 Mar 2014 00:50:12 +0000 (17:50 -0700)]
print MTU size after get MTU request from iphone.
bug:
13568537
Change-Id: If1a4e93b7ef53662bd740da0d3bb91fe747de031
Zhihai Xu [Wed, 16 Apr 2014 18:50:54 +0000 (18:50 +0000)]
Merge "support resolve BLE random mac address for both central and peripheral mode." into klp-wireless-dev
Zhihai Xu [Tue, 11 Mar 2014 22:01:45 +0000 (15:01 -0700)]
support resolve BLE random mac address for both central and peripheral mode.
bug:
12990335
Change-Id: Ica42dcf039487b7b01454ce9255f3f6fae2faaaf
Zhihai Xu [Wed, 16 Apr 2014 18:46:26 +0000 (18:46 +0000)]
Merge "support peripheral mode feature to connect iphone to android device." into klp-wireless-dev
Zhihai Xu [Mon, 10 Feb 2014 18:57:13 +0000 (10:57 -0800)]
support peripheral mode feature to connect iphone to android device.
enable BLE_PERIPHERAL_MODE_SUPPORT flag and fix the bug
the discoverability flag is not set in advertisement packet,
which cause the central device(iphone) can not find android device
during scan.
use displayOnly IO cap for dory, so dory can display the pin,
and the iphone(central device) will input the pin.
bug:
12797572
Change-Id: Ie15ea9e340e3692df846399b9a85aa6d2a97eba7
Zhihai Xu [Wed, 16 Apr 2014 18:22:08 +0000 (18:22 +0000)]
Merge "increase the bluetooth task priority when start a2dp." into klp-wireless-dev
Zhihai Xu [Wed, 8 Jan 2014 19:45:17 +0000 (11:45 -0800)]
increase the bluetooth task priority when start a2dp.
change the BT task priority based on audio play state.
increase the BT task priority to ANDROID_PRIORITY_URGENT_AUDIO
,when start ad2p audio playing.
to better prevent CPU premption by other process/task(UI).
restore the BT task priority when stop a2dp audio playing.
bug:
12082841
Change-Id: I34e8344cffea87f68987149c820cd3e84a4196d1
Conflicts:
utils/include/bt_utils.h
Sharvil Nanavati [Sat, 5 Apr 2014 07:44:19 +0000 (00:44 -0700)]
Compile using C99.
Change-Id: Ifda03e0b3ee3a915bf7ce5a5dce27e5e215dcce2
Conflicts:
bta/Android.mk
gki/Android.mk
main/Android.mk
stack/Android.mk
test/bluedroidtest/Android.mk
Hemant Gupta [Tue, 24 Dec 2013 14:29:57 +0000 (19:59 +0530)]
Bluetooth: GAP: Send bond none in case of temp pairing
Send BOND_NONE in case of temporary device pairing on
receiving the authentication complete, so that UI
doesn't get stucked in pairing for that device.
Change-Id: Ided412ef89d928f958a0052362be145232411c0a
Sharvil Nanavati [Mon, 14 Apr 2014 04:26:56 +0000 (21:26 -0700)]
Remove unused file gki_inet.h and BIG_ENDIAN macro.
Change-Id: I52bb7befdff6138dce4ce74ebd65d4e3763b82aa
Sharvil Nanavati [Sun, 13 Apr 2014 06:01:10 +0000 (23:01 -0700)]
Remove executable bit on source files.
Change-Id: Id456478aea626f57fed3479346e8fb4dbd0fafa6
Andre Eisenbach [Wed, 5 Mar 2014 23:41:53 +0000 (15:41 -0800)]
DO NOT MERGE: LE: Add status callback function for LE broadcasts
Cherry pick from master
Change-Id: If4342050ce1141a9c2111286e7b3cccdc012a91e
Matthew Xie [Wed, 19 Mar 2014 20:47:12 +0000 (13:47 -0700)]
DO NOT MERGE Revert "external/bluetooth/bluedroid: convert LOCAL_MODULE_PATH to LOCAL_MODULE_RELATIVE_PATH"
This reverts commit
273789bb86ef99f5ce022bb64c601ea4ea6f8bb3.
Change-Id: I29e511a8176979a6bf311206427b94a19f06f284
Matthew Xie [Mon, 24 Feb 2014 14:32:10 +0000 (14:32 +0000)]
am
7d006e6b: am
a69e0ae1: am
47128cf9: Merge "Bluetooth: fix sniff issue with multiple rfcomm connection in same ACL"
* commit '
7d006e6b176172cb370cc183b4f2eab501e85c7c':
Bluetooth: fix sniff issue with multiple rfcomm connection in same ACL
Matthew Xie [Mon, 24 Feb 2014 14:32:10 +0000 (14:32 +0000)]
am
9d368b68: am
1292dd82: am
6bc62a36: Merge "Bluetooth: GAP: Update bonded device info immediately on NVRAM"
* commit '
9d368b68841dee6e6f581a3890433dcdc022fbb8':
Bluetooth: GAP: Update bonded device info immediately on NVRAM
Matthew Xie [Mon, 24 Feb 2014 14:29:02 +0000 (14:29 +0000)]
am
a69e0ae1: am
47128cf9: Merge "Bluetooth: fix sniff issue with multiple rfcomm connection in same ACL"
* commit '
a69e0ae1107e3148542a76ac840f3bcaa8ee2661':
Bluetooth: fix sniff issue with multiple rfcomm connection in same ACL
Matthew Xie [Mon, 24 Feb 2014 14:29:02 +0000 (14:29 +0000)]
am
1292dd82: am
6bc62a36: Merge "Bluetooth: GAP: Update bonded device info immediately on NVRAM"
* commit '
1292dd82812d76c54568796fbdafd1fdadc816eb':
Bluetooth: GAP: Update bonded device info immediately on NVRAM
Matthew Xie [Mon, 24 Feb 2014 14:25:25 +0000 (14:25 +0000)]
am
47128cf9: Merge "Bluetooth: fix sniff issue with multiple rfcomm connection in same ACL"
* commit '
47128cf95b72eedb7a6da29687b1ff6b668b59d6':
Bluetooth: fix sniff issue with multiple rfcomm connection in same ACL
Matthew Xie [Mon, 24 Feb 2014 14:25:25 +0000 (14:25 +0000)]
am
6bc62a36: Merge "Bluetooth: GAP: Update bonded device info immediately on NVRAM"
* commit '
6bc62a368da0a434a4e01f7bcb3273bd52a07568':
Bluetooth: GAP: Update bonded device info immediately on NVRAM
Matthew Xie [Mon, 24 Feb 2014 01:29:59 +0000 (01:29 +0000)]
Merge "Bluetooth: fix sniff issue with multiple rfcomm connection in same ACL"
Matthew Xie [Sun, 23 Feb 2014 22:34:33 +0000 (22:34 +0000)]
Merge "Bluetooth: GAP: Update bonded device info immediately on NVRAM"
Matthew Xie [Sat, 22 Feb 2014 01:22:04 +0000 (01:22 +0000)]
Merge "Bluetooth: Cleanup bt_config.xml during BLE Scan"
Sunny Kapdi [Wed, 20 Nov 2013 22:41:41 +0000 (14:41 -0800)]
Bluetooth: Cleanup bt_config.xml during BLE Scan
BLE Scan needs to clean up the bt_config.xml periodically so that
new entries could be added to bt_config.xml. Also, there is no
need to add BLE Random bd_addr to the bt_config.xml
Change-Id: I00ba2e50dd39a021099052cb569786f6cb2cc8b9
b/
12912008
Pradeep Panigrahi [Wed, 18 Sep 2013 06:55:37 +0000 (12:25 +0530)]
Bluetooth: fix sniff issue with multiple rfcomm connection in same ACL
When there are multiple rfcomm connection with same appid and peer bd
address, we were clearing jv pm control block in addition to closing
connection to service with registered app id. Since appid is common
across all the client and server app registed to jv. we were clearing
entire connection to service for this appid and peer bdadres which was
corrupting sniff logic, when another rfcomm connection exists with same
peer bdaddress. So avoiding closing of connection to the service when
multiple rfcomm connection exists in same ACL.
Change-Id: Id71ef5f7f0e00f1bda3b70a4826c31bef46caafc
Srinu Jella [Thu, 21 Nov 2013 11:40:02 +0000 (17:10 +0530)]
Bluetooth: GAP: Update bonded device info immediately on NVRAM
As the new bt-if config updation have delay in writing items
to the NVRAM, some times there is a chance of not updating
incase of battery removal happens before the actual write onto
NVRAM.
As the bonded device info is important data to store, it will
update bonded info immediately onto NVRAM.
Change-Id: I9699c8225a77d9272420334b4c76b3c1df87e459
Mike J. Chen [Fri, 14 Feb 2014 23:41:11 +0000 (23:41 +0000)]
am
0c880603: am
6a342588: am
136e4b1c: Merge "Don\'t send a dummy 0 HID event when connection is closed" into klp-modular-dev
* commit '
0c880603dba9f9f09523509f19c6808ba8f52e12':
Don't send a dummy 0 HID event when connection is closed
Mike J. Chen [Fri, 14 Feb 2014 23:13:26 +0000 (23:13 +0000)]
am
6a342588: am
136e4b1c: Merge "Don\'t send a dummy 0 HID event when connection is closed" into klp-modular-dev
* commit '
6a342588e9a161ec5347035df7472f1f4a8e86ea':
Don't send a dummy 0 HID event when connection is closed
Mike J. Chen [Fri, 14 Feb 2014 22:54:44 +0000 (22:54 +0000)]
am
136e4b1c: Merge "Don\'t send a dummy 0 HID event when connection is closed" into klp-modular-dev
* commit '
136e4b1c73af97b3f743395da47f137356776090':
Don't send a dummy 0 HID event when connection is closed
Mike J. Chen [Fri, 14 Feb 2014 19:18:11 +0000 (19:18 +0000)]
Merge "Don't send a dummy 0 HID event when connection is closed" into klp-modular-dev
Mike J. Chen [Fri, 14 Feb 2014 02:17:56 +0000 (18:17 -0800)]
Don't send a dummy 0 HID event when connection is closed
This was put there by Broadcom to try to handle held keys
on disconnect, but the best place to do that is in InputManager.
There's no guarantee that a report_id 1 is keyboard, and
in particular it causes game controllers to get a false
joystick move event to 0,0.
Bug
12785130
Change-Id: I85fbafa44d927082def3848132344550e8b94bb0
Signed-off-by: Mike J. Chen <mjchen@google.com>
Matthew Xie [Wed, 12 Feb 2014 23:03:13 +0000 (23:03 +0000)]
am
081ff423: am
aa58a6cf: am
13a01401: Merge "Don\'t leak process-private data on Change Local Name command."
* commit '
081ff423d4b55098adcdb4d1e7786fb719b410bd':
Don't leak process-private data on Change Local Name command.
Matthew Xie [Wed, 12 Feb 2014 22:21:38 +0000 (22:21 +0000)]
am
aa58a6cf: am
13a01401: Merge "Don\'t leak process-private data on Change Local Name command."
* commit '
aa58a6cffbdeed8381d309a4dcb093c42b29fd84':
Don't leak process-private data on Change Local Name command.
Matthew Xie [Wed, 12 Feb 2014 19:22:38 +0000 (19:22 +0000)]
am
13a01401: Merge "Don\'t leak process-private data on Change Local Name command."
* commit '
13a01401b8caaa28c407fe21726adc020cfb6169':
Don't leak process-private data on Change Local Name command.
Mike J. Chen [Wed, 12 Feb 2014 01:25:28 +0000 (01:25 +0000)]
am
dd7e09cf: am
17e40430: am
fab294ca: Merge "Fix bug introduced in last change. GATT define was after BLE_INCLUDED define." into klp-modular-dev
* commit '
dd7e09cf861b0da60a09dfabb665d4001d540bdf':
Fix bug introduced in last change. GATT define was after BLE_INCLUDED define.
Mike J. Chen [Wed, 12 Feb 2014 01:23:02 +0000 (01:23 +0000)]
am
17e40430: am
fab294ca: Merge "Fix bug introduced in last change. GATT define was after BLE_INCLUDED define." into klp-modular-dev
* commit '
17e4043072c6860a58c8da57a6e93685bd6bd14a':
Fix bug introduced in last change. GATT define was after BLE_INCLUDED define.
Mike J. Chen [Wed, 12 Feb 2014 01:20:03 +0000 (01:20 +0000)]
am
fab294ca: Merge "Fix bug introduced in last change. GATT define was after BLE_INCLUDED define." into klp-modular-dev
* commit '
fab294ca83b62c8e258f2bcd91018c31daf377bf':
Fix bug introduced in last change. GATT define was after BLE_INCLUDED define.
Mike J. Chen [Wed, 12 Feb 2014 01:16:10 +0000 (01:16 +0000)]
Merge "Fix bug introduced in last change. GATT define was after BLE_INCLUDED define." into klp-modular-dev
Mike J. Chen [Wed, 12 Feb 2014 01:15:21 +0000 (01:15 +0000)]
am
a89616ab: Fix bug introduced in last change. GATT define was after BLE_INCLUDED define.
* commit '
a89616ab221841dc6f053672e5e1226201dcf209':
Fix bug introduced in last change. GATT define was after BLE_INCLUDED define.
Mike J. Chen [Wed, 12 Feb 2014 00:23:31 +0000 (16:23 -0800)]
Fix bug introduced in last change. GATT define was after BLE_INCLUDED define.
The bug caused GATT_INCLUDED to be FALSE if not already defined.
Looks like the bluedroid codebase really can't support no GATT or SMP
if BLE_INCLUDED is true. Lots of compiler failures (not just warnings but
real failures) if GATT or SMP are not set to TRUE when BLE_INCLUDED is true.
Not cleaning that up at this time but just fixing basic build breakage.
Change-Id: I561cd4dcb2b39e90158ce74e55dca294f55cb540
Signed-off-by: Mike J. Chen <mjchen@google.com>
Mike J. Chen [Wed, 12 Feb 2014 00:23:31 +0000 (16:23 -0800)]
Fix bug introduced in last change. GATT define was after BLE_INCLUDED define.
The bug caused GATT_INCLUDED to be FALSE if not already defined.
Looks like the bluedroid codebase really can't support no GATT or SMP
if BLE_INCLUDED is true. Lots of compiler failures (not just warnings but
real failures) if GATT or SMP are not set to TRUE when BLE_INCLUDED is true.
Not cleaning that up at this time but just fixing basic build breakage.
Change-Id: I561cd4dcb2b39e90158ce74e55dca294f55cb540
Signed-off-by: Mike J. Chen <mjchen@google.com>
Mike J. Chen [Tue, 11 Feb 2014 23:32:35 +0000 (23:32 +0000)]
am
e80e4df1: am
b45e54f0: Merge "Fix warnings when BLE_INCLUDED is FALSE" into klp-modular-dev
* commit '
e80e4df1f9b9b60be37ef8ad854d7b46b9dbfb83':
Fix warnings when BLE_INCLUDED is FALSE
Mike J. Chen [Tue, 11 Feb 2014 23:32:35 +0000 (23:32 +0000)]
am
d382270c: am
ef04b8a2: Merge "Set some SMP and GATT to FALSE if BLE is FALSE" into klp-modular-dev
* commit '
d382270cede9188b3cd493cfd9fac093a5efdc9a':
Set some SMP and GATT to FALSE if BLE is FALSE
Mike J. Chen [Tue, 11 Feb 2014 23:29:03 +0000 (23:29 +0000)]
am
b45e54f0: Merge "Fix warnings when BLE_INCLUDED is FALSE" into klp-modular-dev
* commit '
b45e54f09e5496f298bb5c36354dff8ae27925f2':
Fix warnings when BLE_INCLUDED is FALSE
Mike J. Chen [Tue, 11 Feb 2014 23:29:02 +0000 (23:29 +0000)]
am
ef04b8a2: Merge "Set some SMP and GATT to FALSE if BLE is FALSE" into klp-modular-dev
* commit '
ef04b8a221df70ac205184c1f5098c409306556d':
Set some SMP and GATT to FALSE if BLE is FALSE
Mike J. Chen [Tue, 11 Feb 2014 23:24:33 +0000 (23:24 +0000)]
Merge "Fix warnings when BLE_INCLUDED is FALSE" into klp-modular-dev
Mike J. Chen [Tue, 11 Feb 2014 23:24:22 +0000 (23:24 +0000)]
Merge "Set some SMP and GATT to FALSE if BLE is FALSE" into klp-modular-dev
Matthew Xie [Tue, 11 Feb 2014 22:54:03 +0000 (22:54 +0000)]
Merge "Don't leak process-private data on Change Local Name command."
Sharvil Nanavati [Mon, 10 Feb 2014 06:21:48 +0000 (22:21 -0800)]
Don't leak process-private data on Change Local Name command.
The Change Local Name command to the controller doesn't zero out
the contents of the command buffer before setting the contents. As
a result, the controller sees both the new device name and then
random data left over on the heap from previous operations. This
random data may contain sensitive information. Although the device
name is null-terminated, a controller may inadvertently leak the
additional information over the network.
This change fixes the behaviour by zeroing out the entire command
buffer before copying the device name into it. If the device name
doesn't occupy the entire buffer, the rest of it will be zero.
http://b/
12949068
Change-Id: I4dd6ce6dd6709589fcd1e23f974db820820bc8e1
Mike J. Chen [Tue, 11 Feb 2014 21:51:29 +0000 (13:51 -0800)]
Fix warnings when BLE_INCLUDED is FALSE
Change-Id: I27c8ac6d5410159ed6bbef6aca5bf3b9692b0467
Signed-off-by: Mike J. Chen <mjchen@google.com>
Mike J. Chen [Tue, 11 Feb 2014 21:50:32 +0000 (13:50 -0800)]
Set some SMP and GATT to FALSE if BLE is FALSE
Also add asserts that SMP and GATT are never TRUE
if BLE is FALSE since that would make no sense.
Change-Id: I6530fe0255da0b237daa82b1b0dff286f2c5225d
Signed-off-by: Mike J. Chen <mjchen@google.com>
Mike J. Chen [Tue, 11 Feb 2014 21:17:06 +0000 (21:17 +0000)]
am
04276619: am
384d6211: Merge "Revert "Make all warnings fatal compilation errors"" into klp-modular-dev
* commit '
04276619fc4286c2b2926adfa5a5f87ab5065ee5':
Revert "Make all warnings fatal compilation errors"