Closed Bug 1142007 Opened 9 years ago Closed 6 years ago

Bluedroid function |get_remote_service_record| doesn't work

Categories

(Firefox OS Graveyard :: Bluetooth, defect)

Product:
Firefox OS Graveyard
Component:
Bluetooth
Platform:
All
Gonk (Firefox OS)
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

(Not tracked)

Status:
RESOLVED WONTFIX

People

(Reporter: tzimmermann, Unassigned)

References

Details

Attachments

(1 obsolete file) 

Calling |get_remote_service_record| of the Bluedroid API does not return a service record. The returned UUID id 0, the channel is -1. The logcat contains the following error message.

> E/        (  207): ### ASSERT : external/bluetooth/bluedroid/main/../btif/src/btif_dm.c line 1621 unhandled remote service record event (4) ###
Here's a stack trace from where this error is reported.

#0  tgkill () at bionic/libc/arch-arm/bionic/tgkill.S:46
#1  0xb6e8e13c in pthread_kill (t=<optimized out>, sig=6) at bionic/libc/bionic/pthread_kill.cpp:49
#2  0xb6e8e350 in raise (sig=6) at bionic/libc/bionic/raise.cpp:32
#3  0xb6e8d086 in __libc_android_abort () at bionic/libc/bionic/abort.cpp:55
#4  0xb6e9c9a0 in abort () at bionic/libc/arch-arm/bionic/abort_arm.S:41
#5  0xb6c7ea7c in btif_dm_remote_service_record_evt (event=<optimized out>, 
    p_param=0xb729591e "\300\030\205\334\331\237Ͷ\234\377(\267̯\340\266\204\255Ӷ\234\377(\267\f")
    at external/bluetooth/bluedroid/main/../btif/src/btif_dm.c:1622
#6  0xb6c7d658 in btif_context_switched (p_msg=0xb7295910) at external/bluetooth/bluedroid/main/../btif/src/btif_core.c:189
#7  btif_task (params=<optimized out>) at external/bluetooth/bluedroid/main/../btif/src/btif_core.c:349
#8  0xb6ca6726 in gki_task_entry (params=3068215700) at external/bluetooth/bluedroid/gki/./ulinux/gki_ulinux.c:157
#9  0xb6e8822c in __thread_entry (func=0xb6ca66c9 <gki_task_entry>, arg=0xb6e14194 <gki_pthread_info+12>, tls=0xb6766dd0)
    at bionic/libc/bionic/pthread_create.cpp:105
#10 0xb6e883c4 in pthread_create (thread_out=0xb6e12df8 <gki_cb+8>, attr=<optimized out>, start_routine=0xb6ca66c9 <gki_task_entry>, arg=0x78)
    at bionic/libc/bionic/pthread_create.cpp:224
#11 0x00000000 in ?? ()
The final HCI messages on my notebook are listed below. It looks like the request for the service record has been received and responded correctly by the computer.

> ACL data: handle 12 flags 0x02 dlen 38
    L2CAP(d): cid 0x0040 len 34 [psm 1]
        SDP SSA Req: tid 0x0 len 0x1d
          pat uuid-128 00001105-0000-1000-8000-00805f9b34fb
          max 240
          aid(s) 0x0000 - 0xffff
          cont 00
< ACL data: handle 12 flags 0x00 dlen 115
    L2CAP(d): cid 0x0044 len 111 [psm 1]
        SDP SSA Rsp: tid 0x0 len 0x6a
          count 103
          record #0
              aid 0x0000 (SrvRecHndl)
                 uint 0x10006
              aid 0x0001 (SrvClassIDList)
                 < uuid-16 0x1105 (OBEXObjPush) >
              aid 0x0004 (ProtocolDescList)
                 < < uuid-16 0x0100 (L2CAP) > <
                 uuid-16 0x0003 (RFCOMM) uint 0x9 > <
                 uuid-16 0x0008 (OBEX) > >
              aid 0x0005 (BrwGrpList)
                 < uuid-16 0x1002 (PubBrwsGrp) >
              aid 0x0009 (BTProfileDescList)
                 < < uuid-16 0x1105 (OBEXObjPush) uint 0x100 > >
              aid 0x0100 (SrvName)
                 str "OBEX Object Push"
              aid 0x0303 (SuppFormatsList)
                 < uint 0x1 uint 0x2 uint 0x3 uint 0x4 uint 0x5 uint 0x6 uint 0xff >
          cont 00
> ACL data: handle 12 flags 0x02 dlen 12
    L2CAP(s): Disconn req: dcid 0x0040 scid 0x0044
< ACL data: handle 12 flags 0x00 dlen 12
    L2CAP(s): Disconn rsp: dcid 0x0040 scid 0x0044
> HCI Event: Number of Completed Packets (0x13) plen 5
    handle 12 packets 2
Michael,

I can trigger this problem when I call |get_remote_service_record|. I give a correct UUID and device address to the function, but it always returns an error (i.e., UUID of 0, channel of -1, empty record name). Comment 2 indicates that the lookup from the remote device worked correctly.

Could this be a driver problem? Have you seen this before?
Flags: needinfo?(mvines)
We've not seen this crash in our builds yet, no.  I assume this is on a Flame?  Can you try on a v2.2 reference device?
Flags: needinfo?(mvines)
It's actually not a crash. The function simply doesn't return a meaningful result. We didn't use it until now, so the problem never showed up.

I prepared this path to trigger the function. The STR is

  - open Settings app and pair with Notebook/PC
  - open Camera app and take a picture
  - from within Camera app, send picture to paired BT device

Expected result

  - both devices will start the transmission and copy the picture

Actual result

  - there's a Gaia message that the transfer started, but nothing happens

The logcat will show the error message. and with the patch applied, Gecko should print lines like these

> I/GeckoBluetooth(  206): RemoteDevicePropertiesNotification: virtual void mozilla::dom::bluetooth::BluetoothServiceBluedroid::RemoteDevicePropertiesNotification(mozilla::dom::bluetooth::BluetoothStatus, const nsAString_internal&, int, const mozilla::dom::bluetooth::BluetoothProperty*):1587 00001105-0000-1000-8000-00805f9b34fb
> I/GeckoBluetooth(  206): RemoteDevicePropertiesNotification: virtual void mozilla::dom::bluetooth::BluetoothServiceBluedroid::RemoteDevicePropertiesNotification(mozilla::dom::bluetooth::BluetoothStatus, const nsAString_internal&, int, const mozilla::dom::bluetooth::BluetoothProperty*):1610 00000000-0000-0000-0000-000000000000

The first line shows the UUID of the service (OPP in this case) that was given to |get_remote_service_record|. The second line is the UUID in the returned service record. It's always zero. I observed this with in-Gecko and bluetoothd backends, and also checked the UUID structure when it comes out of Bluedroid.

Since my PC's HCI data indicates that the request was handled correctly, I assume that there's a problem in the Bluedroid driver or something's wrong in the way I call the function. BTW, the patch explicitly calls |get_remote_services| first to make sure that the local data/state is up-to-date.
Attachment #8577200 - Flags: feedback?(mvines)
How does this impact v2.2?  I don't see anything on the bug attributes that tie it to our current efforts.  Providing such a link would really help to prioritize.
We don't use this function in v2.2, so this version is not affected.
Comment on attachment 8577200 [details] [diff] [review]
Lookup service channel in Bluedroid's OPP manager

Review of attachment 8577200 [details] [diff] [review]:
-----------------------------------------------------------------

(Sry, hit me up for post v2.2 stuff later in the year)
Attachment #8577200 - Flags: feedback?(mvines)
See Also: → 1170509
No longer blocks: 1141616
Attachment #8577200 - Attachment is obsolete: true
Firefox OS is not being worked on
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: