Bluetooth stopped working after kernel update 4.4.0-91

[Doug Morse]

Bluetooth connection to Bose Soundlink II have been fine for years. Applied a few new updates in the last day or two --- system was generally up to date, I don't often wait more than a week or so --- and now I can no longer connect. Removed the device and re-paired just fine, although the speaker did not connect as it usually does after the pairing (i.e., it stayed in pairing mode).

Messages in syslog:

Aug 15 20:32:07 n1 bluetoothd[1075]: connect error: Connection refused (111)
Aug 15 20:32:07 n1 bluetoothd[1075]: GLib: Source ID 276 was not found when attempting to remove it
Aug 15 20:32:11 n1 bluetoothd[1075]: connect error: Connection refused (111)
Aug 15 20:32:42 n1 bluetoothd[1075]: connect error: Connection refused (111)
Aug 15 20:32:42 n1 bluetoothd[1075]: GLib: Source ID 287 was not found when attempting to remove it
Aug 15 20:32:46 n1 bluetoothd[1075]: connect error: Connection refused (111)
Aug 15 20:33:07 n1 bluetoothd[1075]: connect error: Host is down (112)
Aug 15 20:33:07 n1 bluetoothd[1075]: GLib: Source ID 305 was not found when attempting to remove it

ProblemType: Bug
DistroRelease: Ubuntu 16.04
Package: bluez 5.37-0ubuntu5
ProcVersionSignature: Ubuntu 4.4.0-91.114-generic 4.4.76
Uname: Linux 4.4.0-91-generic x86_64
ApportVersion: 2.20.1-0ubuntu2.10
Architecture: amd64
CurrentDesktop: Unity
Date: Tue Aug 15 20:34:34 2017
InstallationDate: Installed on 2015-12-26 (597 days ago)
InstallationMedia: Ubuntu 15.10 "Wily Werewolf" - Release amd64 (20151021)
InterestingModules: rfcomm bnep btusb bluetooth
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-4.4.0-91-generic root=UUID=5950fbba-cde1-49ac-a918-04e517894c57 ro quiet splash vt.handoff=7
SourcePackage: bluez
UpgradeStatus: Upgraded to xenial on 2017-02-28 (168 days ago)
dmi.bios.date: 01/18/2017
dmi.bios.vendor: Intel Corporation
dmi.bios.version: RYBDWi35.86A.0362.2017.0118.0940
dmi.board.name: NUC5i5RYB
dmi.board.vendor: Intel Corporation
dmi.board.version: H40999-504
dmi.chassis.type: 3
dmi.modalias: dmi:bvnIntelCorporation:bvrRYBDWi35.86A.0362.2017.0118.0940:bd01/18/2017:svn:pn:pvr:rvnIntelCorporation:rnNUC5i5RYB:rvrH40999-504:cvn:ct3:cvr:
hciconfig:
 hci0: Type: BR/EDR Bus: USB
  BD Address: DC:53:60:97:55:17 ACL MTU: 1021:5 SCO MTU: 96:6
  UP RUNNING PSCAN
  RX bytes:6896 acl:46 sco:0 events:294 errors:0
  TX bytes:4254 acl:32 sco:0 commands:156 errors:0
mtime.conffile..etc.bluetooth.main.conf: 2015-12-29T07:24:08.439240

[Daniel van Vugt (vanvugt)]

If you're using 16.04 then there have been no high-level bluetooth-related updates at all recently:

 bluez 5.37-0ubuntu5 ... 2016-03-09
 pulseaudio 1:8.0-0ubuntu3.3 ... 2017-06-30

So if your system was working for the previous month or more then the problem is not caused by bluez or pulseaudio at least.

You might want to assign it to "linux (Ubuntu)" in case it was caused by a recent kernel update like:

 linux 4.4.0-91.114 ... 2017-08-07

which sounds more likely by the date.

[Daniel van Vugt (vanvugt)]

You can try previous kernel versions your system had prior to the problem, using this technique:
  https://wiki.ubuntu.com/RecoveryMode

No need to select "recovery mode" but other older kernel versions should still be listed for you to use.

[Doug Morse (dm-dougmorse)]

Thanks Daniel. Rolling back to kernel 4.4.0-87-generic #110-Ubuntu SMP seemed to fix the problem. Upgrading to 4.4.0-91 also seemed to cause other problems: My system started overheating, eventually slowed to a crawl (i.e., over a matter of minutes after each reboot slowed to the point of being unresponsive), and I saw 4 instances (one for each core, I presume) of kidle_interrupt, which I had never seen before.

I notice that 4.4.0-91 is no longer showing up as an update in the repos for 16.04.

For what it's worth, I also downgraded libgd3 to version 2.1.1-4build2, as it also upgraded around the same time as my recent kernel updates (-87 and -89, but I don't think I ever rebooted into -89).

[Ubuntu Kernel Bot (ubuntu-kernel-bot)]

This change was made by a bot.

[Thomas Schwinge (tschwinge)]

I may be running into the same issue as reported here. Has any progress been made, or any workaround been found?