[desktop] Bluetooth is always enabled after reboot even if it was disabled

[Alexey Kulik]

If bluetooth was disabled using the switcher in the bluetooth indicator, after reboot it becomes enabled.

ProblemType: Bug
DistroRelease: Ubuntu 14.04
Package: indicator-bluetooth 0.0.6+14.04.20140207-0ubuntu1
ProcVersionSignature: Ubuntu 3.13.0-18.38-generic 3.13.6
Uname: Linux 3.13.0-18-generic x86_64
ApportVersion: 2.13.3-0ubuntu1
Architecture: amd64
CurrentDesktop: Unity
Date: Sat Mar 22 22:34:52 2014
InstallationDate: Installed on 2014-01-27 (54 days ago)
InstallationMedia: Ubuntu 14.04 LTS "Trusty Tahr" - Alpha amd64 (20140124)
SourcePackage: indicator-bluetooth
UpgradeStatus: No upgrade log present (probably fresh install)
upstart.indicator-bluetooth.log: (bluetooth-wizard:2251): Gtk-DEBUG: Connecting to session manager

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Alberto Salvia Novella (es20490446e)]

It's a non-ideal default configuration.

[Tamir (sugip)]

Situation really has not changed (if not off from the context menu of the indicator, and bluetooth settings menu).
 Alberto, I can give information from my laptop, but do not know how best to do it.

[Zero (zeugentor)]

Before installing of skype it seems to be fixed. But if you set up skype the bug is back

[Nazar Mokrynskyi (nazar-pc)]

I'm suffering from tis bug too, I do not use bluetooth on laptop, why this is so difficult to remember this?

[Renato Araujo Oliveira Filho (renatofilho)]

This is happening on ubuntu phone too.

[Mathieu Trudel-Lapierre (cyphermox)]

This isn't up to the indicator. The indicator can toggle the killswitches, but something else needs to persist the settings across a reboot.

On phone, that's urfkill.

On computers, that's the rfkill-store/rfkill-restore jobs in the rfkill package.

I'm setting the tasks for those, but things may change during the development cycle to make it all managed by urfkill, which is the "right" way to fix this for computers, and will still need a bit more work on phone.

[Tony Espy (awe)]

Just tried this on a recent image ( #183 ) on mako, and I can reproduce it.

It looks like the bluetooth indicator doesn't talk to urfkill directly. When I disable BT via the indicator, the rfkill state is indeed toggled, as if I run 'rfkill list', it shows as soft-blocked.

[Tony Espy (awe)]

Confirmed on krillin RTM #72 with the urfkill changes currently in RTM silo-024.

That said, it's not currently InProgress by either myself or Mathieu, thus 2014-10-02 is over-optimistic. As such I'm pushing to the following weeks' milestone.

[Tony Espy (awe)]

So I think this bug should be split into two different bugs.

First, the original reporter filed this on an x86 machine running Desktop, not Touch. As urfkill is still not used on the Desktop as of the latest utopic images, I'm splitting this into a urfkill bug for the problems described by myself and Renato on phones, and a rfkill bug for the desktop issue ( although this could actually be a driver/hardware-specific bug vs. a bug in the rfkill package ).

Bug #1376063 has been created to track the urfkill problem seen with the phone images and the urfkill task in this bug marked Invalid. All phone-related tags have been copied to the new bug and removed from this bug.

[Mathieu Trudel-Lapierre (cyphermox)]

I'm keeping this open because some fixes will still need to happen in both rfkill and urfkill for this to work properly; just like there will be at least the need to also start shipping urfkill by default on desktop before this works better.

One of the issues is that rfkill ships scripts that currently work fairly well for many systems, but don't work on others because of how killswitches work on these systems. Once urfkill starts shipping on desktop, these scripts will also need to be removed.

I'm updating the bug title to denote clearly that this is specifically a desktop bug; there is a different bug open for touch issues: bug 1376063.

[Mathieu Trudel-Lapierre (cyphermox)]

Now, one issue we'll need to clear up is exactly which killswitches are available on the system where this bug appears.

Could you please attach the output of 'rfkill list' to this bug report, along with the output of 'lsmod' so we know what drivers are used by the kernel?

Thanks!

[Tamir (sugip)]

ON Ubuntu 14.04 turned off and after a reboot, but in Ubuntu 14.10 on again

[Nazar Mokrynskyi (nazar-pc)]

Just ping to fix this annoying bug finally, should not be difficult.

[Joschi Poschi (joschiposchi)]

In Ubuntu 15.04 the bug is still persistent.

[Michael Weimann (m982)]

This still happens on 15.10

[Joan CiberSheep (cibersheep)]

I'm experiencing this bug in 16.04 (updated from 15.10) on a slimbook:
 CPU Intel® Core™ i5-5200U CPU @ 2.20GHz × 4

How?
Turn off bluetooth from indicator
Reboot
Indicator several times shows bluetooth is on

I don't know how to send a useful log.

Also in Ubuntu touch OTA10.1 happens getting out of flight mode or reboot.

[spike speigel (frail-knight)]

I'm experiencing this in 16.04 LTS as well...

$ uname -a
Linux blackhole 4.10.0-19-generic #21~16.04.1-Ubuntu SMP Fri Apr 7 08:20:02 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux

[spike speigel (frail-knight)]

@cibersheep

This might be the solution. Worked for me:

https://github.com/blueman-project/blueman/issues/682