On Sat, 11 Feb 2017 13:08:23 +0900 Nobuhiro Iwamatsu <iwama...@nigauri.org> wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA256 > > Control: tags 853207 + moreinfo > > Hi, > > > Most likely the bluetooth support for your laptop is broken under linux. > > > > The bluetooth.service file contains > > > > ConditionPathIsDirectory=/sys/class/bluetooth > > > > which means that systemd only tries to start the bluetooth daemon when > > the bluetooth module from the kernel is loaded which creates the above > > directory in sysfs. The bluetooth module is in turn only loaded by the > > kernel/udev when bluetooth hardware is detected. So the .service file > > behaves correctly as to only start the daemon when there is actual > > bluetooth hardware available. > > > > Could you check if the above reasoning is correct by first directly > > loading the bluetooth kernel module (modprobe bluetooth), which should > > create /sys/class/bluetooth and then trying to start the bluetooth > > daemon (systemctl restart bluetooth.service)? > > > > If that is the case the bug should be reassigned as a wishlist bug to > > systemd. If a condition for starting a service is not satisfied when a > > user explicitly tries to explicitly start it on the command line, it > > would be useful to report the reason for the startup failure to the > > user. This would significantly help a user to debug such an issue. > > > > Matthias, thanks for your support. > I confirmed this bug with stretch, but the same problem did not happen. > > Russell, could you confirm again with Matthias's comment?
Cc'ing Russel, as apparently you didn't Cc him. Cheers, Emilio