Hi, On Sun, Sep 27, 2015 at 09:01:07PM +0530, Ritesh Raj Sarraf wrote: > That said, I'm not sure what you meant by systemd killing forked > processes. Me and Michael Biebl exchanged emails when he was helping me > with systemd integration for LMT. So, earlier, LMT was invoked through > udev. And udev had never guaranteed to allow a long running process to > complete. > > So to overcome that buggy approach, we now have /lib/udev/lmt-udev, > which if it detects is running on systemd, will invoke systemd. And > systemd, afaik, does not have any such behavior for programs that it is > invoking by itself.
I'm not sure if it kills forked children of processes invoked via ExecStart in a one-shot service (although I rather suspect it does), but it certainly does for things in /lib/systemd/system-sleep/. For some reason I replaced "systemctl restart laptop-mode" with direct invocation of "laptop_mode auto force". I think it was not behaving well, but can't remember. :-( I'll try to put systemctl restart there instead and remove the wait and see if things keep working or if something breaks. -- Tomáš Janoušek, a.k.a. Pivník, a.k.a. Liskni_si, http://work.lisk.in/
