Ahoj, Dňa Sat, 2 Aug 2014 18:21:48 +0200 Slavko <linux@slavino.sk> napísal: > Ahoj, > > Dňa Sat, 2 Aug 2014 12:43:27 +0100 Brian <ad44@cityscape.co.uk> > napísal: > > > On Sat 02 Aug 2014 at 10:49:32 +0100, Brian wrote: > > > > This is on a minimal Jessie with systemd-sys installed and nouveau. > > You were right about systemd not using pm-utils, but it does give a > > second testing route. > > > > root@debian:~# /lib/systemd/systemd-sleep --version > > systemd 208 > > +PAM +LIBWRAP +AUDIT +SELINUX +IMA +SYSVINIT +LIBCRYPTSETUP > > +GCRYPT +ACL +XZ > > As i mention, i am on the 204 still (because 208 prevent to use the > SysV): > > /lib/systemd/systemd-sleep --version > systemd 204 > +PAM +LIBWRAP +AUDIT +SELINUX +IMA +SYSVINIT +LIBCRYPTSETUP +GCRYPT > +ACL +XZ > > > Suspending with > > > > /lib/systemd/systemd-sleep suspend > > > > works. Resuming also works here. journalctl has the full story. > > Because i afraid if i will remove all nvidia's things properly, i > tried it partialy - i end the X session, stop the DM and remove the > nvidia module (i hope that this is enough to eliminate the suspend & > nvidia interaction). When i make sure, that the nvidia module is not > loaded i try to hibernate: > > systemctl hibernate > > System hibernates and then after some small time it resume without any > visible problems. Similarly as you, i can see records in the > journalctl. After this i tried to suspend it: > > systemctl suspend > > System suspends, but don't resume and after forced (by button) reboot > i cannot see any records in the journalctl (i suspend it at cca > 17:46): > > journalctl --until=17:50 > -- Logs begin at So 2014-08-02 17:51:06 CEST, end at So 2014-08-02 > 18:01:01 CEST. -- OK, i did small progress :-) The issue is (seems) not related to systemd, it is caused by switching from pm-utils to kernel's PM. I follow Debugging hibernation and suspend at [1], where i did tests via /sys/power/pm_test. The first i try the "freezer", which works, then i try the "devices" option, where the same problems happen, then it seems. IIUC, the problem is with suspending the devices. Now the question is: which device? Any help is welcomed :) [1] https://www.kernel.org/doc/Documentation/power/basic-pm-debugging.txt -- Slavko http://slavino.sk
Attachment:
signature.asc
Description: PGP signature