Am 10.03.2010 18:33, schrieb Andrey Borzenkov: > Unfortunately, I still experience occasional lockups which require me to > press power off button. What I find annoying - NM consistently comes up > disabled after this. Simple "cnetworkmanager -o yes" is enough, but I am > curious - why? Is it intentional design decision or should it be > considered a bug? How NM decides it should be disabled in the first > place? It's a bug. When you are going into suspend state, pm-utils will send NM a signal to disable all interfaces. This state is written to /var/lib/NetworkManager/NetworkManager.state. Upon (successful) resume, pm-utils will tell NM again, to reenable all devices. If your system crashes on resume, NetworkManager.state will contain stale information. Iirc, Dan proposed to solve this, using two different kind of disabled states: user initiated (which should be preserved on reboots) and automatic/system state. The latter could probably simply be put into /var/run, which would ensure it is cleaned up on boot. Cheers, Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth?
Attachment:
signature.asc
Description: OpenPGP digital signature