On Tue, Aug 15, 2017 at 05:05:59PM +0200, Alessandro DE LAURENZIS wrote:
> Mike, all,
> 
> On Sat 12/08/2017 19:01, Alessandro DE LAURENZIS wrote:
> > Hello Mike,
> > 
> > > > root on sd0a (ff014e14e96d5c40.a) swap on sd0b dump on sd0b
> > > > WARNING: / was not properly unmounted
> > 
> > so: it tries to unpack a hibernated image, but then it is like the
> > previous hibernation didn't complete, since a fsck is forced.
> > 
> > But I was just reviewing this stuff, and can give some additional
> > information: the IntelDRM doesn't play any roles here; actually what I'm
> > observing is that an unmodified /bsd works flawlessly, instead the
> > hibernation "fails" after a modification with config(8); I just do the
> > following:
> > 
> > [....................snip....................]
> > sh> config -o /bsd.noulpt -e /bsd
> > OpenBSD 6.1-current (GENERIC.MP) #44: Thu Aug  3 12:12:07 MDT 2017
> >   [email protected]:/usr/src/sys/arch/amd64/compile/GENERIC.MP
> >   Enter 'help' for information
> >   ukc> disable ulpt*
> >   299 ulpt* disabled
> >   ukc> quit
> > [....................snip....................]
> > 
> > and then rebooting with /bsd.noulpt and hibernating/resuming, I see the
> > problem.
> > 
> > Does that make any sense?
> 
> after a discussion with semarie@ on #openbsd, the point here is that in
> /etc/rc we have:
> 
> ln -fh /bsd /bsd.booted
> 
> so in order to be compatible with current KARL/hibernation/resume mechanism,
> the booted kernel *must* be called "/bsd".
> 

Although this may be true, that's the default and IMO doesn't need to be
documented as such. But thanks nonetheless for clarifying this.

-ml

> I think a mention in https://www.openbsd.org/faq/current.html could be
> beneficial to other users, too.
> 
> Thanks for your time (and sorry again for the poor report).
> 
> -- 
> Alessandro DE LAURENZIS
> [mailto:[email protected]]
> LinkedIn: http://it.linkedin.com/in/delaurenzis

Reply via email to