Errmm...... okay, yes, I guess that makes sense.


On Mon, 15 Oct 2001, Manuel Bouyer wrote:

> On Sun, Oct 14, 2001 at 02:19:45PM -0700, Matthew Jacob wrote:
> >
> > Really? Are you sure this is sensible? The completion thread *will* start
> > eventually, right?
>
> Yes, but registering a callback will also freeze the adapter queue.
> We may deadlock in autoconfig because of this, no ?
> I think doing this avoid checking SCSIPI_CHAN_TACTIVE in every driver ...
>
> >
> >
> > On Mon, 15 Oct 2001, Manuel BOUYER wrote:
> >
> > >
> > > Module Name:	syssrc
> > > Committed By:	bouyer
> > > Date:		Sun Oct 14 21:17:42 UTC 2001
> > >
> > > Modified Files:
> > > 	syssrc/sys/dev/scsipi: scsipi_base.c
> > >
> > > Log Message:
> > > Refuse to register a callback if the completion thread isn't started yet.
> > >
> > >
> > > To generate a diff of this commit:
> > > cvs rdiff -r1.60 -r1.61 syssrc/sys/dev/scsipi/scsipi_base.c
> > >
> > > Please note that diffs are not public domain; they are subject to the
> > > copyright notices on the relevant files.
> > >
> > >
> --
> Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
>      {Net,Free}BSD: 23 ans d'experience feront toujours la difference
> --
>