On Thu, Jun 19, 2014 at 10:00:03PM -0600, Warner Losh wrote: > > On Jun 18, 2014, at 4:32 PM, Craig Rodrigues <rodr...@freebsd.org> wrote: > > > On Tue, Jun 17, 2014 at 11:59 PM, Dimitry Andric <d...@freebsd.org> wrote: > >> > >> See r257658 which had fixed this, but it was reverted by Warner, > >> unfortunately. This was one of the uses of the EARLY_BUILD flag which > >> was *not* redundant. :-/ > > > > > > Warner, > > > > Is there a way to bring back the behavior of r257658 but without > > the way it was implemented with the EARLY_BUILD flag? > > It would be handy, as we can see by this thread discussion. > > OK. I must be daft, or maybe just missing something. But I can build 9.3 > almost branch point on a current jail running on a 10.x system (to simulate > the 9 on current case). I don’t see the problem being talked about at all. As > for 9 on 10, I was also able to do a build, but my 10.x system may not have > been pristine. >
Removal of EARLY_BUILD is not the issue here, I have no idea where the hell the came into play. It is a race in the chain with what make(1) gets built for the stable/9 userland. It is why the 'make make buildworld' thing I mentioned works. Glen
pgpYeCFRFGObO.pgp
Description: PGP signature
