svn commit: r352520 - head/usr.sbin/pkg

[Ian Lepore]

On Tue, 2019-10-08 at 12:26 +0200, Emmanuel Vadot wrote:
>  Hi Glen,
> 
> On Thu, 19 Sep 2019 16:43:12 +0000 (UTC)
> Glen Barber <gjb at FreeBSD.org> wrote:
> 
> > Author: gjb
> > Date: Thu Sep 19 16:43:12 2019
> > New Revision: 352520
> > URL: https://svnweb.freebsd.org/changeset/base/352520
> > 
> > Log:
> >   Apply r346792 (cperciva) from stable/12 to head.  The original commit
> >   message:
> >   
> >    On non-x86 systems, use "quarterly" packages.
> >   
> >    x86 architectures have "latest" package builds on stable/*, so keep using
> >    those (they'll get switched over to "quarterly" during releases).
> >   
> >   The original commit was a direct commit to stable/12, as at the time it
> >   was presumed it would not be necessary for head.  However, when it is time
> >   to create a releng branch or switch from PRERELEASE/STABLE to BETA/RC, the
> >   pkg(7) Makefile needs further adjusting.  This commit includes those
> >   further adjustments, evaluating the BRANCH variable from release/Makefile
> >   to determine the pkg(7) repository to use.
> >   
> >   MFC after:	immediate (if possible)
> >   Sponsored by:	Rubicon Communications, LLC (Netgate)
> > 
> > Modified:
> >   head/usr.sbin/pkg/Makefile
> > 
> > Modified: head/usr.sbin/pkg/Makefile
> > ==============================================================================
> > --- head/usr.sbin/pkg/Makefile	Thu Sep 19 15:12:32 2019	(r352519)
> > +++ head/usr.sbin/pkg/Makefile	Thu Sep 19 16:43:12 2019	(r352520)
> > @@ -1,6 +1,16 @@
> >  # $FreeBSD$
> >  
> > +.if ${MACHINE} != "amd64" && ${MACHINE} != "i386"
> > +PKGCONFBRANCH?=	quarterly
> > +.else
> > +_BRANCH!=	${MAKE} -C ${SRCTOP}/release -V BRANCH
> > +BRANCH?=	${_BRANCH}
> > +. if ${BRANCH:MBETA*} || ${BRANCH:MRC*} || ${BRANCH:MRELEASE*}
> > +PKGCONFBRANCH?=	quarterly
> > +. else
> >  PKGCONFBRANCH?=	latest
> > +. endif
> > +.endif
> >  CONFS=		FreeBSD.conf.${PKGCONFBRANCH}
> >  CONFSNAME=	FreeBSD.conf
> >  CONFSDIR=	/etc/pkg
> 
>  Tier 2 (and weird tier1 like aarch64) only have latest for current so
> this doesn't work.
>  Also this depends on MACHINE and iirc MACHINE is always the host when
> cross compiling.
>  I think this need to be reverted.
> 
>  Cheers,
> 

MACHINE is the build host when you first launch make(1), but the
crossbuild code in Makefile.inc1 overrides MACHINE to TARGET (and
MACHINE_ARCH to TARGET_ARCH) and launches a sub-make that way.

-- Ian