Date: Tue, 21 Mar 2006 23:56:53 +0200 From: Ruslan Ermilov <ru@freebsd.org> To: Doug Barton <dougb@freebsd.org> Cc: freebsd-current@freebsd.org, rwatson@freebsd.org, Steve Kargl <sgk@troutmask.apl.washington.edu> Subject: Re: mergemaster broken? Message-ID: <20060321215653.GA77045@ip.net.ua> In-Reply-To: <44206E9F.6020702@FreeBSD.org> References: <20060320221741.GA13699@troutmask.apl.washington.edu> <20060320230009.GA55254@troutmask.apl.washington.edu> <20060320231604.GH83781@ip.net.ua> <20060320232746.GI83781@ip.net.ua> <20060320234006.GV35129@funkthat.com> <20060320234145.GB33003@troutmask.apl.washington.edu> <20060321064528.GD77849@ip.net.ua> <44206E9F.6020702@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--fdj2RfSjLxBAspz7
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Tue, Mar 21, 2006 at 01:22:39PM -0800, Doug Barton wrote:
> I'm including rwatson here since the MACHINE_ARCH stuff was his idea.
>=20
MACHINE_ARCH should be spelled TARGET_ARCH when doing cross
stuff properly.
> Ruslan Ermilov wrote:
>=20
> >> On Mon, Mar 20, 2006 at 03:40:06PM -0800, John-Mark Gurney wrote:
> >>> Should we also document that -m is suppose to be src's etc dir instead
> >>> of src? I've accidentally pointed -m at src, and then it does a make
> >>> which is quite ammuzing as it's completely the wrong thing... Or now
> >>> that we call outside of /etc, should we make -m really point to src,
> >>> and have the proper calls add etc to the directory?
>=20
> I strongly dislike the idea of changing the semantics of the -m option. I=
t's
> been the way it is since day 1, and I really hate to make changes to
> something like that. I can see a case for making the man page more clear,
> but I'd rather work around the problem with -m than change the semantics.
>=20
> > Doesn't really matter, mergemaster(8) was broken because it was
> > written when we didn't have correct wrappers for "distrib-dirs"
> > and "distribution", upgrade and cross-arch friendly, at the top
> > level.
> >=20
> > Anyway, attached is the patch I'd like to commit after a nod
> > from Doug. It fixes mergemaster(8) to use src/Makefile wrappers
> > for distrib-dirs and distribution targets, and makes it use
> > TARGET_ARCH instead of faking up MACHINE_ARCH, now that it uses
> > the correct wrappers. It also makes ${SOURCEDIR} and -m point
> > at the src/ top, as documented in a manpage.
>=20
> Forgive me if I'm being dense here, but why do the changes you describe
> require that we run make in src/?
>=20
You didn't read this thread from beginning, did you? OK, I will
explain it in a bit more detail.
The time you're running "mergemaster -p", you run src/etc/Makefile
which may be incompatible with installed make(1) binary and/or
/usr/share/mk files (like is the case for this thread). By using
the top level wrappers, all these problems are solved. Wrappers
execute src/etc/Makefile targets in upgrade- and cross-friendly
environment, with bootstrapped make(1) binary and src/share/mk
files.
> Or, alternatively, if it is _absolutely_
> necessary to do so, why do we have to redefine SOURCEDIR to be src, and w=
hy
> can't we just strip /etc from SOURCEDIR where needed?
>=20
My original diff didn't change that, I think someone else
asked to change it, and it was pretty logical (see below).
> In short, I have no objections to fixing mergemaster to work with the new
> world order, but it needs to be done in a way that does not change semant=
ics
> of an existing option. I'd also like confirmation from Robert that we're =
not
> breaking any of the behavior that he added by doing it the way you propos=
e.
>=20
OK, you have my patches, do what you want with them. I don't
care which directory the -m option uses, but there's not much
sense keeping it pointing to src/etc/ if we're going to use
the top-level Makefile and its targets. Absolutely what needs
to be done:
1) Fix mergemaster(8) to use top-level targets for distrib-dirs,
distribute, and obj/all in "etc". The latter should be more
obvious -- my changes fixes it so it runs obj/all in etc in
the same environment as buildworld does, i.e., using adequate
tools.
2) s/MACHINE_ARCH/TARGET_ARCH/, i.e., don't pretend we're a
TARGET_ARCH. When needed top-level makefiles will
substitute MACHINE_ARCH=3D${TARGET_ARCH}.
3) Grow support for TARGET for pc98, better through the
"arch[:machine]" format of the -A option's argument.
Some reference commits from src/Makefile:
: revision 1.320
: date: 2005/07/15 14:38:54; author: ru; state: Exp; lines: +1 -1
: Make the `distrib-dirs' target official and working properly
: when cross-building (installing). (This is useful for NFS
: installing world/kernel to another architecture machine.)
:=20
: revision 1.317
: date: 2005/02/27 12:11:34; author: ru; state: Exp; lines: +2 -1
: Added the convenience "distribution" target which calls the
: target of the same name from src/etc/Makefile with a proper
: environment, suitable to be used during upgrades and cross-
: builds.
Cheers,
--=20
Ruslan Ermilov
ru@FreeBSD.org
FreeBSD committer
--fdj2RfSjLxBAspz7
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (FreeBSD)
iD8DBQFEIHakqRfpzJluFF4RAsRFAKCcOh8QPmZBsN06q0N2JSY3wojxFwCeI2LU
pU/Mm3YQlv1EcepqWqGCjjc=
=WO8a
-----END PGP SIGNATURE-----
--fdj2RfSjLxBAspz7--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060321215653.GA77045>