Date: Fri, 29 Sep 2006 17:43:32 +0400 From: Ruslan Ermilov <ru@FreeBSD.org> To: Robert Watson <rwatson@FreeBSD.org> Cc: current@FreeBSD.org Subject: Re: lockf in installworld -- not a good idea Message-ID: <20060929134332.GD4776@rambler-co.ru> In-Reply-To: <20060929141709.E70454@fledge.watson.org> References: <20060929141709.E70454@fledge.watson.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--Km1U/tdNT/EmXiR1
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Fri, Sep 29, 2006 at 02:20:06PM +0100, Robert Watson wrote:
>=20
> >revision 1.72
> >date: 2005/11/03 08:56:39; author: ru; state: Exp; lines: +1 -0
> >Serialize access to the info/dir file; needed for parallel installs.
> >
> >Reported by: scottl
> >
> >I'm not very fond of using the non-standard lockf(1) here, but I
> >have no better idea at the moment. NetBSD uses ln(1) to create a
> >lock file, but this approach can result in a deadlock if make is
> >interrupted, leaving an orphaned lock file.
>=20
> I'm not sure why this suddenly showed up in my configuration, but this is=
=20
> preventing me from installing world on an NFS mounted file system without=
=20
> rpc.lockd running. I get the following:
>=20
> =3D=3D=3D> lib/libcom_err/doc (install)
> lockf -k /usr/share/info/dir install-info --quiet =20
> --defsection=3D"Programming &
> development tools." --defentry=3D"* libcom_err: (com_err). A Comm=
on=20
> Error
> Description Library for UNIX." com_err.info /usr/share/info/dir
> lockf: cannot open /usr/share/info/dir: Operation not supported
> *** Error code 73
>=20
> Stop in /zoo/rwatson/netperf/RELENG_6/src/lib/libcom_err/doc.
> *** Error code 1
>=20
> Stop in /zoo/rwatson/netperf/RELENG_6/src/lib/libcom_err.
> *** Error code 1
>=20
The necessity to run rpc.lockd is documented in the build(7)
manpage. Quote:
: installworld Install everything built by a preceding buildworld step
: into the directory hierarchy pointed to by make(1) vari-
: able DESTDIR.
:=20
: If installing onto an NFS file system, make sure that
: rpc.lockd(8) is running on both client and server. See
: rc.conf(5) on how to make it start at boot time.
> I've noticed an increasing intolerance in our tools for system install an=
d=20
> maintenance to locking not being implemented over the past few years. I =
no=20
> longer get working cron on boxes with neither rpc.lockd nor local locking=
=20
> enabled, for example. Installworld represents a bigger problem, since I=
=20
> don't want to have to depend on a completely working rpc chain in order t=
o=20
> installworld, nor depend on running in what would effectively be multiuse=
r=20
> mode. Surely there's a better fix for this than adding lockf use?
>=20
I don't know of a better fix. Another approach is that mentioned
in the commit log and used by NetBSD. I tried it, and it was very
fragile -- it could easily leave the temporary file around, and
that would stuck forever another instances.
The problem at hand is that multiple instances of install-info(1)
are attempting to write to the ${DESTDIR}/usr/share/info/dir file.
If you have a better idea, don't hesitate to let me know. I'd
very much like to get rid of that as well.
Cheers,
--=20
Ruslan Ermilov
ru@FreeBSD.org
FreeBSD committer
--Km1U/tdNT/EmXiR1
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (FreeBSD)
iD8DBQFFHSMEqRfpzJluFF4RAsKRAKCZx4lQziJRibkA15786ZmqHsB+zgCeI77P
SNAAg2xagRiUMTtwlr11gnY=
=mXpp
-----END PGP SIGNATURE-----
--Km1U/tdNT/EmXiR1--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060929134332.GD4776>