Date: Fri, 5 Apr 2002 17:37:38 +0100 From: Josef Karthauser <joe@tao.org.uk> To: Greg 'groggy' Lehey <grog@FreeBSD.org> Cc: "J. Mallett" <jmallett@FreeBSD.org>, Dag-Erling Smorgrav <des@ofug.org>, Kenneth Culver <culverk@alpha.yumyumyum.org>, Matthew Dillon <dillon@apollo.backplane.com>, Mike Barcroft <mike@FreeBSD.org>, "M. Warner Losh" <imp@village.org>, jake@locore.ca, cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: Compiler cache (was: style(9) bike shed, 371st edition (was: cvs commit: src/sys/i386/i386 critical.c src/sys/i386/include cpufunc.h critical.h src/sys/i386/isa apic_vector.s icu_vector.s src/sys/kern kern_fork.c kern_proc.c kern_switch.c src/sys/alpha/alph)) Message-ID: <20020405163738.GA72974@genius.tao.org.uk> In-Reply-To: <20020405135843.S93816@wantadilla.lemis.com> References: <20020402154308.F46914-100000@alpha.yumyumyum.org> <xzplmc5u6rp.fsf@flood.ping.uio.no> <20020403100131.Q26122@wantadilla.lemis.com> <xzp7knoq591.fsf@flood.ping.uio.no> <20020404120805.U44513@wantadilla.lemis.com> <20020405034604.A3087@FreeBSD.ORG> <20020405135843.S93816@wantadilla.lemis.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--0F1p//8PRICkK4MW Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Fri, Apr 05, 2002 at 01:58:43PM +0930, Greg 'groggy' Lehey wrote: > On Friday, 5 April 2002 at 3:46:05 +0000, J. Mallett wrote: > > On Thu, Apr 04, 2002 at 12:08:05PM +0930, Greg 'groggy' Lehey wrote: > >> > >> On a similar note, Andrew Tridgell has written a compiler cache which > >> claims to be able to speed up things by recognizing already-compiled > >> sources. I'm a little sceptical myself, but everybody who has tried > >> it says it's very useful. http://ccache.samba.org/ for more details. > > > > Why don't we look into precompiled headers, thus avoiding analysis of > > well, crap? >=20 > That wouldn't do the same thing. The cache apparently contains object > files. If it detects preprocessor output identical to something it > has compiled before, it just grabs the object file out of the cache. >=20 > I'm dubious too. But it seems that the Samba and Apache people love > it. I suppose it would help when you have multiple source trees, all > of which are mainly the same. A mate of mine is doing a very similar thing. It works wonders if you've got loads of people all 'make world'ing on the same cluster of machines - say with shared file space. There's no point rebuilding an object if someone did so a few minutes ago. It's probably more use in a multideveloper local environment rather than for individuals. Joe --0F1p//8PRICkK4MW Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.6 (FreeBSD) Comment: For info see http://www.gnupg.org iEYEARECAAYFAjyt0tIACgkQXVIcjOaxUBaGXgCdHfVEfLqDhblCIqonkAAt3H3a 42YAoMBgribOgZlRKT7OGWQqTgfKJDuk =DFOR -----END PGP SIGNATURE----- --0F1p//8PRICkK4MW-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020405163738.GA72974>