Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 27 Apr 2006 10:06:21 -0700
From:      Brooks Davis <brooks@one-eyed-alien.net>
To:        Scott Long <scottl@samsco.org>
Cc:        src-committers@FreeBSD.org, Pawel Jakub Dawidek <pjd@FreeBSD.org>, Brooks Davis <brooks@one-eyed-alien.net>, cvs-src@FreeBSD.org, cvs-all@FreeBSD.org, John-Mark Gurney <jmg@FreeBSD.org>
Subject:   Re: cvs commit: src/sys/kern sched_4bsd.c
Message-ID:  <20060427170621.GC12971@odin.ac.hmc.edu>
In-Reply-To: <4450F84D.4080601@samsco.org>
References:  <200604261942.k3QJgdnb086990@repoman.freebsd.org> <20060427102026.GC20828@garage.freebsd.pl> <20060427163351.GB12971@odin.ac.hmc.edu> <4450F84D.4080601@samsco.org>

next in thread | previous in thread | raw e-mail | index | archive | help

--gr/z0/N6AeWAPJVB
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Apr 27, 2006 at 10:58:53AM -0600, Scott Long wrote:
> Brooks Davis wrote:
> >On Thu, Apr 27, 2006 at 12:20:26PM +0200, Pawel Jakub Dawidek wrote:
> >
> >>On Wed, Apr 26, 2006 at 07:42:39PM +0000, John-Mark Gurney wrote:
> >>+> jmg         2006-04-26 19:42:38 UTC
> >>+>=20
> >>+>   FreeBSD src repository
> >>+>=20
> >>+>   Modified files:
> >>+>     sys/kern             sched_4bsd.c=20
> >>+>   Log:
> >>+>   move remaining sysctl into the kern.sched tree...
> >>[...]
> >>+> -SYSCTL_INT(_kern, OID_AUTO, ccpu, CTLFLAG_RD, &ccpu, 0, "");
> >>+> +SYSCTL_INT(_kern_sched, OID_AUTO, ccpu, CTLFLAG_RD, &ccpu, 0, "");
> >>[...]
> >>
> >>Too fast.
> >>
> >># grep -r kern.ccpu /usr/src/
> >
> >
> >We should retain the old location as well under COMPAT_FREEBSD[456].  It
> >doesn't cost anything and we should treat the sysctl tree like any other
> >API to the extent possible.
>=20
> Well, sysctl's are a second class API.  The ease by they can be changed,=
=20
> as well as the completely uncontrolled nature that they are developed=20
> in, means that we really start to hamstring ourselves if we declare them
> as a first class API.  What we should do instead is define a subset of
> the sysctl tree that is stable and guaranteed to exist, and declare that
> the rest are transient and therefore not part of the API.  Doing this by
> popular vote is impractical, what we need is someone to sit down and=20
> audit what is there now and make appropriate recommendations.  sysctls=20
> that are used by libkvm are a good candidate, for example, but random
> driver sysctls might not be.

I agree we don't want to be too restrictive here, but for cases like
this where there's effectively no cost I think we should at least try
to retain compatibility under #ifs.  If we break that compatibility at
some later date due to actually changing the data or because something
else needs to use that location in the tree, that's probably OK.

-- Brooks

--=20
Any statement of the form "X is the one, true Y" is FALSE.
PGP fingerprint 655D 519C 26A7 82E7 2529  9BF0 5D8E 8BE9 F238 1AD4

--gr/z0/N6AeWAPJVB
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQFEUPoMXY6L6fI4GtQRAndlAJ4gXRoEjYEgrTXABLmbOOyMVxK57wCgpJlD
BNK8OUVugyAQlkHb/IQ9QFc=
=e/14
-----END PGP SIGNATURE-----

--gr/z0/N6AeWAPJVB--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060427170621.GC12971>