Subject: Re: workaround for FreeBSD binary that hardcodes a device major (was: Re: CVS commit: src/sys/conf)
To: matthew green <mrg@eterna.com.au>
From: Bill Studenmund <wrstuden@netbsd.org>
List: tech-kern
Date: 09/24/2006 00:36:23
--ZPt4rx8FFjLCG7dd
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sun, Sep 24, 2006 at 08:13:46AM +0100, David Laight wrote:
> On Sun, Sep 24, 2006 at 03:13:31PM +1000, matthew green wrote:
> >    2) Use a MD major for twa, and patch freebsd_sys_mknod() so that 187
> >    gets translated into the right major
>=20
> What happens when 187 is allocated for a real device as well?
> How can only some of the accesses get translated?

I think the idea is that we only care about attempts to generate a node=20
187 from FreeBSD-emulated code. 187 won't be used by anything else on the=
=20
FreeBSD emulation side of things.

I think we're only talking about mapping mknod requests. Everything else=20
will be using NetBSD node values.

Take care,

Bill

--ZPt4rx8FFjLCG7dd
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (NetBSD)

iD8DBQFFFjV3Wz+3JHUci9cRAnAqAKCYq9ic+Wd7CGSWVO+f3v6EiMTnYwCeLaa7
+jJQNLnosH0ACrHyHKB+VOA=
=qcpR
-----END PGP SIGNATURE-----

--ZPt4rx8FFjLCG7dd--