Subject: Re: Why does mode_t need 32 bits?
To: Christos Zoulas <christos@deshaw.com>
From: Jim Wise <jimw@numenor.turner.com>
List: tech-kern
Date: 10/30/1997 14:31:42
-----BEGIN PGP SIGNED MESSAGE-----

On Thu, 30 Oct 1997, Christos Zoulas wrote:

> On Oct 30, 11:40am, gwr@mc.com ("Gordon W. Ross") wrote:
> -- Subject: Re: Why does mode_t need 32 bits?
> 
> | > From: Jason Thorpe <thorpej@nas.nasa.gov>
> | > Date: Wed, 29 Oct 1997 19:45:18 -0800
> | > 
> | [ Why 32-bits for mode_t and nlinks_t? ]
> | > 
> | > So we can actually build lint libraries.
> | 
> | I'm still confused.  Why does lint care about this?
> 
> Because it is much easier to deal with integral types being passed
> to library functions rather than promotion cases.

Surely this isn't the only reason?  I find it hard to believe that we
are changing our API because our lint is broken...

- --
				Jim Wise
				jim.wise@turner.com

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 5.0i for non-commercial use
Charset: noconv

iQEVAwUBNFjgpRg+dMhCouwfAQGLlQf9EQHN6GH6+RkIxj7gUszuSZN1haf33KmP
ZnBbQpEQnzxTi0coCO/xpA6adZLpJ9ZR0vgotji7qtmE9FmfMYTF5M0L5J9fSaXD
fUbsoOLNweu7jaQKhF4GolRA4LCn/dVRkCXySFsERIljYRIIbUkzzLKwGV7fsFII
QuuCnCBG95ARjdKbnXyyY5JLQuhn7Nj4Ud9MGJMty9gSBFTsb0u+N1FReeZ/MBMq
f18a82md9y8Mn13b+2lUl8W5x5h05jU5NM8GcvtoB7xgyQBYJQA9E5VBIy96M9ll
NLrSIx23j3/kINuO7e6T4zgz5Q2dPbc0hKO+6n4ZDQThdnBOadmqKg==
=u/xh
-----END PGP SIGNATURE-----