Subject: Re: CVS commit: [elad-kernelauth] src/sys/ufs/ext2fs
To: Christos Zoulas <christos@zoulas.com>
From: Jason Thorpe <thorpej@shagadelic.org>
List: source-changes
Date: 04/20/2006 09:31:41
On Apr 20, 2006, at 9:17 AM, Christos Zoulas wrote:
> The problem with this approach is that it needs <sys/kauth.h> to be
> present in all the header files that use the type. Unfortunately this
No, the problem is that prototypes and other structures that use a
kauth_cred_t are being exposed to userland. If you eliminate that,
then you eliminate the need for userland to know anything kauth_cred-
anything, and you only include <sys/kauth.h> #ifdef _KERNEL.
> is a problem with the `c' language. There are no opaque types; for
> the purist, the best you can do is to use a struct pointer which is
> what the rest of the kernel does for all the data structures.
>
> So you have the following choices:
>
> 1. Revolting hacks [actual code snippet]:
>
> #ifndef _UFS_FFS_FFS_EXTERN_H_
> #define _UFS_FFS_FFS_EXTERN_H_
>
> #ifndef HAVE_NBTOOL_CONFIG_H
> #include <sys/kauth.h>
> #else
> typedef void *kauth_cred_t;
> #endif /* !HAVE_NBTOOL_CONFIG_H */
>
> 2. Include <sys/kauth.h> everywhere and install it in userland.
>
> 3. Use "struct kauth_cred *" instead of "kauth_cred_t"
>
> 4. Rewrite the kernel in a different language.
>
> I choose [3]; In fact, I have almost finished fixing all the header
> files
> and adding explicit '#include <sys/kauth.h>' in the files that need it
> rather than including <sys/kauth.h> in <sys/resource.h> and <sys/
> file.h>
> so that everyone gets it by side-effect.
>
> christos
-- thorpej