Subject: [cmu.andrew.internet.computing.netbsd.tech-kern] Re: More config changes, for ro source tree and disjoint build trees
To: Jason Thorpe <thorpej@nas.nasa.gov>
From: None <rvb@IGW.TRUST.CS.CMU.EDU>
List: tech-kern
Date: 08/13/1996 14:20:32
------- Start of forwarded message -------
On Mon, 12 Aug 1996 16:10:43 -0700
Jason Thorpe <thorpej@nas.nasa.gov wrote:
> # Machine-independent kernel options
> option compat_12 # binary compatibility with NetBSD 1.2
> [ . . . ]
> This would emit a file "option_compat_12.h" which would contain one of
> two things:
> #define COMPAT_12 0
> - or -
> #define COMPAT_12 1
#include <option_compat_12.h>
> #if (COMPAT_12 > 0)
> [ . do something . ]
This is exactly what mach did, except we used "options" as the keyword
and we did not prefix the generated file with "option_". I suppose
this could prevent some confusion, but the kernel hacker would always
have to remember to prepend "option_" to an option name. The question
is how likely is it that the header file namespace would conflict with
the option name namespace.