cvs commit: src/sys/compat/linux linux_ioctl.c

[Poul-Henning Kamp]

In message <20050324210504.GL46915 at wombat.fafoe.narf.at>, Stefan Farfeleder wri
tes:
>On Thu, Mar 24, 2005 at 03:55:59PM -0500, John Baldwin wrote:
>> On Thursday 24 March 2005 03:19 pm, Brooks Davis wrote:
>> > On Thu, Mar 24, 2005 at 09:09:43PM +0100, Stefan Farfeleder wrote:
>> > > On Thu, Mar 24, 2005 at 12:02:10PM -0800, Brooks Davis wrote:
>> > > > Prior to C99, I think #error wasn't specified.  C99 does specify it.
>> > >
>> > > C89/90 had #error too.
>> >
>> > I stand corrected.  Now that I think about it, all but the perverse
>> > preprocessors would implement #error just by blowing up on an unknown
>> > directive so even that isn't an issue.  In any case, I've made the
>> > change and the new code is at least shorter.  It's really only there for
>> > people mucking with kernel internals who should be able to figure out
>> > what it means.
>> 
>> We could always just change CTASSERT() to use #error if that is better.
>
>Unfortunately you can't use preprocessor directives in the replacement
>list of a macro.

And you can't use all sorts of repressions in #if either.

that's why CTASSERT() is such a magic beast.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.