cvs commit: src/sys/conf kern.post.mk kmod.mk

[M. Warner Losh]

In message: <200311141635.hAEGZqiT031464 at green.bikeshed.org>
            "Brian F. Feldman" <green at FreeBSD.org> writes:
: "M. Warner Losh" <imp at bsdimp.com> wrote:
: > In message: <200311141604.hAEG4BCg041862 at repoman.freebsd.org>
: >             Brian Feldman <green at FreeBSD.org> writes:
: > :   Include opt_global.h in the modules build, when building from a normal
: > :   kernel build.  This makes it possible for me not to get pissed off that
: > :   random.ko crashes the system trying to rdtsc() when the i386/cpu.h
: > :   support code decides it's okay to call that op when neither I386_CPU or
: > :   I486_CPU is defined.  I guess it also makes WITNESS/INVARIANTS defines
: > :   get picked up by the modules.
: > 
: > I've been polishing similar things that I'd hoped to get into this
: > release.  I'll try to finish them up today.  You should really talk to
: > people before doing these things.
: 
: I know that you have been working on that, and I'm very happy to see that.  
: If this makes things difficult of course there's no reason not to remove it 
: other than being an interim solution for crashes people might see now.  If I 
: didn't know you were working on the smarter module build, I would have just 
: gone through all the headers and tried to remove all the "harmful" uses of 
: constants defined by opt_global.h (which I feel is very much not a 
: "solution", but either way... getting rid of the problem is important, and I 
: don't think that the importance is particularly diminished just because the 
: bugs have existed for a while, IMHO).

This actually is a good part of the 'first step' that I've been
wanting to get done.  It just suprised me, that's all, to see it
committed the day we're trying to freeze for 5.2.

Warner