Re: Patch to ask if user has egcs/pgcc / add me to credits

Niels Kristian Bech Jensen (nkbj@image.dk)
Sat, 27 Jun 1998 08:50:49 +0200 (CEST)


On Sat, 27 Jun 1998, James Michael Mastros wrote:

> On Sat, Jun 27, 1998 at 07:35:51 AM +0200, Niels Kristian Bech Jensen wrote:
> > On Fri, 26 Jun 1998, James Michael Mastros wrote:
> >
> Because you defined CFLAGS based on HOSTCFLAGS. The two are unrelated.
> Perhaps I'm building a kernel for a 386 on a Pentium.
>
> I would have no problem with:
> CFLAGS = -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer
> HOSTCFLAGS = -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer
>
I see your point. I've changed it in my patch.

> > > [arch/i386/Makefile]
> > > > -
> > > > -ifdef SMP
> > > > -CFLAGS := $(CFLAGS) -D__SMP__
> > > > endif
> > > You just killed SMP as we know it.
> >
> > Defined twice. That's why I remove it.
> Whops... Then why don't I see it twice when (accedently, in my case)
> compiling a SMP kernel?
>
I don't know. The same lines are in the toplevel Makefile, so -D__SMP__
should be seen twice in the compiler flags. It is on my system.

[...]
> > Why. __SMP__ isn't used in any of the files built by this Makefile. It's
> > defined in the toplevel Makefile anyway.
> Many makefiles, including arch/i386/Makefile explicitly set CFLAGS themselves.
>
But they all include CFLAGS from the previous level.

I've used the patch for both UP and SMP kernels without problems.

--
Med venlig hilsen / Best regards
                                                          nkbj@image.dk
Niels Kristian Bech Jensen                   http://www.image.dk/~nkbj/

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu