Re: config.guess manufacturer string?

[Ralf Corsepius <rc040203 freenet de>]

Dan Nicholson wrote:
> On Wed, Feb 18, 2009 at 9:13 AM, Stepan Kasal <skasal redhat com> wrote:
> > Actually, if you care about the host triplet used for rpm builds,
> > that's something where config.guess is not directly involved.
> >
> > Most spec files call %configure.  That macro calls configure with
> > option --build.  With that option given, configure does not call
> > config.guess.
> >
> > So it is possible that you would be satisfied if the %configure macro
> > used
> >  ./configure  --build=i686-redhat-linux-gnu --host=i686-redhat-linux-gnu
> >
> > To achive that, it is sufficient to modify the macro %{_host}.
> > (Macro %{_host_vendor} should probably be modified as well.)
> >
> > No, I do not see the reason for doing this.  Should all Linux
> > distributions do the same?
>
> I think most distributions do set the --build correctly for how
> they've configured their systems. Regardless of whether the redhat is
> there or not, %configure should be setting --build and --host at least
> the same as --target.
Which is a bug. %configure should not set any of these.

> Instead, rpm passes the values that _it_ was
> built with as --build and --host.
>
> $ rpm -E%configure | grep -e --build -e --host -e --target
>   ./configure --build=x86_64-unknown-linux-gnu --host=x86_64-unknown-linux-gnu \
>         --target=x86_64-redhat-linux-gnu \
>
> The advantage of setting --build correctly is that autoconf macros
> will use test against the triplet will work as expected. For instance
> AC_PROG_CC will search for gcc with the host triplet prefixed.
Which is a side-effect originating from the bug above.

configure scripts should be using the uncanonicalized versions of *_TOOLS.

Ralf