---------- Forwarded message ----------
From: Thibaud Fabre <[email protected]>
Date: 2010/6/20
Subject: Re: [Info] LFS Temp System - GCC 4.5.0 with MPFR > 2.4.2
To: LFS Developers Mailinglist <[email protected]>


2010/6/20 Matthew Burgess <[email protected]>

On Thu, 17 Jun 2010 11:07:19 +0200, Thibaud Fabre <[email protected]>
> wrote:
> > GCC compilation will fail even WITH the
> > gmp/mpc/mpfr sources present in gcc source folder when make tries to run
> > mpfr-configure. Copying the two missing GMP source files into
> > /gcc-build/gmp will fix this.
>
> This certainly sounds like an upstream bug.  It may be related to this
> change:
>
> "The detection of a GMP build directory (more precisely, the internal
> header
> files of GMP) was previously done separately from the use of the
> --with-gmp-build
> configure option. This was not consistent with the documentation and with
> other
> parts of the configure script. So, as of MPFR 3.0.0, the internal header
> files of
> GMP are now used if and only if the --with-gmp-build configure option is
> given"
>
> So, could you try passing '--with-gmp-build=/mnt/lfs/sources/gcc-4.5.0/gmp'
> to
> GCC's configure invocation in chapter 5 please?  I'm hoping that gets
> passed
> through to MPFR's configure!
>

Unfortunately, it does not. Make still fails with the same error :

checking for gmp internal files... configure: error: header files gmp-impl.h
> and longlong.h not found
> make[1]: *** [configure-mpfr] Error 1
> make[1]: Leaving directory `/mnt/lfs/sources/gcc-build'
> make: *** [all] Error 2
>

Configure command was :

../gcc-4.5.0/configure \
>     --target=$LFS_TGT --prefix=/tools-test \
>     --disable-nls --disable-shared --disable-multilib \
>     --disable-decimal-float --disable-threads \
>     --disable-libmudflap --disable-libssp \
>     --disable-libgomp --enable-languages=c \
>
>     --with-gmp-build=/mnt/lfs/sources/gcc-4.5.0/gmp
>

I also tried with relative path to gmp source, with the same result.

I tried compiling using --with-gmp-include (as suggested as suggested by V.
Lefevre on http://gcc.gnu.org/ml/gcc/2010-06/msg00280.html), then with
--with-gmp, successfully with both options. However, I have not tested the
gcc build any further, and I'm not sure how to do so (unless I keep going to
chap 5.7 to run the dummy test), so I can't tell if there are any
shortcomings to compiling with these options, and from the build log, it
seems that mpfr does not look for gmp-impl.h and longlong.h anymore.

I will get back to you to confirm whether or not the dummy.c test succeeds
with those builds.

Best regards,
Thibaud F.
-- 
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page

Reply via email to