Re: [gentoo-alt] Restructuring the bootstrap (or: getting rid of gcc 4.2)

[Ruud Koolen]

On Monday 03 February 2014 10:16:09 Fabian Groffen wrote:
> On 02-02-2014 20:35:28 +0100, Fabian Groffen wrote:
> > Let me see if I can build latest gcc on Solaris for you on a vanilla
> > system.
>
> On an OpenIndiana box (that already comes with a more decent toolchain
> than e.g. Solaris 10):
>
> [ra:/export/gentoo/temp]64 % uname -a
> SunOS ra 5.11 oi_151a7 i86pc i386 i86pc Solaris
> [ra:/export/gentoo/temp]64 % tar -jxf
> /export/gentoo/distfiles/gcc-4.7.3.tar.bz2
> [ra:/export/gentoo/temp]64 % env -i /bin/bash -l
> bash: keychain: command not found
> bash-4.0$ pwd
> /export/gentoo/temp
> bash-4.0$ cd gcc-4.7.3/
> bash-4.0$ ./configure
> checking build system type... i386-pc-solaris2.11
> checking host system type... i386-pc-solaris2.11
> checking target system type... i386-pc-solaris2.11
> checking for a BSD-compatible install... /bin/ginstall -c
> checking whether ln works... yes
> checking whether ln -s works... yes
> checking for a sed that does not truncate output... /usr/gnu/bin/sed
> checking for gawk... gawk
> checking for libitm support... yes
> checking for gcc... gcc
> checking for C compiler default output file name... a.out
> checking whether the C compiler works... yes
> checking for suffix of object files... o
> checking whether we are using the GNU C compiler... yes
> checking whether gcc accepts -g... yes
> checking for gcc option to accept ISO C89... none needed
> checking for g++... g++
> checking whether we are using the GNU C++ compiler... yes
> checking whether g++ accepts -g... yes
> checking for gnatbind... no
> checking for gnatmake... no
> checking whether compiler driver understands Ada... no
> checking how to compare bootstrapped objects... cmp --ignore-initial=16
> $$f1 $$f2
> checking for objdir... .libs
> checking for the correct version of gmp.h... no
> configure: error: Building GCC requires GMP 4.2+, MPFR 2.3.1+ and MPC
> 0.8.0+.
> Try the --with-gmp, --with-mpfr and/or --with-mpc options to specify
> their locations.  Source code for these libraries can be found at
> their respective hosting sites as well as at
> ftp://gcc.gnu.org/pub/gcc/infrastructure/.  See also
> http://gcc.gnu.org/install/prerequisites.html for additional info.  If
> you obtained GMP, MPFR and/or MPC from a vendor distribution package,
> make sure that you have installed both the libraries and the header
> files.  They may be located in separate packages.
>
> Basically the journey already ends here.
>
> I went a bit through gmp and mpfr, but they depend on eachother, and
> this is basically where the bootstrap process is carefully crafted the
> way it is, because stuff depends on stuff and gmp for instance compiles,
> but creates a 64-bits library (for some reason), while plain invoked
> configure for gcc will target a 32-bits compiler, so it will break down
> the line.
>
> Fabian

Sure, because you used neither the CFLAGS and LDFLAGS the bootstrap script 
uses to find the right library versions, nor all the wizardry that 
toolchain.eclass does to build the right toolchain. If you build things the 
way the bootstrap script does, using portage, things should work.

Case in point: after some unrelated fixes to the bootstrap script (mostly 
regarding baselayout-prefix, which you seem to have fixed independently for 
the mainline bootstrap a few days ago) my bootstrap now works flawlessly on 
solaris 10 x86 (using gcc 3.4), as well as SLES :)

I'll test on a couple of other solaris-ish versions (OpenIndiana, solaris 11, 
maybe solaris sparc if I can get my hands on one) and then it's on to tougher 
platforms!

-- Ruud