This is the mail archive of the
gcc-prs@gcc.gnu.org
mailing list for the GCC project.
Re: bootstrap/6825: [Sun OS 4.1.4] gcc 3.1 fails to build
- From: Wolfgang Bangerth <bangerth at ticam dot utexas dot edu>
- To: nobody at gcc dot gnu dot org
- Cc: gcc-prs at gcc dot gnu dot org,
- Date: 5 Dec 2002 23:26:00 -0000
- Subject: Re: bootstrap/6825: [Sun OS 4.1.4] gcc 3.1 fails to build
- Reply-to: Wolfgang Bangerth <bangerth at ticam dot utexas dot edu>
The following reply was made to PR bootstrap/6825; it has been noted by GNATS.
From: Wolfgang Bangerth <bangerth@ticam.utexas.edu>
To: jason andrade <jason@dstc.edu.au>
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: bootstrap/6825: [Sun OS 4.1.4] gcc 3.1 fails to build
Date: Thu, 5 Dec 2002 17:23:06 -0600 (CST)
> the problem seems to be because gcc3 doesn't ship with (or sunos doesn't have?)
> definitions for a bunch of functions that it thinks it should find, e.g for
> div_t and so on.
Sorry, but I think I can't help you here.
> > > ./configure --prefix=/opt/local/stow/gcc-3.2.1 --enable-languages=c,c++
> > > and
> > > make -j2 CFLAGS=-O2 LIBCFLAGS=-O2 LIBCXXFLAGS=-O2 -fno-implicit-templates bootstrap-lean
> >
> > First, since this failed, could you retry by building in a separate
> > build-directory, as the installation instructions recommend? I.e., if
> > gcc-3.2.1 is you src dir, then make a parallel directory build and call
> > ../gcc-3.2.1/configure --...
> > make bootstrap
>
> i'll try that. will it make much of a difference ?
Maybe. It is supposed to work to build in-tree, but it does not get much
testing, and it seems to provoke all kinds of strange and hard to track
down errors. The recommended version that most of the developers use is to
build in a separate build-dir.
> > Also, what happens if you don't specify CFLAGS et al? As far as I know the
> > stage2 build already uses optimization flags, so there is no need to also
> > use them for the first two stages (apart from possibly tripping your
> > bootstrapping compiler).
>
> haven't tried without the cflags. at this stage i am assuming the problem
> is specifically with function definition issues for c++, which wouldn't
> be influenced by the cflags (but please let me know if i am wrong)
Some optimization flags switch on builtins and/or select different math
functions. I don't know whether it makes a difference in your case, but it
would rule out one possibility for breakage.
> when i get a little more time, i am going to try and find all the missing
> functions on another more up to date OS and see if i can reuse them on
> sunos4.
I think this would be appreciated!
Regards
Wolfgang
-------------------------------------------------------------------------
Wolfgang Bangerth email: bangerth@ticam.utexas.edu
www: http://www.ticam.utexas.edu/~bangerth