This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: 2.95,2 bug; can not create cross compiler
- To: Franz Sirl <Franz dot Sirl-kernel at lauterbach dot com>
- Subject: Re: 2.95,2 bug; can not create cross compiler
- From: Jonathan Larmour <jlarmour at cygnus dot co dot uk>
- Date: Thu, 18 Nov 1999 20:13:26 +0000
- CC: Mike Stump <mrs at windriver dot com>, gcc-bugs at gcc dot gnu dot org, gcc at gcc dot gnu dot org, schweitz at nortelnetworks dot com
- Organization: Cygnus Solutions
- References: <4.2.2.19991118205920.04500390@mail.lauterbach.com>
Franz Sirl wrote:
>
> At 20:44 18.11.99 , Mike Stump wrote:
> > > Date: Thu, 18 Nov 1999 16:40:36 GMT
> > > To: mrs@windriver.com
> > > From: jlarmour@cygnus.co.uk (Jonathan Larmour)
> >
> > > Isn't this what --without-newlib is for?
> >
> >I don't understand the relevance of the question. No is the only
> >answer I can imagine. Or put another way, I am not using newlib, and
> >I don't want to inhibit_libc as I have one. I don't want to lie to
> >the compiler and claim I have newlib, and currently, --without-newlib
> >is how it my tree is configured and that is the default for me.
>
> He probably means --with-newlib and is referring to the original problem.
> If you just want a naked crosscompiler with no target libraries or headers
> around (and no newlib), this will define you inhibit_libc (see
> gcc/configure). --with-newlib probably should be duplicated with a more
> sensible name like --no-target-stuff :-).
Sorry, yes. And my other point being that using autoconf to automatically
determine whether to define inhibit_libc or not would not be the right thing
to do (which was the particular bit in Mike's message I was following up
to).
Jifl