PCH merge bootstrap failure on systems without flex

[Geoff Keating]

> Date: Wed, 5 Jun 2002 15:22:57 -0400 (EDT)
> From: "Kaveh R. Ghazi" <ghazi@caip.rutgers.edu>
> Cc: gcc-bugs@gcc.gnu.org, gcc@gcc.gnu.org
> 
> I'm getting bootstrap failures in stage1 because there's no flex on my
> box.
> 
>  > checking for flex... false
>  > [...]
>  > false  -o../../egcc-CVS20020605/gcc/gengtype-lex.c
>  > ../../egcc-CVS20020605/gcc/gengtype-lex.l \
>  >  || ( rm -f ../../egcc-CVS20020605/gcc/gengtype-lex.c && false )
>  > make[2]: *** [../../egcc-CVS20020605/gcc/gengtype-lex.c] Error 255
> 
> Are we now requiring flex as part of the bootstrap process?  Was this
> decision made knowingly or by accident?

I intentionally used flex.  I wasn't the first to introduce it into
mainline, though; that honour goes to Tim Josling who added
gcc/treelang/lex.l.  If I got the makefiles right, gengtype-lex.c
should be pregenerated in the release tarball, just like the uses of
bison.

Even before that, flex was required to build ld from CVS, so it's not
a new tool in the toolchain.

-- 
- Geoffrey Keating <geoffk@geoffk.org> <geoffk@redhat.com>