This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: [RFA]: Top-level configure patch to build Cygwin native newlib
- From: jjohnstn <jjohnstn at redhat dot com>
- To: Corinna Vinschen <vinschen at redhat dot com>
- Cc: gdb-patches at sources dot redhat dot com, <newlib at sources dot redhat dot com>, <binutils at sources dot redhat dot com>
- Date: Wed, 16 Jun 2004 13:47:33 -0400 (EDT)
- Subject: Re: [RFA]: Top-level configure patch to build Cygwin native newlib
On Wed, 16 Jun 2004, Corinna Vinschen wrote:
> On Jun 16 12:23, jjohnstn wrote:
> > Hi Corinna,
> >
> > I have no problem with the patch, but would it be possible to add
> > to the patch to issue a warning message to the user to tell them that
> > winsup is missing?
>
> No, that shouldn't be a problem. In that case the
>
> test -d winsup && skipdirs=`echo " ${skipdirs} " | sed -e 's/ target-newlib / /'`
>
> would become a
>
> if test -d winsup
> then
> skipdirs=`echo " ${skipdirs} " | sed -e 's/ target-newlib / /'`
> else
> echo "Warning: winsup is missing so newlib can't be build."
> fi
>
Thanks, just change "build" to "built".
>
> If that's fine with everyone...
>
> Erm... btw., should I Cc the gcc-patches list, perhaps?
>
It would be a good idea. IIRC, most of the top-level files are
copied by sources from gcc. Did I get that correct, Chris?
>
> Corinna
>
>
>
> >
> > -- Jeff J.
> >
> > On Wed, 16 Jun 2004, Corinna Vinschen wrote:
> >
> > > Hi,
> > >
> > > perhaps I could just apply the patch but I would like to ask people first
> > > if there something is in the way of applying the below patch.
> > >
> > > The idea is the following: If you have a source tree containing the newlib
> > > subdir but the source tree does not contain the winsup subdir, then it's
> > > impossible to build a native Cygwin newlib. The reason is, that building a
> > > Cygwin native newlib requires a bunch of header files from the winsup/cygwin
> > > directory. If the winsup directory is missing, the build will break when
> > > trying to build newlib.
> > >
> > > For that reason, the below patch to configure.in checks if the winsup
> > > directory is available and removes newlib from the target_configdirs
> > > if winsup is missing.
> > >
> > > Is that ok to apply?
> > >
> > >
> > > Corinna
> > >
> > >
> > > * configure.in: Don't build Cygwin native newlib if winsup
> > > directory is missing.
> > > * configure: Regenerate.
> > >
> > >
> > > Index: configure.in
> > > ===================================================================
> > > RCS file: /cvs/cvsfiles/gnupro/configure.in,v
> > > retrieving revision 1.32
> > > diff -p -u -r1.32 configure.in
> > > --- configure.in 9 Jun 2004 17:31:01 -0000 1.32
> > > +++ configure.in 16 Jun 2004 15:19:40 -0000
> > > @@ -548,8 +548,8 @@ case "${target}" in
> > > *-*-cygwin*)
> > > target_configdirs="$target_configdirs target-libtermcap target-winsup"
> > > noconfigdirs="$noconfigdirs target-gperf target-libgloss ${libgcj}"
> > > - # always build newlib.
> > > - skipdirs=`echo " ${skipdirs} " | sed -e 's/ target-newlib / /'`
> > > + # always build newlib if winsup directory is present.
> > > + test -d winsup && skipdirs=`echo " ${skipdirs} " | sed -e 's/ target-newlib / /'`
> > >
> > > # Can't build gdb for Cygwin if not native.
> > > case "${host}" in
> > >
> > >
> > >
>
>