https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116400
--- Comment #14 from Iain Sandoe <iains at gcc dot gnu.org> ---
(In reply to Iain Sandoe from comment #13)
> (In reply to Francois-Xavier Coudert from comment #12)
> > Created attachment 61532 [details]
> > Regeneration script
> >
> > Attached is a shell script, to be placed in libgfortran/, that can be run in
> > that directory to regenerated all files. If people agree with this approach,
> > I would then remove the hackish "maintainer mode" code from
> > libgfortran/Makefile.am, leading to behavior that is identical to other GCC
> > runtime libraries:
> > - maintainer mode regenerates automake/autoconf files
> > - regeneration of library code is done manually by developers, through a
> > script
> >
> > I think this is definitely more robust. Let me know what you think.
> cool ...
> what might be nice is to share the list by including it in both the regen
> script and the Makefile... that seems like it should be possible?
In addition to sharing the input list
I'd think that the Makefile can have
-include $(srdir)/generated_sources_list
and the script can just
. generated_sources_list
====
1/
I think that the regen script should collect all the inputs into one list and
do something like:
for f in ${all_generated} ; do
$(M4) -Dfile=$f -I./m4 powu.m4 > $f.tmp
../move-if-changed $f.tmp $f
done
otherwise it's going to touch all the files?
2/ if this is called from the Makefile then we definitely need to use $(M4); I
guess the script can fallback to 'm4' if that is unset?
