Re: newlib: require autoconf-2.69

Mike Frysinger <[email protected]> Mon, 10 Jan 2022 19:58:47 -0500

Newsgroups	gmane.comp.lib.newlib
Message-ID	<YdzWR9EFxM6Rvsr5@vapier>

On 10 Jan 2022 10:02, Corinna Vinschen wrote:
> On Jan  7 14:09, Mike Frysinger wrote:
> > On 07 Jan 2022 10:58, Corinna Vinschen wrote:
> > > On Jan  6 01:10, Mike Frysinger wrote:
> > > > On 05 Jan 2022 20:46, Mike Frysinger wrote:
> > > > > do you want me to update newlib/ & libgloss/ to leverage this macro ?
> > > > 
> > > > looks like it's just as easy as passing -I to the top-level config dir when
> > > > running aclocal.  since override.m4 overrides AC_INIT, it forces the version
> > > > lock logic.  example diff for newlib/ is below.
> > > > 
> > > > lmk what you prefer here.
> > > 
> > > No worries.  I guess it's really better to stick to the same autoconf
> > > requirements as gcc/binutils-gdb.
> > 
> > is that an approval to push the override.m4 approach ? :)
> 
> That was an approval to push your original patch as is, but I'd like
> an approach to tweak and use override.m4 better, me thinks.  Do you
> want to submit a patch for this?

i've posted two patches at this point.  the first one in this thread which only
changed the AC_PREREQ lines, and a second one to leverage override.m4:
https://sourceware.org/pipermail/newlib/2022/018829.html
granted, it wasn't every single subdir, but that example patch is what i used
everywhere and the resulting diff was basically copied & pasted.

or do you want a patch that updates AC_PREREQ & uses override.m4 ?

or do you want a patch that deletes AC_PREREQ & uses override.m4 ? 
with the override.m4 logic, AC_PREREQ is completely redundant.
-mike

signature.asc (application/pgp-signature, 833 B)