[Bug libstdc++/86450] Bootstrap failure due to -Wabi

[sgk at troutmask dot apl.washington.edu]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86450

--- Comment #16 from Steve Kargl <sgk at troutmask dot apl.washington.edu> ---
On Tue, Jul 10, 2018 at 09:40:16PM +0000, redi at gcc dot gnu.org wrote:
> 
> >  There is an obvious patch.
> > I will commit my patch by the end of the day,
> > or revert the patch causing the problem.
> 
> OK. But gfortran being unable to cope with any warnings elsewhere in gcc
> without crying P1-blocker is the real problem, you really should address that.
> 

I cannot control other developers, who commit patches,
that lack adequately testing.  I am unsure how to address
the issue other than raising a PR.  Being unable to
bootstrap, certainly seems to be a blocking problem.

Breaking the tree (and it seems that this issues has been
known for a couple of weeks) and turning a blind eye seems
to be rather unfriendly to other cwcontributors.

PS: The description of --enable-maintainer-mode does not
state that -Werror is added to the options.  Why have
an option that "regenerate[s] the Autoconf and Automake
output files" if it cannot not be used?