[sr #111380] autoconf-2.72.90 C vs C++ test failure with Oracle Studio 12.6

--8323329-424238335-1770219544=:1412222
Content-Type: TEXT/plain; CHARSET=utf-8
Content-Transfer-Encoding: QUOTED-PRINTABLE
Content-Disposition: inline

Update of sr #111380 (group autoconf):=0A=0A                Priority:      =
 5 - Unprioritized =3D> 7 - Release N (Desirable)=0A                  Statu=
s:                    None =3D> Confirmed=0A             Assigned to:      =
              None =3D> zackw=0A=0A    ____________________________________=
___________________=0A=0AFollow-up Comment #1:=0A=0AIn the 2.70 release run=
up I found a bunch of bugs where macros would succeed=0Aunder AC_LANG(C) bu=
t fail _incorrectly_ under AC_LANG(C++), because the test=0Aprogram didn't =
stick to the intersection of C and C++.  That's why=0AAT_CHECK_MACRO requir=
es the C and C++ compilers to give the same result.=0A=0AIn this case, it l=
ooks like AC_C_FLEXIBLE_ARRAY_MEMBER _should_ give a=0Adifferent result for=
 AC_LANG(C++) than AC_LANG(C), so it should be added to=0Athe list, in test=
s/mktests.pl, of macros not to test with AT_CHECK_MACRO, and=0Awe should ad=
d a hand-written test of this macro to semantics.at.  I'll do that=0Alater =
today.=0A=0A=0A    _______________________________________________________=
=0A=0AReply to this item at:=0A=0A  <https://savannah.gnu.org/support/?1113=
80>=0A=0A_______________________________________________=0AMessage sent via=
 Savannah=0Ahttps://savannah.gnu.org/=0A
--8323329-424238335-1770219544=:1412222
Content-Type: APPLICATION/pgp-signature; name=signature.asc

-----BEGIN PGP SIGNATURE-----

iHUEABYIAB0WIQQk97aszIMMAvLLwm6qLAuaBUf3TgUCaYNoGAAKCRCqLAuaBUf3
TljqAQCKrzAJsLpc6vKd85CSeIzfmzfC0CGreKye0mlszEctVwD/QGLe9rnuIXVP
HZEVIeRFc0VHo+lVhW7GjEKt4N/PWQQ=
=gkkU
-----END PGP SIGNATURE-----

--8323329-424238335-1770219544=:1412222--