[sr #111243] stdbool conftest program (and possibly others) is broken with GCC 15

Sergio Durigan Junior <[email protected]> Tue, 13 May 2025 22:11:41 -0400 (EDT)

Newsgroups	gmane.comp.sysutils.autoconf.bugs
Message-ID	<[email protected]>

Follow-up Comment #4, sr #111243 (group autoconf):

[comment #3 comment #3:]
> The "long long reliablity" test is not part of autoconf, so any bug reports
> need to be sent to whoever maintains that test.

Apologies for the noise; my brief inspection of the code led me to believe
that it was part of autoconf.

> I cannot reproduce the problem with stdbool.h using Autoconf 2.72 (as shipped
> with Fedora 42), when building cifs-tools commit
> 5ba5b16a73e70b586355b591f9feefbe3c493bc2 (aka version 7.3).

Hm, that's interesting.  And I'm assuming that you're using GCC 15 as well?

> The message "checking for stdbool.h that conforms to C99" listed in the
> original bug report is not generated by Autoconf 2.72; it is generated by
> older versions of Autoconf. So it appears the original bug report is
> mistaken, in that the bug is about older Autoconf versions (which are indeed
> buggy in this area), not about Autoconf 2.72 (where the bugs are fixed).

I just checked and the build is indeed using Autoconf 2.72, so it wasn't a
mistake on my end.

I will perform a few more tests here and report back if I find something.


    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/support/?111243>

_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/

signature.asc (application/pgp-signature, 228 B)