> On 12 Aug 2025, at 22:13, Joseph Myers via Gcc <gcc@gcc.gnu.org> wrote:
>
> On Tue, 12 Aug 2025, Andrew Pinski via Gcc wrote:
>
>> I think we should have a requirement of the bare minimum for a port is a
>> maintainer.
>> I also vote to have a testresults for the target at least once a year.
>
> Maintainers should also show some sign of action on things requiring
> changes across all ports.
>
> One GCC-internal example would be the move to LRA.
>
> One user-visible example would be defining an ABI for _BitInt and enabling
> _BitInt support for the port, it's a required C23 feature. Here many of
> the less-active ports actually have it *easier* than the more-active
> ports, because they probably don't have any maintained ABI document or
> expectation of being ABI-compatible with other implementations for such
> new features, so just need to write down what the ABI is that GCC uses in
> the commit message when enabling the feature - whereas for more-active
> ports, there might be a maintained ABI document to update, and other
> active implementations to maintain ABI compatibility with, both of which
> are liable to slow down defining the ABI and adding the feature.
and there are some ports (like Darwin) where the GCC maintainers are not
the ABI owners and might have limited ability to influence priority in adopting
changes.
Iain
