Possible suboptimal code generated in 32-bit ABI mode

[Richard Bradfield]

On Sun, Nov 19, 2017 at 04:19:45PM +0000, bradfirj@fstab.me wrote:
>For reference, I am compiling everything using gcc trunk, at commit
>254929 from Sun Nov 19, and I am benchmarking on a Skylake i7-6700K
>at 4.0GHz.

And something else I should definitely have included in the first place,
I cannot reproduce this on the GCC 7.2 shipped with my distribution, so
is it possible this is a regression?

Richard