[Bug middle-end/118097] [15 regression] recent bug with -O2, but not -O1

[sjames at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118097

--- Comment #8 from Sam James <sjames at gcc dot gnu.org> ---
(In reply to Sam James from comment #7)
> Sure, but I can't bisect it if I can't reproduce it...
> 
> On godbolt, even, I see the same results for gcc 14.1 and trunk:
> https://godbolt.org/z/o8orfaq6x.

And for your original: https://godbolt.org/z/hM5Th9KdK. What am I missing?