[Bug lto/83201] [7/8 Regression] SPEC CPU2017 505.mcf_r produces incorrect output when built with -flto and FDO

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83201

--- Comment #9 from Martin Liška <marxin at gcc dot gnu.org> ---
Using just a single ltrans, I see first divergence in mcf_r.ltrans0.088t.dom1.
Richi, how possible is the revision real culprit?