https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61225
--- Comment #21 from Jeffrey A. Law <law at redhat dot com> --- >From my experimentations, I don't see a reasonable way to fix this in stage4 except via some horrid define_peep2. We don't have the LOG_LINKS we need to do a good job on this during the combiner. I'll certainly support xfailing the test and probably a downgrade to at least a P2, possibly a P4.
