https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85999
Bug ID: 85999 Summary: 416.gamess slowed down by BB vectorization Product: gcc Version: 8.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: tree-optimization Assignee: unassigned at gcc dot gnu.org Reporter: rguenth at gcc dot gnu.org Target Milestone: --- Adding -ftree-slp-vectorize to -O2 -march=haswell [-mprefer-avx128] on 416.gamess slows down the benchmark significantly. Adding -fno-tree-slp-vectorize to -Ofast -march=haswell speeds it up. Thus somehow BB vectorization slows down 416.gamess.