https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119925
--- Comment #3 from Filip Kastl <pheeck at gcc dot gnu.org> --- (In reply to Filip Kastl from comment #1) > I've just noticed that there is also a 10% slowdown of 433.milc SPEC 2006 > benchmark also with -Ofast -march=native PGO. Looking at the graphs, I'd > say these two slowdowns probably have the same cause. > > https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=583.507.0 Bad graph, sorry. Here is the one I wanted to send: https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.9=578.70.0 The comparison graph is ok.
