https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119010
--- Comment #6 from Richard Biener <rguenth at gcc dot gnu.org> --- (In reply to Jan Hubicka from comment #5) > I had patch to reduce max issue back to 4 (exactly because of the compile > time slowdown and because the way we model decoder we own't issue more than > 4 instructions). Seems I forgot to push it, will re-benchmark and fix that. Note sapphirerapids, also with 6 max issue, doesn't exhibit the compile time slowdown, so I think it's because of the missing insn reservations for some of the moves. > I will check why movdf gets nothing, that is odd indeed. Thanks.
