On Wed, Aug 28, 2019 at 3:54 PM Richard Biener <rguent...@suse.de> wrote: > > >>> 2019-08-23 Richard Biener <rguent...@suse.de> > > >>> > > >>> PR target/91522 > > >>> PR target/91527 > > >>> * config/i386/i386-features.h (general_scalar_chain::defs_map): > > >>> New member. > > >>> (general_scalar_chain::replace_with_subreg): Remove. > > >>> (general_scalar_chain::replace_with_subreg_in_insn): Likewise. > > >>> (general_scalar_chain::convert_reg): Adjust signature. > > >>> * config/i386/i386-features.c (scalar_chain::add_insn): Do not > > >>> iterate over all defs of a reg. > > >>> (general_scalar_chain::replace_with_subreg): Remove. > > >>> (general_scalar_chain::replace_with_subreg_in_insn): Likewise. > > >>> (general_scalar_chain::make_vector_copies): Populate defs_map, > > >>> place copy only after defs that are used as vectors in the chain. > > >>> (general_scalar_chain::convert_reg): Emit a copy for a specific > > >>> def in a specific instruction. > > >>> (general_scalar_chain::convert_op): All reg uses are converted > > >>> here. > > >>> (general_scalar_chain::convert_insn): Emit copies for scalar > > >>> uses of defs here. Replace uses with the copies we created. > > >>> Replace and convert the def. Adjust REG_DEAD notes, remove > > >>> REG_EQUIV/EQUAL notes. > > >>> (general_scalar_chain::convert_registers): Only handle copies > > >>> into the chain here. > > > > > > Rubberstamped with LGTM. It looks you are the master of this domain now ;) > > > > This breaks bootstrap for i686-darwin (and most likely is the cause of > > bootstrap fail on > > i686-linux i686-linux-gnu at > > https://gcc.gnu.org/ml/gcc-regression/2019-08/msg00398.html > > et al) > > It gives a bunch of compare errors spread around the tree - so no specific > > pointer there. > > > > There’s a secondary fail overlaying it between 274933-or so and 274980 > > which confused > > my initial search. > > Please file a bugreport. Don't have time today to dig into but will do > tomorrow (but now at least try to reproduce).
Looking at HJ's testreports, it looks configuring for i686-linux --with-fpmath=sse should be enough to trigger comparison failure. Uros.