https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64460
--- Comment #5 from ktkachov at gcc dot gnu.org --- (In reply to Joel Sherrill from comment #4) > (In reply to ktkachov from comment #3) > > I have a fix in the works. The bug is down to the "shift" attribute value in > > one of the patterns, although only xscale_sched_adjust_cost uses that > > attribute. > > Awesome! I was about to start another attempt at a git bisect but it sounds > like you don't need it. Let me know if it is worth running. Thanks, but no need. I think I've got the root cause and I got a fix that fixes the testcase, I just need to put it through testing. A testsuite-appropriate reduced testcase would be nice I guess, although I'll kick off a reducer soon.
