http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57309
--- Comment #2 from Bill Schmidt <wschmidt at gcc dot gnu.org> --- (In reply to Richard Biener from comment #1) > Can you isolate a testcase for the worst loop? Not yet. It's one of these horrible gargantuan functions (leslie3d is one big file and fluxi, fluxj, fluxk are all quite large functions) and the issue appears to involve global register allocation, so it will be sensitive to control flow changes. Not an easy thing to attack with delta, either. Peter is planning to look at this at some point, and I have a bunch of dump files saved away to help with the analysis. Not sure what else to do for now.
