https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113939
--- Comment #5 from Jeffrey A. Law <law at gcc dot gnu.org> --- So the m68k bootstrap with LRA enabled blows up. It looks like the stage1 compiler is mis-compiling the stage2 compiler. The result is a metric ton of errors building the stage2 libgcc. A .o bisection landed on fold-const.o (ugh). I'll try to narrow down further at some point, but it's way down on my priority list.
