http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54146
--- Comment #47 from Steven Bosscher <steven at gcc dot gnu.org> 2012-08-15 15:07:05 UTC --- (In reply to comment #46) > Created attachment 28020 [details] > Faster rewrite_into_loop_closed_ssa After this patch, IRA is the only major bottle-neck left, although there are still a few passes that take more time than I think is reasonable (out-of-SSA in particular). New top 10 spenders: integrated RA : 227.36 (30%) usr 678602 kB (19%) ggc tree SSA incremental : 66.17 ( 9%) usr 151828 kB ( 4%) ggc out of ssa : 43.76 ( 6%) usr 608 kB ( 0%) ggc tree PTA : 43.06 ( 6%) usr 27466 kB ( 1%) ggc df live regs : 35.03 ( 5%) usr 0 kB ( 0%) ggc df live&initialized regs: 34.25 ( 5%) usr 0 kB ( 0%) ggc tree SSA rewrite : 18.37 ( 2%) usr 24888 kB ( 1%) ggc remove unused locals : 17.97 ( 2%) usr 0 kB ( 0%) ggc dominance computation : 15.06 ( 2%) usr 0 kB ( 0%) ggc tree CFG cleanup : 11.79 ( 2%) usr 1123 kB ( 0%) ggc