d7c702b451
In the motivating case from PR35681 and represented by the macro-fuse-cmp test: https://bugs.llvm.org/show_bug.cgi?id=35681 ...there's a 37 -> 31 byte size win for the loop because we eliminate the big base address offsets. SPEC2017 on Ryzen shows no significant perf difference. Differential Revision: https://reviews.llvm.org/D42607 llvm-svn: 324289 |
||
|---|---|---|
| .. | ||
| 2008-08-14-ShadowIV.ll | ||
| 2009-11-10-LSRCrash.ll | ||
| 2011-07-20-DoubleIV.ll | ||
| 2011-11-29-postincphi.ll | ||
| 2011-12-04-loserreg.ll | ||
| 2012-01-13-phielim.ll | ||
| bin_power.ll | ||
| canonical-2.ll | ||
| canonical.ll | ||
| incorrect-offset-scaling.ll | ||
| ivchain-X86.ll | ||
| ivchain-stress-X86.ll | ||
| lit.local.cfg | ||
| lsr-expand-quadratic.ll | ||
| lsr-filtering-scaledreg.ll | ||
| lsr-insns-1.ll | ||
| lsr-insns-2.ll | ||
| macro-fuse-cmp.ll | ||
| nested-loop.ll | ||
| no_superflous_induction_vars.ll | ||
| pr17473.ll | ||
| pr28719.ll | ||
| sibling-loops.ll | ||