[PATCH] D15559: [LoopUnrollRuntime] Do unroll when the unroll factor is 2

Charlie Turner via llvm-commits llvm-commits at lists.llvm.org
Fri Dec 18 01:43:13 PST 2015


chatur01 added a comment.

I've just seen the following regressions for spec2006 on the A57 in T32 mode,

  spec.cpu2006.ref.456_hmmer 	13.34%
  spec.cpu2006.ref.471_omnetpp 	7.87%
  spec.cpu2006.ref.458_sjeng 	3.72%
  spec.cpu2006.ref.429_mcf 	3.48%
  spec.cpu2006.ref.462_libquantum 	3.47%
  spec.cpu2006.ref.482_sphinx3 	2.27%
  spec.cpu2006.ref.445_gobmk 	1.90%
  spec.cpu2006.ref.453_povray 	1.61%
  spec.cpu2006.ref.447_dealII 	1.10%

hmmer is clearly a good test case here, must be some register pressure issues.


http://reviews.llvm.org/D15559





More information about the llvm-commits mailing list