[PATCH] D64760: [x86] use more phadd for reductions

[Simon Pilgrim via Phabricator via llvm-commits]

RKSimon added inline comments.


================
Comment at: llvm/test/CodeGen/X86/vector-reduce-add.ll:680
 ; AVX1-FAST-NEXT:    vphaddw %xmm0, %xmm0, %xmm0
 ; AVX1-FAST-NEXT:    vmovd %xmm0, %eax
 ; AVX1-FAST-NEXT:    # kill: def $ax killed $ax killed $eax
----------------
spatel wrote:
> RKSimon wrote:
> > ideally we'd have these cases combine as ll
> 'll' - do you mean alter/override the ExpandReductions pass in IR?
sorry - "ideally we'd have these cases combine as well"


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D64760/new/

https://reviews.llvm.org/D64760