[llvm-commits] [llvm] r126424 - in /llvm/trunk: docs/LangRef.html lib/CodeGen/SelectionDAG/DAGCombiner.cpp test/CodeGen/X86/vec_anyext.ll test/CodeGen/X86/vec_sext.ll

[Duncan Sands]

Hi Nadav,

> -  if (ISD::isNON_EXTLoad(N0.getNode())&&
> +  // None of the supported targets knows how to perform load and sign extend
> +  // in one instruction.  We only perform this transformation on scalars.

I think in your comment (and the other similar ones) you should make clear that
you mean that no supported targets know how to do this for vectors.  Right now
it sounds like no target can do it for scalars either.

> +  // Currenly we only perform this optimization on scalars because vectors

Currenly -> Currently

Ciao, Duncan.