[PATCH] D55182: InstCombine: Scalarize single use icmp/fcmp

[Sanjay Patel via Phabricator via llvm-commits]

spatel added a comment.

In D55182#1316746 <https://reviews.llvm.org/D55182#1316746>, @spatel wrote:

> 2. There should be a test that shows the benefit of using 'isCheapToScalarize()'. IIUC, that would involve a sequence with multiple binops and/or compares between an insertelement and the trailing extract.


I added a binop test like this here:
rL348417 <https://reviews.llvm.org/rL348417>
...and it shows a missed opportunity. AFAICT, there was no test coverage for anything beyond the simplest cases for isCheapToScalarize(), so please do add something like that for cmp.

Also, I used 'match' to reduce the binop code:
rL348418 <https://reviews.llvm.org/rL348418>

cmp should be similar:

  Value *X, *Y;
  CmpInst::Predicate Pred;
  if (match(SrcVec, m_Cmp(Pred, m_Value(X), m_Value(Y))) &&
      cheapToScalarize(SrcVec, IndexC)) {
    // extelt (cmp X, Y), Index --> cmp (extelt X, Index), (extelt Y, Index)
    Value *E0 = Builder.CreateExtractElement(X, Index);
    Value *E1 = Builder.CreateExtractElement(Y, Index);
    return CmpInst::Create(cast<CmpInst>(SrcVec)->getOpcode(), Pred, E0, E1);
  }


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D55182/new/

https://reviews.llvm.org/D55182