[llvm] [RISCV][CodeGen][TTI]Use processShuffleMasks for cost estimations/actual per-register shuffles (PR #118103)

[Luke Lau via llvm-commits]

================
@@ -5098,58 +5097,70 @@ static SDValue lowerShuffleViaVRegSplitting(ShuffleVectorSDNode *SVN,
 
   MVT ElemVT = VT.getVectorElementType();
   unsigned ElemsPerVReg = *VLen / ElemVT.getFixedSizeInBits();
-  unsigned VRegsPerSrc = NumElts / ElemsPerVReg;
-
-  SmallVector<std::pair<int, SmallVector<int>>>
-    OutMasks(VRegsPerSrc, {-1, {}});
-
-  // Check if our mask can be done as a 1-to-1 mapping from source
-  // to destination registers in the group without needing to
-  // write each destination more than once.
-  for (unsigned DstIdx = 0; DstIdx < Mask.size(); DstIdx++) {
-    int DstVecIdx = DstIdx / ElemsPerVReg;
-    int DstSubIdx = DstIdx % ElemsPerVReg;
-    int SrcIdx = Mask[DstIdx];
-    if (SrcIdx < 0 || (unsigned)SrcIdx >= 2 * NumElts)
-      continue;
-    int SrcVecIdx = SrcIdx / ElemsPerVReg;
-    int SrcSubIdx = SrcIdx % ElemsPerVReg;
-    if (OutMasks[DstVecIdx].first == -1)
-      OutMasks[DstVecIdx].first = SrcVecIdx;
-    if (OutMasks[DstVecIdx].first != SrcVecIdx)
-      // Note: This case could easily be handled by keeping track of a chain
-      // of source values and generating two element shuffles below.  This is
-      // less an implementation question, and more a profitability one.
-      return SDValue();
-
-    OutMasks[DstVecIdx].second.resize(ElemsPerVReg, -1);
-    OutMasks[DstVecIdx].second[DstSubIdx] = SrcSubIdx;
-  }
 
   EVT ContainerVT = getContainerForFixedLengthVector(DAG, VT, Subtarget);
   MVT OneRegVT = MVT::getVectorVT(ElemVT, ElemsPerVReg);
   MVT M1VT = getContainerForFixedLengthVector(DAG, OneRegVT, Subtarget);
   assert(M1VT == getLMUL1VT(M1VT));
   unsigned NumOpElts = M1VT.getVectorMinNumElements();
-  SDValue Vec = DAG.getUNDEF(ContainerVT);
+  unsigned NormalizedVF = ContainerVT.getVectorMinNumElements();
----------------
lukel97 wrote:

I don't think VF is a term used outside of the vectorizers much. Maybe use something like ContainerNumOpElts?

https://github.com/llvm/llvm-project/pull/118103