[AMDGPU] Minor change to d16 buffer load implementation

[tpr]

Summary

By not reconstructing the operand list of the SDNode, this change makes
it easier to add the forthcoming new tbuffer and buffer intrinsics.

Change-Id: I0cb79ef0801532645d7dd954a6d7355139db7b38

[nhaehnle]

LGTM

llvm/trunk/lib/Target/AMDGPU/SIISelLowering.h

private:
   SDValue LowerFDIV(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerINT_TO_FP(SDValue Op, SelectionDAG &DAG, bool Signed) const;
   SDValue LowerSTORE(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerTrig(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerATOMIC_CMP_SWAP(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerBRCOND(SDValue Op, SelectionDAG &DAG) const;
 
   SDValue adjustLoadValueType(unsigned Opcode, MemSDNode *M,
-                              SelectionDAG &DAG,
+                              SelectionDAG &DAG, ArrayRef<SDValue> Ops,
                               bool IsIntrinsic = false) const;
 
   SDValue handleD16VData(SDValue VData, SelectionDAG &DAG) const;
 
   /// Converts \p Op, which must be of floating point type, to the
   /// floating point type \p VT, by either extending or truncating it.
   SDValue getFPExtOrFPTrunc(SelectionDAG &DAG,
                             SDValue Op,

llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp

static SDValue adjustLoadValueTypeImpl(SDValue Result, EVT LoadVT,
 
   // Cast back to the original packed type.
   return DAG.getNode(ISD::BITCAST, DL, LoadVT, Result);
 }
 
 SDValue SITargetLowering::adjustLoadValueType(unsigned Opcode,
                                               MemSDNode *M,
                                               SelectionDAG &DAG,
+                                              ArrayRef<SDValue> Ops,
                                               bool IsIntrinsic) const {
   SDLoc DL(M);
-  SmallVector<SDValue, 10> Ops;
-  Ops.reserve(M->getNumOperands());
-
-  Ops.push_back(M->getOperand(0));
-  if (IsIntrinsic)
-    Ops.push_back(DAG.getConstant(Opcode, DL, MVT::i32));
-
-  // Skip 1, as it is the intrinsic ID.
-  for (unsigned I = 2, E = M->getNumOperands(); I != E; ++I)
-    Ops.push_back(M->getOperand(I));
 
   bool Unpacked = Subtarget->hasUnpackedD16VMem();
   EVT LoadVT = M->getValueType(0);
 
   EVT EquivLoadVT = LoadVT;
   if (Unpacked && LoadVT.isVector()) {
     EquivLoadVT = LoadVT.isVector() ?
       EVT::getVectorVT(*DAG.getContext(), MVT::i32,
case Intrinsic::amdgcn_buffer_load_format: {
     };
 
     unsigned Opc = (IntrID == Intrinsic::amdgcn_buffer_load) ?
         AMDGPUISD::BUFFER_LOAD : AMDGPUISD::BUFFER_LOAD_FORMAT;
     EVT VT = Op.getValueType();
     EVT IntVT = VT.changeTypeToInteger();
     auto *M = cast<MemSDNode>(Op);
     EVT LoadVT = Op.getValueType();
-    bool IsD16 = LoadVT.getScalarType() == MVT::f16;
-    if (IsD16)
-      return adjustLoadValueType(AMDGPUISD::BUFFER_LOAD_FORMAT_D16, M, DAG);
 
+    if (LoadVT.getScalarType() == MVT::f16)
+      return adjustLoadValueType(AMDGPUISD::BUFFER_LOAD_FORMAT_D16,
+                                 M, DAG, Ops);
     return DAG.getMemIntrinsicNode(Opc, DL, Op->getVTList(), Ops, IntVT,
                                    M->getMemOperand());
   }
   case Intrinsic::amdgcn_tbuffer_load: {
     MemSDNode *M = cast<MemSDNode>(Op);
     EVT LoadVT = Op.getValueType();
-    bool IsD16 = LoadVT.getScalarType() == MVT::f16;
-    if (IsD16) {
-      return adjustLoadValueType(AMDGPUISD::TBUFFER_LOAD_FORMAT_D16, M, DAG);
-    }
 
     SDValue Ops[] = {
       Op.getOperand(0),  // Chain
       Op.getOperand(2),  // rsrc
       Op.getOperand(3),  // vindex
       Op.getOperand(4),  // voffset
       Op.getOperand(5),  // soffset
       Op.getOperand(6),  // offset
       Op.getOperand(7),  // dfmt
       Op.getOperand(8),  // nfmt
       Op.getOperand(9),  // glc
       Op.getOperand(10)   // slc
     };
 
+    if (LoadVT.getScalarType() == MVT::f16)
+      return adjustLoadValueType(AMDGPUISD::TBUFFER_LOAD_FORMAT_D16,
+                                 M, DAG, Ops);
     return DAG.getMemIntrinsicNode(AMDGPUISD::TBUFFER_LOAD_FORMAT, DL,
                                    Op->getVTList(), Ops, LoadVT,
                                    M->getMemOperand());
   }
   case Intrinsic::amdgcn_buffer_atomic_swap:
   case Intrinsic::amdgcn_buffer_atomic_add:
   case Intrinsic::amdgcn_buffer_atomic_sub:
   case Intrinsic::amdgcn_buffer_atomic_smin: