[BPF] expand cttz, ctlz for i32, i64

llvm/lib/Target/BPF/BPFISelLowering.cpp

@@ -69,7 +69,7 @@ BPFTargetLowering::BPFTargetLowering(const TargetMachine &TM,
   setOperationAction(ISD::BRIND, MVT::Other, Expand);
   setOperationAction(ISD::BRCOND, MVT::Other, Expand);
 
-  setOperationAction(ISD::GlobalAddress, MVT::i64, Custom);
+  setOperationAction({ISD::GlobalAddress, ISD::ConstantPool}, MVT::i64, Custom);
 
   setOperationAction(ISD::DYNAMIC_STACKALLOC, MVT::i64, Custom);
   setOperationAction(ISD::STACKSAVE, MVT::Other, Expand);
@@ -111,6 +111,10 @@ BPFTargetLowering::BPFTargetLowering(const TargetMachine &TM,
     setOperationAction(ISD::SRL_PARTS, VT, Expand);
     setOperationAction(ISD::SRA_PARTS, VT, Expand);
     setOperationAction(ISD::CTPOP, VT, Expand);
+    setOperationAction(ISD::CTTZ, VT, Expand);
+    setOperationAction(ISD::CTLZ, VT, Expand);
+    setOperationAction(ISD::CTTZ_ZERO_UNDEF, VT, Expand);
+    setOperationAction(ISD::CTLZ_ZERO_UNDEF, VT, Expand);
 
     setOperationAction(ISD::SETCC, VT, Expand);
     setOperationAction(ISD::SELECT, VT, Expand);
@@ -123,11 +127,6 @@ BPFTargetLowering::BPFTargetLowering(const TargetMachine &TM,
                        STI.getHasJmp32() ? Custom : Promote);
   }
 
-  setOperationAction(ISD::CTTZ, MVT::i64, Custom);
-  setOperationAction(ISD::CTLZ, MVT::i64, Custom);
-  setOperationAction(ISD::CTTZ_ZERO_UNDEF, MVT::i64, Custom);
-  setOperationAction(ISD::CTLZ_ZERO_UNDEF, MVT::i64, Custom);
-
   setOperationAction(ISD::SIGN_EXTEND_INREG, MVT::i1, Expand);
   if (!STI.hasMovsx()) {
     setOperationAction(ISD::SIGN_EXTEND_INREG, MVT::i8, Expand);
@@ -305,6 +304,8 @@ SDValue BPFTargetLowering::LowerOperation(SDValue Op, SelectionDAG &DAG) const {
     return LowerBR_CC(Op, DAG);
   case ISD::GlobalAddress:
     return LowerGlobalAddress(Op, DAG);
+  case ISD::ConstantPool:
+    return LowerConstantPool(Op, DAG);
   case ISD::SELECT_CC:
     return LowerSELECT_CC(Op, DAG);
   case ISD::DYNAMIC_STACKALLOC:
@@ -670,18 +671,41 @@ const char *BPFTargetLowering::getTargetNodeName(unsigned Opcode) const {
   return nullptr;
 }
 
+static SDValue getTargetNode(GlobalAddressSDNode *N, const SDLoc &DL, EVT Ty,
+                             SelectionDAG &DAG, unsigned Flags) {
+  return DAG.getTargetGlobalAddress(N->getGlobal(), DL, Ty, 0, Flags);
+}
+
+static SDValue getTargetNode(ConstantPoolSDNode *N, const SDLoc &DL, EVT Ty,
+                             SelectionDAG &DAG, unsigned Flags) {
+  return DAG.getTargetConstantPool(N->getConstVal(), Ty, N->getAlign(),
+                                   N->getOffset(), Flags);
+}
+
+template <class NodeTy>
+SDValue BPFTargetLowering::getAddr(NodeTy *N, SelectionDAG &DAG,
+                                   unsigned Flags) const {
+  SDLoc DL(N);
+
+  SDValue GA = getTargetNode(N, DL, MVT::i64, DAG, Flags);
+
+  return DAG.getNode(BPFISD::Wrapper, DL, MVT::i64, GA);
+}
+
 SDValue BPFTargetLowering::LowerGlobalAddress(SDValue Op,
                                               SelectionDAG &DAG) const {
-  auto *N = cast<GlobalAddressSDNode>(Op);
+  GlobalAddressSDNode *N = cast<GlobalAddressSDNode>(Op);
   if (N->getOffset() != 0)
     report_fatal_error("invalid offset for global address: " +
                        Twine(N->getOffset()));
+  return getAddr(N, DAG);
+}
 
-  SDLoc DL(Op);
-  const GlobalValue *GV = N->getGlobal();
-  SDValue GA = DAG.getTargetGlobalAddress(GV, DL, MVT::i64);
+SDValue BPFTargetLowering::LowerConstantPool(SDValue Op,
+                                             SelectionDAG &DAG) const {
+  ConstantPoolSDNode *N = cast<ConstantPoolSDNode>(Op);
 
-  return DAG.getNode(BPFISD::Wrapper, DL, MVT::i64, GA);
+  return getAddr(N, DAG);
 }
 
 unsigned

llvm/lib/Target/BPF/BPFISelLowering.h

@@ -75,8 +75,13 @@ class BPFTargetLowering : public TargetLowering {
 
   SDValue LowerBR_CC(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerSELECT_CC(SDValue Op, SelectionDAG &DAG) const;
+
+  SDValue LowerConstantPool(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerGlobalAddress(SDValue Op, SelectionDAG &DAG) const;
 
+  template <class NodeTy>
+  SDValue getAddr(NodeTy *N, SelectionDAG &DAG, unsigned Flags = 0) const;
+
   // Lower the result values of a call, copying them out of physregs into vregs
   SDValue LowerCallResult(SDValue Chain, SDValue InGlue,
                           CallingConv::ID CallConv, bool IsVarArg,

llvm/lib/Target/BPF/BPFInstrInfo.td

@@ -727,6 +727,7 @@ let usesCustomInserter = 1, isCodeGenOnly = 1 in {
 
 // load 64-bit global addr into register
 def : Pat<(BPFWrapper tglobaladdr:$in), (LD_imm64 tglobaladdr:$in)>;
+def : Pat<(BPFWrapper tconstpool:$in), (LD_imm64 tconstpool:$in)>;
 
 // 0xffffFFFF doesn't fit into simm32, optimize common case
 def : Pat<(i64 (and (i64 GPR:$src), 0xffffFFFF)),

llvm/lib/Target/BPF/BPFMCInstLower.cpp

@@ -74,6 +74,9 @@ void BPFMCInstLower::Lower(const MachineInstr *MI, MCInst &OutMI) const {
     case MachineOperand::MO_GlobalAddress:
       MCOp = LowerSymbolOperand(MO, GetGlobalAddressSymbol(MO));
       break;
+    case MachineOperand::MO_ConstantPoolIndex:
+      MCOp = LowerSymbolOperand(MO, Printer.GetCPISymbol(MO.getIndex()));
+      break;
     }
 
     OutMI.addOperand(MCOp);