[AArch64] Implement __builtin_thread_pointer

[zatrazz]

Summary

This path add the aarch64 lowering of __builtin_thread_pointer. It uses
the already implemented AArch64ISD::THREAD_POINTER used in TLS generation.

[rengolin]

This looks surprisingly simple. LGTM. Thanks!

test/CodeGen/AArch64/arm64-builtins-linux.ll

1

better to use "aarch64", since "arm64" is an alias.

[zatrazz]

test/CodeGen/AArch64/arm64-builtins-linux.ll

1

right, I will change that.

include/llvm/IR/IntrinsicsAArch64.td

 //===- IntrinsicsAARCH64.td - Defines AARCH64 intrinsics ---*- tablegen -*-===//
 //
 //                     The LLVM Compiler Infrastructure
 //
 // This file is distributed under the University of Illinois Open Source
 // License. See LICENSE.TXT for details.
 //
 //===----------------------------------------------------------------------===//
 //
 // This file defines all of the AARCH64-specific intrinsics.
 //
 //===----------------------------------------------------------------------===//
 
 let TargetPrefix = "aarch64" in {
 
+def int_aarch64_thread_pointer : GCCBuiltin<"__builtin_thread_pointer">,
+            Intrinsic<[llvm_ptr_ty], [], [IntrNoMem]>;
+
 def int_aarch64_ldxr : Intrinsic<[llvm_i64_ty], [llvm_anyptr_ty]>;
 def int_aarch64_ldaxr : Intrinsic<[llvm_i64_ty], [llvm_anyptr_ty]>;
 def int_aarch64_stxr : Intrinsic<[llvm_i32_ty], [llvm_i64_ty, llvm_anyptr_ty]>;
 def int_aarch64_stlxr : Intrinsic<[llvm_i32_ty], [llvm_i64_ty, llvm_anyptr_ty]>;
 
 def int_aarch64_ldxp : Intrinsic<[llvm_i64_ty, llvm_i64_ty], [llvm_ptr_ty]>;
 def int_aarch64_ldaxp : Intrinsic<[llvm_i64_ty, llvm_i64_ty], [llvm_ptr_ty]>;
 def int_aarch64_stxp : Intrinsic<[llvm_i32_ty],

lib/Target/AArch64/AArch64ISelLowering.h

SDValue LowerCall(CallLoweringInfo & /*CLI*/,
                     SmallVectorImpl<SDValue> &InVals) const override;
 
   SDValue LowerCallResult(SDValue Chain, SDValue InFlag,
                           CallingConv::ID CallConv, bool isVarArg,
                           const SmallVectorImpl<ISD::InputArg> &Ins, SDLoc DL,
                           SelectionDAG &DAG, SmallVectorImpl<SDValue> &InVals,
                           bool isThisReturn, SDValue ThisVal) const;
 
+  SDValue LowerINTRINSIC_WO_CHAIN(SDValue Op, SelectionDAG &DAG) const;
+
   bool isEligibleForTailCallOptimization(
       SDValue Callee, CallingConv::ID CalleeCC, bool isVarArg,
       bool isCalleeStructRet, bool isCallerStructRet,
       const SmallVectorImpl<ISD::OutputArg> &Outs,
       const SmallVectorImpl<SDValue> &OutVals,
       const SmallVectorImpl<ISD::InputArg> &Ins, SelectionDAG &DAG) const;
 
   /// Finds the incoming stack arguments which overlap the given fixed stack

lib/Target/AArch64/AArch64ISelLowering.cpp

AArch64TargetLowering::AArch64TargetLowering(const TargetMachine &TM,
   EnableExtLdPromotion = true;
 
   setMinFunctionAlignment(2);
 
   RequireStrictAlign = (Align == StrictAlign);
 
   setHasExtractBitsInsn(true);
 
+  setOperationAction(ISD::INTRINSIC_WO_CHAIN, MVT::Other, Custom);
+
   if (Subtarget->hasNEON()) {
     // FIXME: v1f64 shouldn't be legal if we can avoid it, because it leads to
     // silliness like this:
     setOperationAction(ISD::FABS, MVT::v1f64, Expand);
     setOperationAction(ISD::FADD, MVT::v1f64, Expand);
     setOperationAction(ISD::FCEIL, MVT::v1f64, Expand);
     setOperationAction(ISD::FCOPYSIGN, MVT::v1f64, Expand);
     setOperationAction(ISD::FCOS, MVT::v1f64, Expand);
static SDValue LowerMUL(SDValue Op, SelectionDAG &DAG) {
   EVT Op1VT = Op1.getValueType();
   return DAG.getNode(N0->getOpcode(), DL, VT,
                      DAG.getNode(NewOpc, DL, VT,
                                DAG.getNode(ISD::BITCAST, DL, Op1VT, N00), Op1),
                      DAG.getNode(NewOpc, DL, VT,
                                DAG.getNode(ISD::BITCAST, DL, Op1VT, N01), Op1));
 }
 
+SDValue AArch64TargetLowering::LowerINTRINSIC_WO_CHAIN(SDValue Op,
+                                                     SelectionDAG &DAG) const {
+  unsigned IntNo = cast<ConstantSDNode>(Op.getOperand(0))->getZExtValue();
+  SDLoc dl(Op);
+  switch (IntNo) {
+  default: return SDValue();    // Don't custom lower most intrinsics.
+  case Intrinsic::aarch64_thread_pointer: {
+    EVT PtrVT = getPointerTy(DAG.getDataLayout());
+    return DAG.getNode(AArch64ISD::THREAD_POINTER, dl, PtrVT);
+  }
+  }
+}
+
 SDValue AArch64TargetLowering::LowerOperation(SDValue Op,
                                               SelectionDAG &DAG) const {
   switch (Op.getOpcode()) {
   default:
     llvm_unreachable("unimplemented operand");
     return SDValue();
   case ISD::BITCAST:
     return LowerBITCAST(Op, DAG);
case ISD::UINT_TO_FP:
     return LowerINT_TO_FP(Op, DAG);
   case ISD::FP_TO_SINT:
   case ISD::FP_TO_UINT:
     return LowerFP_TO_INT(Op, DAG);
   case ISD::FSINCOS:
     return LowerFSINCOS(Op, DAG);
   case ISD::MUL:
     return LowerMUL(Op, DAG);
+  case ISD::INTRINSIC_WO_CHAIN:
+    return LowerINTRINSIC_WO_CHAIN(Op, DAG);
   }
 }
 
 /// getFunctionAlignment - Return the Log2 alignment of this function.
 unsigned AArch64TargetLowering::getFunctionAlignment(const Function *F) const {
   return 2;
 }
 

test/CodeGen/AArch64/arm64-builtins-linux.ll

+; RUN: llc < %s -march=arm64 -mtriple=arm64-linux-gnu | FileCheck %s
+
+; Function Attrs: nounwind readnone
+declare i8* @llvm.aarch64.thread.pointer() #1
+
+define i8* @thread_pointer() {
+; CHECK: thread_pointer:
+; CHECK: mrs {{x[0-9]+}}, TPIDR_EL0
+  %1 = tail call i8* @llvm.aarch64.thread.pointer()
+  ret i8* %1
+}