[IRBuilder] Add new overload for CreateIntrinsic

clang/lib/CodeGen/CGHLSLRuntime.cpp

@@ -385,8 +385,8 @@ void CGHLSLRuntime::emitEntryFunction(const FunctionDecl *FD,
   SmallVector<OperandBundleDef, 1> OB;
   if (CGM.shouldEmitConvergenceTokens()) {
     assert(EntryFn->isConvergent());
-    llvm::Value *I = B.CreateIntrinsic(
-        llvm::Intrinsic::experimental_convergence_entry, {}, {});
+    llvm::Value *I =
+        B.CreateIntrinsic(llvm::Intrinsic::experimental_convergence_entry, {});
     llvm::Value *bundleArgs[] = {I};
     OB.emplace_back("convergencectrl", bundleArgs);
   }

clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp

@@ -241,21 +241,22 @@ static Value *handleHlslClip(const CallExpr *E, CodeGenFunction *CGF) {
     CMP = CGF->Builder.CreateIntrinsic(
         CGF->Builder.getInt1Ty(), CGF->CGM.getHLSLRuntime().getAnyIntrinsic(),
         {FCompInst});
-  } else
+  } else {
     CMP = CGF->Builder.CreateFCmpOLT(Op0, FZeroConst);
+  }
 
-  if (CGF->CGM.getTarget().getTriple().isDXIL())
+  if (CGF->CGM.getTarget().getTriple().isDXIL()) {
     LastInstr = CGF->Builder.CreateIntrinsic(
         CGF->VoidTy, Intrinsic::dx_discard, {CMP});
-  else if (CGF->CGM.getTarget().getTriple().isSPIRV()) {
+  } else if (CGF->CGM.getTarget().getTriple().isSPIRV()) {
     BasicBlock *LT0 = CGF->createBasicBlock("lt0", CGF->CurFn);
     BasicBlock *End = CGF->createBasicBlock("end", CGF->CurFn);
 
     CGF->Builder.CreateCondBr(CMP, LT0, End);
 
     CGF->Builder.SetInsertPoint(LT0);
 
-    CGF->Builder.CreateIntrinsic(CGF->VoidTy, Intrinsic::spv_discard, {});
+    CGF->Builder.CreateIntrinsic(Intrinsic::spv_discard, {});
 
     LastInstr = CGF->Builder.CreateBr(End);
     CGF->Builder.SetInsertPoint(End);

llvm/include/llvm/IR/IRBuilder.h

@@ -1004,6 +1004,14 @@ class IRBuilderBase {
                             ArrayRef<Value *> Args, FMFSource FMFSource = {},
                             const Twine &Name = "");
 
+  /// Create a call to non-overloaded intrinsic \p ID with \p Args. If
+  /// \p FMFSource is provided, copy fast-math-flags from that instruction to
+  /// the intrinsic.
+  CallInst *CreateIntrinsic(Intrinsic::ID ID, ArrayRef<Value *> Args,
+                            FMFSource FMFSource = {}, const Twine &Name = "") {
+    return CreateIntrinsic(ID, /*Types=*/{}, Args, FMFSource, Name);
+  }
+
   /// Create call to the minnum intrinsic.
   Value *CreateMinNum(Value *LHS, Value *RHS, FMFSource FMFSource = {},
                       const Twine &Name = "") {

llvm/lib/CodeGen/SafeStack.cpp

@@ -367,7 +367,7 @@ Value *SafeStack::getStackGuard(IRBuilder<> &IRB, Function &F) {
 
   if (!StackGuardVar) {
     TL.insertSSPDeclarations(*M);
-    return IRB.CreateIntrinsic(Intrinsic::stackguard, {}, {});
+    return IRB.CreateIntrinsic(Intrinsic::stackguard, {});
   }
 
   return IRB.CreateLoad(StackPtrTy, StackGuardVar, "StackGuard");

llvm/lib/CodeGen/StackProtector.cpp

@@ -542,7 +542,7 @@ static Value *getStackGuard(const TargetLoweringBase *TLI, Module *M,
   if (SupportsSelectionDAGSP)
     *SupportsSelectionDAGSP = true;
   TLI->insertSSPDeclarations(*M);
-  return B.CreateIntrinsic(Intrinsic::stackguard, {}, {});
+  return B.CreateIntrinsic(Intrinsic::stackguard, {});
 }
 
 /// Insert code into the entry block that stores the stack guard
@@ -563,7 +563,7 @@ static bool CreatePrologue(Function *F, Module *M, Instruction *CheckLoc,
   AI = B.CreateAlloca(PtrTy, nullptr, "StackGuardSlot");
 
   Value *GuardSlot = getStackGuard(TLI, M, B, &SupportsSelectionDAGSP);
-  B.CreateIntrinsic(Intrinsic::stackprotector, {}, {GuardSlot, AI});
+  B.CreateIntrinsic(Intrinsic::stackprotector, {GuardSlot, AI});
   return SupportsSelectionDAGSP;
 }
 

llvm/lib/IR/AutoUpgrade.cpp

@@ -1767,7 +1767,7 @@ static Value *upgradeX86VPERMT2Intrinsics(IRBuilder<> &Builder, CallBase &CI,
   if (!IndexForm)
     std::swap(Args[0], Args[1]);
 
-  Value *V = Builder.CreateIntrinsic(IID, {}, Args);
+  Value *V = Builder.CreateIntrinsic(IID, Args);
   Value *PassThru = ZeroMask ? ConstantAggregateZero::get(Ty)
                              : Builder.CreateBitCast(CI.getArgOperand(1),
                                                      Ty);
@@ -2022,8 +2022,8 @@ static Value *upgradeMaskedCompare(IRBuilder<> &Builder, CallBase &CI,
 // Replace a masked intrinsic with an older unmasked intrinsic.
 static Value *upgradeX86MaskedShift(IRBuilder<> &Builder, CallBase &CI,
                                     Intrinsic::ID IID) {
-  Value *Rep = Builder.CreateIntrinsic(
-      IID, {}, {CI.getArgOperand(0), CI.getArgOperand(1)});
+  Value *Rep =
+      Builder.CreateIntrinsic(IID, {CI.getArgOperand(0), CI.getArgOperand(1)});
   return emitX86Select(Builder, CI.getArgOperand(3), Rep, CI.getArgOperand(2));
 }
 
@@ -2280,7 +2280,7 @@ static bool upgradeAVX512MaskToSelect(StringRef Name, IRBuilder<> &Builder,
   SmallVector<Value *, 4> Args(CI.args());
   Args.pop_back();
   Args.pop_back();
-  Rep = Builder.CreateIntrinsic(IID, {}, Args);
+  Rep = Builder.CreateIntrinsic(IID, Args);
   unsigned NumArgs = CI.arg_size();
   Rep = emitX86Select(Builder, CI.getArgOperand(NumArgs - 1), Rep,
                       CI.getArgOperand(NumArgs - 2));
@@ -2510,7 +2510,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
                                           : Intrinsic::x86_avx512_sqrt_pd_512;
 
       Value *Args[] = {CI->getArgOperand(0), CI->getArgOperand(3)};
-      Rep = Builder.CreateIntrinsic(IID, {}, Args);
+      Rep = Builder.CreateIntrinsic(IID, Args);
     } else {
       Rep = Builder.CreateIntrinsic(Intrinsic::sqrt, CI->getType(),
                                     {CI->getArgOperand(0)});
@@ -2637,8 +2637,8 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
       break;
     }
 
-    Rep = Builder.CreateIntrinsic(IID, {},
-                                  {CI->getOperand(0), CI->getArgOperand(1)});
+    Rep =
+        Builder.CreateIntrinsic(IID, {CI->getOperand(0), CI->getArgOperand(1)});
     Rep = applyX86MaskOn1BitsVec(Builder, Rep, CI->getArgOperand(2));
   } else if (Name.starts_with("avx512.mask.fpclass.p")) {
     Type *OpTy = CI->getArgOperand(0)->getType();
@@ -2660,8 +2660,8 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
     else
       llvm_unreachable("Unexpected intrinsic");
 
-    Rep = Builder.CreateIntrinsic(IID, {},
-                                  {CI->getOperand(0), CI->getArgOperand(1)});
+    Rep =
+        Builder.CreateIntrinsic(IID, {CI->getOperand(0), CI->getArgOperand(1)});
     Rep = applyX86MaskOn1BitsVec(Builder, Rep, CI->getArgOperand(2));
   } else if (Name.starts_with("avx512.cmp.p")) {
     SmallVector<Value *, 4> Args(CI->args());
@@ -2689,7 +2689,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
       std::swap(Mask, Args.back());
     Args.push_back(Mask);
 
-    Rep = Builder.CreateIntrinsic(IID, {}, Args);
+    Rep = Builder.CreateIntrinsic(IID, Args);
   } else if (Name.starts_with("avx512.mask.cmp.")) {
     // Integer compare intrinsics.
     unsigned Imm = cast<ConstantInt>(CI->getArgOperand(2))->getZExtValue();
@@ -2905,7 +2905,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
   } else if (Name == "sse42.crc32.64.8") {
     Value *Trunc0 =
         Builder.CreateTrunc(CI->getArgOperand(0), Type::getInt32Ty(C));
-    Rep = Builder.CreateIntrinsic(Intrinsic::x86_sse42_crc32_32_8, {},
+    Rep = Builder.CreateIntrinsic(Intrinsic::x86_sse42_crc32_32_8,
                                   {Trunc0, CI->getArgOperand(1)});
     Rep = Builder.CreateZExt(Rep, CI->getType(), "");
   } else if (Name.starts_with("avx.vbroadcast.s") ||
@@ -3395,7 +3395,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
         IID = Intrinsic::x86_avx512_add_pd_512;
 
       Rep = Builder.CreateIntrinsic(
-          IID, {},
+          IID,
           {CI->getArgOperand(0), CI->getArgOperand(1), CI->getArgOperand(4)});
     } else {
       Rep = Builder.CreateFAdd(CI->getArgOperand(0), CI->getArgOperand(1));
@@ -3411,7 +3411,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
         IID = Intrinsic::x86_avx512_div_pd_512;
 
       Rep = Builder.CreateIntrinsic(
-          IID, {},
+          IID,
           {CI->getArgOperand(0), CI->getArgOperand(1), CI->getArgOperand(4)});
     } else {
       Rep = Builder.CreateFDiv(CI->getArgOperand(0), CI->getArgOperand(1));
@@ -3427,7 +3427,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
         IID = Intrinsic::x86_avx512_mul_pd_512;
 
       Rep = Builder.CreateIntrinsic(
-          IID, {},
+          IID,
           {CI->getArgOperand(0), CI->getArgOperand(1), CI->getArgOperand(4)});
     } else {
       Rep = Builder.CreateFMul(CI->getArgOperand(0), CI->getArgOperand(1));
@@ -3443,7 +3443,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
         IID = Intrinsic::x86_avx512_sub_pd_512;
 
       Rep = Builder.CreateIntrinsic(
-          IID, {},
+          IID,
           {CI->getArgOperand(0), CI->getArgOperand(1), CI->getArgOperand(4)});
     } else {
       Rep = Builder.CreateFSub(CI->getArgOperand(0), CI->getArgOperand(1));
@@ -3461,7 +3461,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
     Intrinsic::ID IID = MinMaxTbl[IsMin][IsDouble];
 
     Rep = Builder.CreateIntrinsic(
-        IID, {},
+        IID,
         {CI->getArgOperand(0), CI->getArgOperand(1), CI->getArgOperand(4)});
     Rep =
         emitX86Select(Builder, CI->getArgOperand(3), Rep, CI->getArgOperand(2));
@@ -3759,7 +3759,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
         IID = Intrinsic::x86_avx512_vfmadd_f64;
       else
         IID = Intrinsic::x86_avx512_vfmadd_f32;
-      Rep = Builder.CreateIntrinsic(IID, {}, Ops);
+      Rep = Builder.CreateIntrinsic(IID, Ops);
     } else {
       Rep = Builder.CreateFMA(A, B, C);
     }
@@ -3812,7 +3812,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
       else
         IID = Intrinsic::x86_avx512_vfmadd_pd_512;
 
-      Rep = Builder.CreateIntrinsic(IID, {}, {A, B, C, CI->getArgOperand(4)});
+      Rep = Builder.CreateIntrinsic(IID, {A, B, C, CI->getArgOperand(4)});
     } else {
       Rep = Builder.CreateFMA(A, B, C);
     }
@@ -3840,7 +3840,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
     Value *Ops[] = {CI->getArgOperand(0), CI->getArgOperand(1),
                     CI->getArgOperand(2)};
     Ops[2] = Builder.CreateFNeg(Ops[2]);
-    Rep = Builder.CreateIntrinsic(IID, {}, Ops);
+    Rep = Builder.CreateIntrinsic(IID, Ops);
   } else if (Name.starts_with("avx512.mask.vfmaddsub.p") ||
              Name.starts_with("avx512.mask3.vfmaddsub.p") ||
              Name.starts_with("avx512.maskz.vfmaddsub.p") ||
@@ -3863,7 +3863,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
       if (IsSubAdd)
         Ops[2] = Builder.CreateFNeg(Ops[2]);
 
-      Rep = Builder.CreateIntrinsic(IID, {}, Ops);
+      Rep = Builder.CreateIntrinsic(IID, Ops);
     } else {
       int NumElts = cast<FixedVectorType>(CI->getType())->getNumElements();
 
@@ -3914,7 +3914,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
 
     Value *Args[] = {CI->getArgOperand(0), CI->getArgOperand(1),
                      CI->getArgOperand(2), CI->getArgOperand(3)};
-    Rep = Builder.CreateIntrinsic(IID, {}, Args);
+    Rep = Builder.CreateIntrinsic(IID, Args);
     Value *PassThru = ZeroMask ? ConstantAggregateZero::get(CI->getType())
                                : CI->getArgOperand(0);
     Rep = emitX86Select(Builder, CI->getArgOperand(4), Rep, PassThru);
@@ -3941,7 +3941,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
 
     Value *Args[] = {CI->getArgOperand(0), CI->getArgOperand(1),
                      CI->getArgOperand(2)};
-    Rep = Builder.CreateIntrinsic(IID, {}, Args);
+    Rep = Builder.CreateIntrinsic(IID, Args);
     Value *PassThru = ZeroMask ? ConstantAggregateZero::get(CI->getType())
                                : CI->getArgOperand(0);
     Rep = emitX86Select(Builder, CI->getArgOperand(3), Rep, PassThru);
@@ -3976,7 +3976,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
 
     Value *Args[] = {CI->getArgOperand(0), CI->getArgOperand(1),
                      CI->getArgOperand(2)};
-    Rep = Builder.CreateIntrinsic(IID, {}, Args);
+    Rep = Builder.CreateIntrinsic(IID, Args);
     Value *PassThru = ZeroMask ? ConstantAggregateZero::get(CI->getType())
                                : CI->getArgOperand(0);
     Rep = emitX86Select(Builder, CI->getArgOperand(3), Rep, PassThru);
@@ -4005,7 +4005,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
 
     Value *Args[] = {CI->getArgOperand(0), CI->getArgOperand(1),
                      CI->getArgOperand(2)};
-    Rep = Builder.CreateIntrinsic(IID, {}, Args);
+    Rep = Builder.CreateIntrinsic(IID, Args);
     Value *PassThru = ZeroMask ? ConstantAggregateZero::get(CI->getType())
                                : CI->getArgOperand(0);
     Rep = emitX86Select(Builder, CI->getArgOperand(3), Rep, PassThru);
@@ -4027,7 +4027,7 @@ static Value *upgradeX86IntrinsicCall(StringRef Name, CallBase *CI, Function *F,
     // Make a call with 3 operands.
     Value *Args[] = {CI->getArgOperand(0), CI->getArgOperand(1),
                      CI->getArgOperand(2)};
-    Value *NewCall = Builder.CreateIntrinsic(IID, {}, Args);
+    Value *NewCall = Builder.CreateIntrinsic(IID, Args);
 
     // Extract the second result and store it.
     Value *Data = Builder.CreateExtractValue(NewCall, 1);
@@ -4095,7 +4095,7 @@ static Value *upgradeAArch64IntrinsicCall(StringRef Name, CallBase *CI,
     Args[1] = Builder.CreateIntrinsic(
         Intrinsic::aarch64_sve_convert_from_svbool, GoodPredTy, Args[1]);
 
-    return Builder.CreateIntrinsic(NewID, {}, Args, /*FMFSource=*/nullptr,
+    return Builder.CreateIntrinsic(NewID, Args, /*FMFSource=*/nullptr,
                                    CI->getName());
   }
 

llvm/lib/Target/AArch64/AArch64ISelLowering.cpp

@@ -28069,7 +28069,7 @@ Value *AArch64TargetLowering::emitLoadLinked(IRBuilderBase &Builder,
         IsAcquire ? Intrinsic::aarch64_ldaxp : Intrinsic::aarch64_ldxp;
 
     Value *LoHi =
-        Builder.CreateIntrinsic(Int, {}, Addr, /*FMFSource=*/nullptr, "lohi");
+        Builder.CreateIntrinsic(Int, Addr, /*FMFSource=*/nullptr, "lohi");
 
     Value *Lo = Builder.CreateExtractValue(LoHi, 0, "lo");
     Value *Hi = Builder.CreateExtractValue(LoHi, 1, "hi");
@@ -28099,7 +28099,7 @@ Value *AArch64TargetLowering::emitLoadLinked(IRBuilderBase &Builder,
 
 void AArch64TargetLowering::emitAtomicCmpXchgNoStoreLLBalance(
     IRBuilderBase &Builder) const {
-  Builder.CreateIntrinsic(Intrinsic::aarch64_clrex, {}, {});
+  Builder.CreateIntrinsic(Intrinsic::aarch64_clrex, {});
 }
 
 Value *AArch64TargetLowering::emitStoreConditional(IRBuilderBase &Builder,

llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp

@@ -1476,7 +1476,7 @@ static std::optional<Instruction *> instCombineRDFFR(InstCombiner &IC,
   auto *PTrue = IC.Builder.CreateIntrinsic(Intrinsic::aarch64_sve_ptrue,
                                            {II.getType()}, {AllPat});
   auto *RDFFR =
-      IC.Builder.CreateIntrinsic(Intrinsic::aarch64_sve_rdffr_z, {}, {PTrue});
+      IC.Builder.CreateIntrinsic(Intrinsic::aarch64_sve_rdffr_z, {PTrue});
   RDFFR->takeName(&II);
   return IC.replaceInstUsesWith(II, RDFFR);
 }

llvm/lib/Target/AMDGPU/AMDGPUAsanInstrumentation.cpp

@@ -73,7 +73,7 @@ static Instruction *genAMDGPUReportBlock(Module &M, IRBuilder<> &IRB,
 
   Trm = SplitBlockAndInsertIfThen(Cond, Trm, false);
   IRB.SetInsertPoint(Trm);
-  return IRB.CreateIntrinsic(Intrinsic::amdgcn_unreachable, {}, {});
+  return IRB.CreateIntrinsic(Intrinsic::amdgcn_unreachable, {});
 }
 
 static Value *createSlowPathCmp(Module &M, IRBuilder<> &IRB, Type *IntptrTy,

llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp

@@ -666,7 +666,7 @@ void AMDGPUAtomicOptimizerImpl::optimizeAtomic(Instruction &I,
     // Record I's original position as the entry block.
     PixelEntryBB = I.getParent();
 
-    Value *const Cond = B.CreateIntrinsic(Intrinsic::amdgcn_ps_live, {}, {});
+    Value *const Cond = B.CreateIntrinsic(Intrinsic::amdgcn_ps_live, {});
     Instruction *const NonHelperTerminator =
         SplitBlockAndInsertIfThen(Cond, &I, false, nullptr, &DTU, nullptr);
 
@@ -698,15 +698,14 @@ void AMDGPUAtomicOptimizerImpl::optimizeAtomic(Instruction &I,
   // using the mbcnt intrinsic.
   Value *Mbcnt;
   if (ST.isWave32()) {
-    Mbcnt = B.CreateIntrinsic(Intrinsic::amdgcn_mbcnt_lo, {},
-                              {Ballot, B.getInt32(0)});
+    Mbcnt =
+        B.CreateIntrinsic(Intrinsic::amdgcn_mbcnt_lo, {Ballot, B.getInt32(0)});
   } else {
     Value *const ExtractLo = B.CreateTrunc(Ballot, Int32Ty);
     Value *const ExtractHi = B.CreateTrunc(B.CreateLShr(Ballot, 32), Int32Ty);
-    Mbcnt = B.CreateIntrinsic(Intrinsic::amdgcn_mbcnt_lo, {},
+    Mbcnt = B.CreateIntrinsic(Intrinsic::amdgcn_mbcnt_lo,
                               {ExtractLo, B.getInt32(0)});
-    Mbcnt =
-        B.CreateIntrinsic(Intrinsic::amdgcn_mbcnt_hi, {}, {ExtractHi, Mbcnt});
+    Mbcnt = B.CreateIntrinsic(Intrinsic::amdgcn_mbcnt_hi, {ExtractHi, Mbcnt});
   }
 
   Function *F = I.getFunction();

llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp

@@ -1033,7 +1033,7 @@ Value *AMDGPUCodeGenPrepareImpl::optimizeWithFDivFast(
   if (!HasFP32DenormalFlush && !NumIsOne)
     return nullptr;
 
-  return Builder.CreateIntrinsic(Intrinsic::amdgcn_fdiv_fast, {}, {Num, Den});
+  return Builder.CreateIntrinsic(Intrinsic::amdgcn_fdiv_fast, {Num, Den});
 }
 
 Value *AMDGPUCodeGenPrepareImpl::visitFDivElement(

llvm/lib/Target/AMDGPU/AMDGPULowerKernelArguments.cpp

@@ -304,7 +304,7 @@ static bool lowerKernelArguments(Function &F, const TargetMachine &TM) {
     return false;
 
   CallInst *KernArgSegment =
-      Builder.CreateIntrinsic(Intrinsic::amdgcn_kernarg_segment_ptr, {}, {},
+      Builder.CreateIntrinsic(Intrinsic::amdgcn_kernarg_segment_ptr, {},
                               nullptr, F.getName() + ".kernarg.segment");
   KernArgSegment->addRetAttr(Attribute::NonNull);
   KernArgSegment->addRetAttr(

llvm/lib/Target/AMDGPU/AMDGPULowerModuleLDSPass.cpp

@@ -532,8 +532,7 @@ class AMDGPULowerModuleLDS {
       auto InsertAt = F->getEntryBlock().getFirstNonPHIOrDbgOrAlloca();
       IRBuilder<> Builder(&*InsertAt);
 
-      It->second =
-          Builder.CreateIntrinsic(Intrinsic::amdgcn_lds_kernel_id, {}, {});
+      It->second = Builder.CreateIntrinsic(Intrinsic::amdgcn_lds_kernel_id, {});
     }
 
     return It->second;