[CIR] Add CIRGen for cir.unreachable and cir.trap (#151363)

Author: Lancern

clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp

@@ -21,6 +21,7 @@
 #include "mlir/Support/LLVM.h"
 #include "clang/AST/Expr.h"
 #include "clang/AST/GlobalDecl.h"
+#include "clang/Basic/Builtins.h"
 #include "clang/CIR/MissingFeatures.h"
 #include "llvm/Support/ErrorHandling.h"
 
@@ -269,6 +270,14 @@ RValue CIRGenFunction::emitBuiltinExpr(const GlobalDecl &gd, unsigned builtinID,
   case Builtin::BI__builtin_rotateright32:
   case Builtin::BI__builtin_rotateright64:
     return emitRotate(e, /*isRotateLeft=*/false);
+
+  case Builtin::BI__builtin_trap:
+    emitTrap(loc, /*createNewBlock=*/true);
+    return RValue::get(nullptr);
+
+  case Builtin::BI__builtin_unreachable:
+    emitUnreachable(e->getExprLoc(), /*createNewBlock=*/true);
+    return RValue::get(nullptr);
   }
 
   // If this is an alias for a lib function (e.g. __builtin_sin), emit

clang/lib/CIR/CodeGen/CIRGenExpr.cpp

@@ -1939,6 +1939,20 @@ LValue CIRGenFunction::emitLoadOfReferenceLValue(Address refAddr,
                         pointeeBaseInfo);
 }
 
+void CIRGenFunction::emitTrap(mlir::Location loc, bool createNewBlock) {
+  cir::TrapOp::create(builder, loc);
+  if (createNewBlock)
+    builder.createBlock(builder.getBlock()->getParent());
+}
+
+void CIRGenFunction::emitUnreachable(clang::SourceLocation loc,
+                                     bool createNewBlock) {
+  assert(!cir::MissingFeatures::sanitizers());
+  cir::UnreachableOp::create(builder, getLoc(loc));
+  if (createNewBlock)
+    builder.createBlock(builder.getBlock()->getParent());
+}
+
 mlir::Value CIRGenFunction::createDummyValue(mlir::Location loc,
                                              clang::QualType qt) {
   mlir::Type t = convertType(qt);

clang/lib/CIR/CodeGen/CIRGenFunction.cpp

@@ -383,6 +383,7 @@ void CIRGenFunction::LexicalScope::emitImplicitReturn() {
         !mayDropFunctionReturn(fd->getASTContext(), fd->getReturnType());
 
     if (shouldEmitUnreachable) {
+      assert(!cir::MissingFeatures::sanitizers());
       if (cgf.cgm.getCodeGenOpts().OptimizationLevel == 0)
         builder.create<cir::TrapOp>(localScope->endLoc);
       else

clang/lib/CIR/CodeGen/CIRGenFunction.h

@@ -1223,8 +1223,28 @@ class CIRGenFunction : public CIRGenTypeCache {
   /// to conserve the high level information.
   mlir::Value emitToMemory(mlir::Value value, clang::QualType ty);
 
+  /// Emit a trap instruction, which is used to abort the program in an abnormal
+  /// way, usually for debugging purposes.
+  /// \p createNewBlock indicates whether to create a new block for the IR
+  /// builder. Since the `cir.trap` operation is a terminator, operations that
+  /// follow a trap cannot be emitted after `cir.trap` in the same block. To
+  /// ensure these operations get emitted successfully, you need to create a new
+  /// dummy block and set the insertion point there before continuing from the
+  /// trap operation.
+  void emitTrap(mlir::Location loc, bool createNewBlock);
+
   LValue emitUnaryOpLValue(const clang::UnaryOperator *e);
 
+  /// Emit a reached-unreachable diagnostic if \p loc is valid and runtime
+  /// checking is enabled. Otherwise, just emit an unreachable instruction.
+  /// \p createNewBlock indicates whether to create a new block for the IR
+  /// builder. Since the `cir.unreachable` operation is a terminator, operations
+  /// that follow an unreachable point cannot be emitted after `cir.unreachable`
+  /// in the same block. To ensure these operations get emitted successfully,
+  /// you need to create a dummy block and set the insertion point there before
+  /// continuing from the unreachable point.
+  void emitUnreachable(clang::SourceLocation loc, bool createNewBlock);
+
   /// This method handles emission of any variable declaration
   /// inside a function, including static vars etc.
   void emitVarDecl(const clang::VarDecl &d);

clang/lib/CIR/Lowering/DirectToLLVM/LowerToLLVM.cpp

@@ -2212,7 +2212,8 @@ void ConvertCIRToLLVMPass::runOnOperation() {
                CIRToLLVMVecShuffleDynamicOpLowering,
                CIRToLLVMVecShuffleOpLowering,
                CIRToLLVMVecSplatOpLowering,
-               CIRToLLVMVecTernaryOpLowering
+               CIRToLLVMVecTernaryOpLowering,
+               CIRToLLVMUnreachableOpLowering
       // clang-format on
       >(converter, patterns.getContext());
 
@@ -2268,6 +2269,13 @@ mlir::LogicalResult CIRToLLVMGetMemberOpLowering::matchAndRewrite(
   }
 }
 
+mlir::LogicalResult CIRToLLVMUnreachableOpLowering::matchAndRewrite(
+    cir::UnreachableOp op, OpAdaptor adaptor,
+    mlir::ConversionPatternRewriter &rewriter) const {
+  rewriter.replaceOpWithNewOp<mlir::LLVM::UnreachableOp>(op);
+  return mlir::success();
+}
+
 mlir::LogicalResult CIRToLLVMTrapOpLowering::matchAndRewrite(
     cir::TrapOp op, OpAdaptor adaptor,
     mlir::ConversionPatternRewriter &rewriter) const {

clang/lib/CIR/Lowering/DirectToLLVM/LowerToLLVM.h

@@ -402,6 +402,16 @@ class CIRToLLVMGetMemberOpLowering
                   mlir::ConversionPatternRewriter &) const override;
 };
 
+class CIRToLLVMUnreachableOpLowering
+    : public mlir::OpConversionPattern<cir::UnreachableOp> {
+public:
+  using mlir::OpConversionPattern<cir::UnreachableOp>::OpConversionPattern;
+
+  mlir::LogicalResult
+  matchAndRewrite(cir::UnreachableOp op, OpAdaptor,
+                  mlir::ConversionPatternRewriter &) const override;
+};
+
 class CIRToLLVMTrapOpLowering : public mlir::OpConversionPattern<cir::TrapOp> {
 public:
   using mlir::OpConversionPattern<cir::TrapOp>::OpConversionPattern;

clang/test/CIR/CodeGen/builtin_call.cpp

@@ -166,3 +166,63 @@ void expect_prob(int x, int y) {
 // LLVM-NEXT:    %[[Y_LONG:.+]] = sext i32 %[[Y]] to i64
 // LLVM-NEXT:    %{{.+}} = call i64 @llvm.expect.with.probability.i64(i64 %[[X_LONG]], i64 %[[Y_LONG]], double 2.500000e-01)
 // LLVM:       }
+
+void unreachable() {
+  __builtin_unreachable();
+}
+
+// CIR-LABEL: @_Z11unreachablev
+// CIR:         cir.unreachable
+// CIR:       }
+
+// LLVM-LABEL: @_Z11unreachablev
+// LLVM:         unreachable
+// LLVM:       }
+
+void f1();
+void unreachable2() {
+  __builtin_unreachable();
+  f1();
+}
+
+// CIR-LABEL: @_Z12unreachable2v
+// CIR:         cir.unreachable
+// CIR-NEXT:  ^{{.+}}:
+// CIR-NEXT:    cir.call @_Z2f1v() : () -> ()
+// CIR:       }
+
+// LLVM-LABEL: @_Z12unreachable2v
+// LLVM:         unreachable
+// LLVM:       {{.+}}:
+// LLVM-NEXT:    call void @_Z2f1v()
+// LLVM:       }
+
+void trap() {
+  __builtin_trap();
+}
+
+// CIR-LABEL: @_Z4trapv
+// CIR:         cir.trap
+// CIR:       }
+
+// LLVM-LABEL: @_Z4trapv
+// LLVM:         call void @llvm.trap()
+// LLVM:       }
+
+void trap2() {
+  __builtin_trap();
+  f1();
+}
+
+// CIR-LABEL: @_Z5trap2v
+// CIR:         cir.trap
+// CIR-NEXT:  ^{{.+}}:
+// CIR-NEXT:    cir.call @_Z2f1v() : () -> ()
+// CIR:       }
+
+// LLVM-LABEL: @_Z5trap2v
+// LLVM:         call void @llvm.trap()
+// LLVM-NEXT:    unreachable
+// LLVM:       {{.+}}:
+// LLVM-NEXT:    call void @_Z2f1v()
+// LLVM:       }