[CIR] Add support for array cleanups (#150499)

This adds support for array cleanups, including the ArrayDtor op.

Author: andykaylor

clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h

@@ -75,6 +75,12 @@ class CIRBaseBuilderTy : public mlir::OpBuilder {
     return getConstant(loc, cir::IntAttr::get(ty, value));
   }
 
+  mlir::Value getSignedInt(mlir::Location loc, int64_t val, unsigned numBits) {
+    auto type = cir::IntType::get(getContext(), numBits, /*isSigned=*/true);
+    return getConstAPInt(loc, type,
+                         llvm::APInt(numBits, val, /*isSigned=*/true));
+  }
+
   mlir::Value getUnsignedInt(mlir::Location loc, uint64_t val,
                              unsigned numBits) {
     auto type = cir::IntType::get(getContext(), numBits, /*isSigned=*/false);

clang/include/clang/CIR/Dialect/IR/CIROps.td

@@ -607,8 +607,8 @@ def CIR_ConditionOp : CIR_Op<"condition", [
 //===----------------------------------------------------------------------===//
 
 defvar CIR_YieldableScopes = [
-  "ArrayCtor", "CaseOp", "DoWhileOp", "ForOp", "IfOp", "ScopeOp", "SwitchOp",
-  "TernaryOp", "WhileOp"
+  "ArrayCtor", "ArrayDtor", "CaseOp", "DoWhileOp", "ForOp", "IfOp", "ScopeOp",
+  "SwitchOp", "TernaryOp", "WhileOp"
 ];
 
 def CIR_YieldOp : CIR_Op<"yield", [
@@ -2227,7 +2227,7 @@ def CIR_TrapOp : CIR_Op<"trap", [Terminator]> {
 }
 
 //===----------------------------------------------------------------------===//
-// ArrayCtor
+// ArrayCtor & ArrayDtor
 //===----------------------------------------------------------------------===//
 
 class CIR_ArrayInitDestroy<string mnemonic> : CIR_Op<mnemonic> {
@@ -2258,7 +2258,9 @@ def CIR_ArrayCtor : CIR_ArrayInitDestroy<"array.ctor"> {
   let description = [{
     Initialize each array element using the same C++ constructor. This
     operation has one region, with one single block. The block has an
-    incoming argument for the current array index to initialize.
+    incoming argument for the current array element to initialize.
+
+    Example:
 
     ```mlir
     cir.array.ctor(%0 : !cir.ptr<!cir.array<!rec_S x 42>>) {
@@ -2270,6 +2272,25 @@ def CIR_ArrayCtor : CIR_ArrayInitDestroy<"array.ctor"> {
   }];
 }
 
+def CIR_ArrayDtor : CIR_ArrayInitDestroy<"array.dtor"> {
+  let summary = "Destroy array elements with C++ dtors";
+  let description = [{
+    Destroy each array element using the same C++ destructor. This
+    operation has one region, with one single block. The block has an
+    incoming argument for the current array element to destruct.
+
+    Example:
+
+    ```mlir
+    cir.array.dtor(%0 : !cir.ptr<!cir.array<!rec_S x 42>>) {
+      ^bb0(%arg0: !cir.ptr<!rec_S>):
+        cir.call @some_dtor(%arg0) : (!cir.ptr<!rec_S>) -> ()
+        cir.yield
+    }
+    ```
+  }];
+}
+
 //===----------------------------------------------------------------------===//
 // VecCreate
 //===----------------------------------------------------------------------===//

clang/lib/CIR/CodeGen/CIRGenDecl.cpp

@@ -649,6 +649,38 @@ void CIRGenFunction::emitNullabilityCheck(LValue lhs, mlir::Value rhs,
   assert(!cir::MissingFeatures::sanitizers());
 }
 
+/// Destroys all the elements of the given array, beginning from last to first.
+/// The array cannot be zero-length.
+///
+/// \param begin - a type* denoting the first element of the array
+/// \param end - a type* denoting one past the end of the array
+/// \param elementType - the element type of the array
+/// \param destroyer - the function to call to destroy elements
+void CIRGenFunction::emitArrayDestroy(mlir::Value begin, mlir::Value end,
+                                      QualType elementType,
+                                      CharUnits elementAlign,
+                                      Destroyer *destroyer) {
+  assert(!elementType->isArrayType());
+
+  // Differently from LLVM traditional codegen, use a higher level
+  // representation instead of lowering directly to a loop.
+  mlir::Type cirElementType = convertTypeForMem(elementType);
+  cir::PointerType ptrToElmType = builder.getPointerTo(cirElementType);
+
+  // Emit the dtor call that will execute for every array element.
+  cir::ArrayDtor::create(
+      builder, *currSrcLoc, begin, [&](mlir::OpBuilder &b, mlir::Location loc) {
+        auto arg = b.getInsertionBlock()->addArgument(ptrToElmType, loc);
+        Address curAddr = Address(arg, cirElementType, elementAlign);
+        assert(!cir::MissingFeatures::dtorCleanups());
+
+        // Perform the actual destruction there.
+        destroyer(*this, curAddr, elementType);
+
+        cir::YieldOp::create(builder, loc);
+      });
+}
+
 /// Immediately perform the destruction of the given object.
 ///
 /// \param addr - the address of the object; a type*
@@ -658,10 +690,34 @@ void CIRGenFunction::emitNullabilityCheck(LValue lhs, mlir::Value rhs,
 ///   elements
 void CIRGenFunction::emitDestroy(Address addr, QualType type,
                                  Destroyer *destroyer) {
-  if (getContext().getAsArrayType(type))
-    cgm.errorNYI("emitDestroy: array type");
+  const ArrayType *arrayType = getContext().getAsArrayType(type);
+  if (!arrayType)
+    return destroyer(*this, addr, type);
+
+  mlir::Value length = emitArrayLength(arrayType, type, addr);
+
+  CharUnits elementAlign = addr.getAlignment().alignmentOfArrayElement(
+      getContext().getTypeSizeInChars(type));
+
+  auto constantCount = length.getDefiningOp<cir::ConstantOp>();
+  if (!constantCount) {
+    assert(!cir::MissingFeatures::vlas());
+    cgm.errorNYI("emitDestroy: variable length array");
+    return;
+  }
+
+  auto constIntAttr = mlir::dyn_cast<cir::IntAttr>(constantCount.getValue());
+  // If it's constant zero, we can just skip the entire thing.
+  if (constIntAttr && constIntAttr.getUInt() == 0)
+    return;
+
+  mlir::Value begin = addr.getPointer();
+  mlir::Value end; // This will be used for future non-constant counts.
+  emitArrayDestroy(begin, end, type, elementAlign, destroyer);
 
-  return destroyer(*this, addr, type);
+  // If the array destroy didn't use the length op, we can erase it.
+  if (constantCount.use_empty())
+    constantCount.erase();
 }
 
 CIRGenFunction::Destroyer *

clang/lib/CIR/CodeGen/CIRGenFunction.h

@@ -848,6 +848,10 @@ class CIRGenFunction : public CIRGenTypeCache {
   /// even if no aggregate location is provided.
   RValue emitAnyExprToTemp(const clang::Expr *e);
 
+  void emitArrayDestroy(mlir::Value begin, mlir::Value end,
+                        QualType elementType, CharUnits elementAlign,
+                        Destroyer *destroyer);
+
   mlir::Value emitArrayLength(const clang::ArrayType *arrayType,
                               QualType &baseType, Address &addr);
   LValue emitArraySubscriptExpr(const clang::ArraySubscriptExpr *e);

clang/lib/CIR/Dialect/Transforms/LoweringPrepare.cpp

@@ -29,7 +29,8 @@ struct LoweringPreparePass : public LoweringPrepareBase<LoweringPreparePass> {
   void runOnOp(mlir::Operation *op);
   void lowerCastOp(cir::CastOp op);
   void lowerUnaryOp(cir::UnaryOp op);
-  void lowerArrayCtor(ArrayCtor op);
+  void lowerArrayDtor(cir::ArrayDtor op);
+  void lowerArrayCtor(cir::ArrayCtor op);
 
   ///
   /// AST related
@@ -172,28 +173,30 @@ void LoweringPreparePass::lowerUnaryOp(cir::UnaryOp op) {
 static void lowerArrayDtorCtorIntoLoop(cir::CIRBaseBuilderTy &builder,
                                        clang::ASTContext *astCtx,
                                        mlir::Operation *op, mlir::Type eltTy,
-                                       mlir::Value arrayAddr,
-                                       uint64_t arrayLen) {
+                                       mlir::Value arrayAddr, uint64_t arrayLen,
+                                       bool isCtor) {
   // Generate loop to call into ctor/dtor for every element.
   mlir::Location loc = op->getLoc();
 
-  // TODO: instead of fixed integer size, create alias for PtrDiffTy and unify
-  // with CIRGen stuff.
+  // TODO: instead of getting the size from the AST context, create alias for
+  // PtrDiffTy and unify with CIRGen stuff.
   const unsigned sizeTypeSize =
       astCtx->getTypeSize(astCtx->getSignedSizeType());
-  auto ptrDiffTy =
-      cir::IntType::get(builder.getContext(), sizeTypeSize, /*isSigned=*/false);
-  mlir::Value numArrayElementsConst = builder.getUnsignedInt(loc, arrayLen, 64);
+  uint64_t endOffset = isCtor ? arrayLen : arrayLen - 1;
+  mlir::Value endOffsetVal =
+      builder.getUnsignedInt(loc, endOffset, sizeTypeSize);
 
-  auto begin = builder.create<cir::CastOp>(
-      loc, eltTy, cir::CastKind::array_to_ptrdecay, arrayAddr);
-  mlir::Value end = builder.create<cir::PtrStrideOp>(loc, eltTy, begin,
-                                                     numArrayElementsConst);
+  auto begin = cir::CastOp::create(builder, loc, eltTy,
+                                   cir::CastKind::array_to_ptrdecay, arrayAddr);
+  mlir::Value end =
+      cir::PtrStrideOp::create(builder, loc, eltTy, begin, endOffsetVal);
+  mlir::Value start = isCtor ? begin : end;
+  mlir::Value stop = isCtor ? end : begin;
 
   mlir::Value tmpAddr = builder.createAlloca(
       loc, /*addr type*/ builder.getPointerTo(eltTy),
       /*var type*/ eltTy, "__array_idx", builder.getAlignmentAttr(1));
-  builder.createStore(loc, begin, tmpAddr);
+  builder.createStore(loc, start, tmpAddr);
 
   cir::DoWhileOp loop = builder.createDoWhile(
       loc,
@@ -202,7 +205,7 @@ static void lowerArrayDtorCtorIntoLoop(cir::CIRBaseBuilderTy &builder,
         auto currentElement = b.create<cir::LoadOp>(loc, eltTy, tmpAddr);
         mlir::Type boolTy = cir::BoolType::get(b.getContext());
         auto cmp = builder.create<cir::CmpOp>(loc, boolTy, cir::CmpOpKind::ne,
-                                              currentElement, end);
+                                              currentElement, stop);
         builder.createCondition(cmp);
       },
       /*bodyBuilder=*/
@@ -213,15 +216,19 @@ static void lowerArrayDtorCtorIntoLoop(cir::CIRBaseBuilderTy &builder,
         op->walk([&](cir::CallOp c) { ctorCall = c; });
         assert(ctorCall && "expected ctor call");
 
-        auto one = builder.create<cir::ConstantOp>(
-            loc, ptrDiffTy, cir::IntAttr::get(ptrDiffTy, 1));
+        // Array elements get constructed in order but destructed in reverse.
+        mlir::Value stride;
+        if (isCtor)
+          stride = builder.getUnsignedInt(loc, 1, sizeTypeSize);
+        else
+          stride = builder.getSignedInt(loc, -1, sizeTypeSize);
 
-        ctorCall->moveAfter(one);
+        ctorCall->moveBefore(stride.getDefiningOp());
         ctorCall->setOperand(0, currentElement);
+        auto nextElement = cir::PtrStrideOp::create(builder, loc, eltTy,
+                                                    currentElement, stride);
 
-        // Advance pointer and store them to temporary variable
-        auto nextElement =
-            builder.create<cir::PtrStrideOp>(loc, eltTy, currentElement, one);
+        // Store the element pointer to the temporary variable
         builder.createStore(loc, nextElement, tmpAddr);
         builder.createYield(loc);
       });
@@ -230,6 +237,18 @@ static void lowerArrayDtorCtorIntoLoop(cir::CIRBaseBuilderTy &builder,
   op->erase();
 }
 
+void LoweringPreparePass::lowerArrayDtor(cir::ArrayDtor op) {
+  CIRBaseBuilderTy builder(getContext());
+  builder.setInsertionPointAfter(op.getOperation());
+
+  mlir::Type eltTy = op->getRegion(0).getArgument(0).getType();
+  assert(!cir::MissingFeatures::vlas());
+  auto arrayLen =
+      mlir::cast<cir::ArrayType>(op.getAddr().getType().getPointee()).getSize();
+  lowerArrayDtorCtorIntoLoop(builder, astCtx, op, eltTy, op.getAddr(), arrayLen,
+                             false);
+}
+
 void LoweringPreparePass::lowerArrayCtor(cir::ArrayCtor op) {
   cir::CIRBaseBuilderTy builder(getContext());
   builder.setInsertionPointAfter(op.getOperation());
@@ -238,13 +257,15 @@ void LoweringPreparePass::lowerArrayCtor(cir::ArrayCtor op) {
   assert(!cir::MissingFeatures::vlas());
   auto arrayLen =
       mlir::cast<cir::ArrayType>(op.getAddr().getType().getPointee()).getSize();
-  lowerArrayDtorCtorIntoLoop(builder, astCtx, op, eltTy, op.getAddr(),
-                             arrayLen);
+  lowerArrayDtorCtorIntoLoop(builder, astCtx, op, eltTy, op.getAddr(), arrayLen,
+                             true);
 }
 
 void LoweringPreparePass::runOnOp(mlir::Operation *op) {
   if (auto arrayCtor = dyn_cast<ArrayCtor>(op))
     lowerArrayCtor(arrayCtor);
+  else if (auto arrayDtor = dyn_cast<cir::ArrayDtor>(op))
+    lowerArrayDtor(arrayDtor);
   else if (auto cast = mlir::dyn_cast<cir::CastOp>(op))
     lowerCastOp(cast);
   else if (auto unary = mlir::dyn_cast<cir::UnaryOp>(op))
@@ -257,7 +278,8 @@ void LoweringPreparePass::runOnOperation() {
   llvm::SmallVector<mlir::Operation *> opsToTransform;
 
   op->walk([&](mlir::Operation *op) {
-    if (mlir::isa<cir::ArrayCtor, cir::CastOp, cir::UnaryOp>(op))
+    if (mlir::isa<cir::ArrayCtor, cir::ArrayDtor, cir::CastOp, cir::UnaryOp>(
+            op))
       opsToTransform.push_back(op);
   });
 

clang/test/CIR/CodeGen/array-ctor.cpp

@@ -33,8 +33,8 @@ void foo() {
 // CIR:   cir.store %[[DECAY]], %[[ITER]] : !cir.ptr<!rec_S>, !cir.ptr<!cir.ptr<!rec_S>>
 // CIR:   cir.do {
 // CIR:     %[[CURRENT:.*]] = cir.load %[[ITER]] : !cir.ptr<!cir.ptr<!rec_S>>, !cir.ptr<!rec_S>
-// CIR:     %[[CONST1:.*]] = cir.const #cir.int<1> : !u64i
 // CIR:     cir.call @_ZN1SC1Ev(%[[CURRENT]]) : (!cir.ptr<!rec_S>) -> ()
+// CIR:     %[[CONST1:.*]] = cir.const #cir.int<1> : !u64i
 // CIR:     %[[NEXT:.*]] = cir.ptr_stride(%[[CURRENT]] : !cir.ptr<!rec_S>, %[[CONST1]] : !u64i), !cir.ptr<!rec_S>
 // CIR:     cir.store %[[NEXT]], %[[ITER]] : !cir.ptr<!rec_S>, !cir.ptr<!cir.ptr<!rec_S>>
 // CIR:     cir.yield