[Mlir-commits] [mlir] 7df7612 - [mlir][NFC] Migrate rest of the dialects to the new fold API
Markus Böck
llvmlistbot at llvm.org
Wed Jan 11 12:47:35 PST 2023
Author: Markus Böck
Date: 2023-01-11T21:47:25+01:00
New Revision: 7df761217cd7d0026ffff23c4bdac846bb60f185
URL: https://github.com/llvm/llvm-project/commit/7df761217cd7d0026ffff23c4bdac846bb60f185
DIFF: https://github.com/llvm/llvm-project/commit/7df761217cd7d0026ffff23c4bdac846bb60f185.diff
LOG: [mlir][NFC] Migrate rest of the dialects to the new fold API
Added:
Modified:
mlir/include/mlir/Dialect/Affine/IR/AffineOps.td
mlir/include/mlir/Dialect/Bufferization/IR/BufferizationBase.td
mlir/include/mlir/Dialect/Complex/IR/ComplexBase.td
mlir/include/mlir/Dialect/EmitC/IR/EmitCBase.td
mlir/include/mlir/Dialect/Func/IR/FuncOps.td
mlir/include/mlir/Dialect/GPU/IR/GPUBase.td
mlir/include/mlir/Dialect/LLVMIR/LLVMOpBase.td
mlir/include/mlir/Dialect/Linalg/IR/LinalgBase.td
mlir/include/mlir/Dialect/Quant/QuantOpsBase.td
mlir/include/mlir/Dialect/SCF/IR/SCFOps.td
mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorBase.td
mlir/include/mlir/Dialect/Transform/IR/TransformDialect.td
mlir/include/mlir/IR/BuiltinDialect.td
mlir/lib/Dialect/Affine/IR/AffineOps.cpp
mlir/lib/Dialect/Bufferization/IR/BufferizationOps.cpp
mlir/lib/Dialect/Complex/IR/ComplexOps.cpp
mlir/lib/Dialect/EmitC/IR/EmitC.cpp
mlir/lib/Dialect/Func/IR/FuncOps.cpp
mlir/lib/Dialect/GPU/IR/GPUDialect.cpp
mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp
mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
mlir/lib/Dialect/Quant/IR/QuantOps.cpp
mlir/lib/Dialect/SCF/IR/SCF.cpp
mlir/lib/Dialect/SparseTensor/IR/SparseTensorDialect.cpp
mlir/lib/Dialect/Transform/IR/TransformOps.cpp
mlir/lib/IR/BuiltinDialect.cpp
mlir/test/lib/Dialect/Test/TestDialect.cpp
mlir/test/lib/Dialect/Test/TestDialect.td
mlir/test/lib/Dialect/Test/TestOps.td
mlir/test/lib/Dialect/Test/TestTraits.cpp
mlir/tools/mlir-linalg-ods-gen/mlir-linalg-ods-yaml-gen.cpp
Removed:
################################################################################
diff --git a/mlir/include/mlir/Dialect/Affine/IR/AffineOps.td b/mlir/include/mlir/Dialect/Affine/IR/AffineOps.td
index d19e4d233c952..4a383ec3a6bf3 100644
--- a/mlir/include/mlir/Dialect/Affine/IR/AffineOps.td
+++ b/mlir/include/mlir/Dialect/Affine/IR/AffineOps.td
@@ -24,6 +24,7 @@ def Affine_Dialect : Dialect {
let cppNamespace = "mlir";
let hasConstantMaterializer = 1;
let dependentDialects = ["arith::ArithDialect"];
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
// Base class for Affine dialect ops.
diff --git a/mlir/include/mlir/Dialect/Bufferization/IR/BufferizationBase.td b/mlir/include/mlir/Dialect/Bufferization/IR/BufferizationBase.td
index 280bfdb380177..ecb8ec918cf72 100644
--- a/mlir/include/mlir/Dialect/Bufferization/IR/BufferizationBase.td
+++ b/mlir/include/mlir/Dialect/Bufferization/IR/BufferizationBase.td
@@ -69,6 +69,7 @@ def Bufferization_Dialect : Dialect {
kEscapeAttrName = "bufferization.escape";
}];
let hasOperationAttrVerify = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
#endif // BUFFERIZATION_BASE
diff --git a/mlir/include/mlir/Dialect/Complex/IR/ComplexBase.td b/mlir/include/mlir/Dialect/Complex/IR/ComplexBase.td
index 31135fc8c8ce7..20bc712b003b3 100644
--- a/mlir/include/mlir/Dialect/Complex/IR/ComplexBase.td
+++ b/mlir/include/mlir/Dialect/Complex/IR/ComplexBase.td
@@ -22,6 +22,7 @@ def Complex_Dialect : Dialect {
let dependentDialects = ["arith::ArithDialect"];
let hasConstantMaterializer = 1;
let useDefaultAttributePrinterParser = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
#endif // COMPLEX_BASE
diff --git a/mlir/include/mlir/Dialect/EmitC/IR/EmitCBase.td b/mlir/include/mlir/Dialect/EmitC/IR/EmitCBase.td
index 375dbcbce1d03..19b2a3281086e 100644
--- a/mlir/include/mlir/Dialect/EmitC/IR/EmitCBase.td
+++ b/mlir/include/mlir/Dialect/EmitC/IR/EmitCBase.td
@@ -31,6 +31,7 @@ def EmitC_Dialect : Dialect {
let hasConstantMaterializer = 1;
let useDefaultTypePrinterParser = 1;
let useDefaultAttributePrinterParser = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
#endif // MLIR_DIALECT_EMITC_IR_EMITCBASE
diff --git a/mlir/include/mlir/Dialect/Func/IR/FuncOps.td b/mlir/include/mlir/Dialect/Func/IR/FuncOps.td
index 4922689fbef60..58a33a124042b 100644
--- a/mlir/include/mlir/Dialect/Func/IR/FuncOps.td
+++ b/mlir/include/mlir/Dialect/Func/IR/FuncOps.td
@@ -23,6 +23,7 @@ def Func_Dialect : Dialect {
let cppNamespace = "::mlir::func";
let dependentDialects = ["cf::ControlFlowDialect"];
let hasConstantMaterializer = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
// Base class for Func dialect ops.
diff --git a/mlir/include/mlir/Dialect/GPU/IR/GPUBase.td b/mlir/include/mlir/Dialect/GPU/IR/GPUBase.td
index d5b36e495493d..7b02ed7c707ee 100644
--- a/mlir/include/mlir/Dialect/GPU/IR/GPUBase.td
+++ b/mlir/include/mlir/Dialect/GPU/IR/GPUBase.td
@@ -56,6 +56,7 @@ def GPU_Dialect : Dialect {
let dependentDialects = ["arith::ArithDialect"];
let useDefaultAttributePrinterParser = 1;
let useDefaultTypePrinterParser = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
def GPU_AsyncToken : DialectType<
diff --git a/mlir/include/mlir/Dialect/LLVMIR/LLVMOpBase.td b/mlir/include/mlir/Dialect/LLVMIR/LLVMOpBase.td
index 80f82650a10f8..37365f648ccc4 100644
--- a/mlir/include/mlir/Dialect/LLVMIR/LLVMOpBase.td
+++ b/mlir/include/mlir/Dialect/LLVMIR/LLVMOpBase.td
@@ -31,6 +31,7 @@ def LLVM_Dialect : Dialect {
let hasRegionArgAttrVerify = 1;
let hasRegionResultAttrVerify = 1;
let hasOperationAttrVerify = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
let extraClassDeclaration = [{
/// Name of the data layout attributes.
diff --git a/mlir/include/mlir/Dialect/Linalg/IR/LinalgBase.td b/mlir/include/mlir/Dialect/Linalg/IR/LinalgBase.td
index 1ee9d2ac241f6..706b2ffe0cee7 100644
--- a/mlir/include/mlir/Dialect/Linalg/IR/LinalgBase.td
+++ b/mlir/include/mlir/Dialect/Linalg/IR/LinalgBase.td
@@ -46,6 +46,7 @@ def Linalg_Dialect : Dialect {
let hasCanonicalizer = 1;
let hasOperationAttrVerify = 1;
let hasConstantMaterializer = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
let extraClassDeclaration = [{
/// Attribute name used to to memoize indexing maps for named ops.
constexpr const static ::llvm::StringLiteral
diff --git a/mlir/include/mlir/Dialect/Quant/QuantOpsBase.td b/mlir/include/mlir/Dialect/Quant/QuantOpsBase.td
index 74d69c046f06b..4a2e11d299d89 100644
--- a/mlir/include/mlir/Dialect/Quant/QuantOpsBase.td
+++ b/mlir/include/mlir/Dialect/Quant/QuantOpsBase.td
@@ -20,6 +20,7 @@ def Quantization_Dialect : Dialect {
let cppNamespace = "::mlir::quant";
let useDefaultTypePrinterParser = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
//===----------------------------------------------------------------------===//
diff --git a/mlir/include/mlir/Dialect/SCF/IR/SCFOps.td b/mlir/include/mlir/Dialect/SCF/IR/SCFOps.td
index d5a150557163c..a610562b281ae 100644
--- a/mlir/include/mlir/Dialect/SCF/IR/SCFOps.td
+++ b/mlir/include/mlir/Dialect/SCF/IR/SCFOps.td
@@ -25,6 +25,7 @@ def SCF_Dialect : Dialect {
let name = "scf";
let cppNamespace = "::mlir::scf";
let dependentDialects = ["arith::ArithDialect"];
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
// Base class for SCF dialect ops.
diff --git a/mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorBase.td b/mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorBase.td
index 2506da4042bd3..12066f3096301 100644
--- a/mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorBase.td
+++ b/mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorBase.td
@@ -83,6 +83,7 @@ def SparseTensor_Dialect : Dialect {
let useDefaultAttributePrinterParser = 1;
let useDefaultTypePrinterParser = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
#endif // SPARSETENSOR_BASE
diff --git a/mlir/include/mlir/Dialect/Transform/IR/TransformDialect.td b/mlir/include/mlir/Dialect/Transform/IR/TransformDialect.td
index a7bb75e9c7e70..05b094cefd992 100644
--- a/mlir/include/mlir/Dialect/Transform/IR/TransformDialect.td
+++ b/mlir/include/mlir/Dialect/Transform/IR/TransformDialect.td
@@ -23,6 +23,8 @@ def Transform_Dialect : Dialect {
"::mlir::pdl_interp::PDLInterpDialect",
];
+ let useFoldAPI = kEmitFoldAdaptorFolder;
+
let extraClassDeclaration = [{
/// Returns the named PDL constraint functions available in the dialect
/// as a map from their name to the function.
diff --git a/mlir/include/mlir/IR/BuiltinDialect.td b/mlir/include/mlir/IR/BuiltinDialect.td
index a3d0f0cbfd9a4..e9d00968c4599 100644
--- a/mlir/include/mlir/IR/BuiltinDialect.td
+++ b/mlir/include/mlir/IR/BuiltinDialect.td
@@ -34,6 +34,8 @@ def Builtin_Dialect : Dialect {
public:
}];
+
+ let useFoldAPI = kEmitFoldAdaptorFolder;
}
#endif // BUILTIN_BASE
diff --git a/mlir/lib/Dialect/Affine/IR/AffineOps.cpp b/mlir/lib/Dialect/Affine/IR/AffineOps.cpp
index be58118d32474..f4c851c1823c6 100644
--- a/mlir/lib/Dialect/Affine/IR/AffineOps.cpp
+++ b/mlir/lib/Dialect/Affine/IR/AffineOps.cpp
@@ -562,7 +562,7 @@ bool AffineApplyOp::isValidSymbol(Region *region) {
});
}
-OpFoldResult AffineApplyOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult AffineApplyOp::fold(FoldAdaptor adaptor) {
auto map = getAffineMap();
// Fold dims and symbols to existing values.
@@ -574,7 +574,7 @@ OpFoldResult AffineApplyOp::fold(ArrayRef<Attribute> operands) {
// Otherwise, default to folding the map.
SmallVector<Attribute, 1> result;
- if (failed(map.constantFold(operands, result)))
+ if (failed(map.constantFold(adaptor.getMapOperands(), result)))
return {};
return result[0];
}
@@ -2135,7 +2135,7 @@ static bool hasTrivialZeroTripCount(AffineForOp op) {
return tripCount && *tripCount == 0;
}
-LogicalResult AffineForOp::fold(ArrayRef<Attribute> operands,
+LogicalResult AffineForOp::fold(FoldAdaptor adaptor,
SmallVectorImpl<OpFoldResult> &results) {
bool folded = succeeded(foldLoopBounds(*this));
folded |= succeeded(canonicalizeLoopBounds(*this));
@@ -2723,7 +2723,7 @@ static void composeSetAndOperands(IntegerSet &set,
}
/// Canonicalize an affine if op's conditional (integer set + operands).
-LogicalResult AffineIfOp::fold(ArrayRef<Attribute>,
+LogicalResult AffineIfOp::fold(FoldAdaptor,
SmallVectorImpl<OpFoldResult> &) {
auto set = getIntegerSet();
SmallVector<Value, 4> operands(getOperands());
@@ -2858,7 +2858,7 @@ void AffineLoadOp::getCanonicalizationPatterns(RewritePatternSet &results,
results.add<SimplifyAffineOp<AffineLoadOp>>(context);
}
-OpFoldResult AffineLoadOp::fold(ArrayRef<Attribute> cstOperands) {
+OpFoldResult AffineLoadOp::fold(FoldAdaptor adaptor) {
/// load(memrefcast) -> load
if (succeeded(memref::foldMemRefCast(*this)))
return getResult();
@@ -2975,7 +2975,7 @@ void AffineStoreOp::getCanonicalizationPatterns(RewritePatternSet &results,
results.add<SimplifyAffineOp<AffineStoreOp>>(context);
}
-LogicalResult AffineStoreOp::fold(ArrayRef<Attribute> cstOperands,
+LogicalResult AffineStoreOp::fold(FoldAdaptor adaptor,
SmallVectorImpl<OpFoldResult> &results) {
/// store(memrefcast) -> store
return memref::foldMemRefCast(*this, getValueToStore());
@@ -3282,8 +3282,8 @@ struct CanonicalizeSingleResultAffineMinMaxOp : public OpRewritePattern<T> {
// %0 = affine.min (d0) -> (1000, d0 + 512) (%i0)
//
-OpFoldResult AffineMinOp::fold(ArrayRef<Attribute> operands) {
- return foldMinMaxOp(*this, operands);
+OpFoldResult AffineMinOp::fold(FoldAdaptor adaptor) {
+ return foldMinMaxOp(*this, adaptor.getOperands());
}
void AffineMinOp::getCanonicalizationPatterns(RewritePatternSet &patterns,
@@ -3310,8 +3310,8 @@ void AffineMinOp::print(OpAsmPrinter &p) { printAffineMinMaxOp(p, *this); }
// %0 = affine.max (d0) -> (1000, d0 + 512) (%i0)
//
-OpFoldResult AffineMaxOp::fold(ArrayRef<Attribute> operands) {
- return foldMinMaxOp(*this, operands);
+OpFoldResult AffineMaxOp::fold(FoldAdaptor adaptor) {
+ return foldMinMaxOp(*this, adaptor.getOperands());
}
void AffineMaxOp::getCanonicalizationPatterns(RewritePatternSet &patterns,
@@ -3431,7 +3431,7 @@ void AffinePrefetchOp::getCanonicalizationPatterns(RewritePatternSet &results,
results.add<SimplifyAffineOp<AffinePrefetchOp>>(context);
}
-LogicalResult AffinePrefetchOp::fold(ArrayRef<Attribute> cstOperands,
+LogicalResult AffinePrefetchOp::fold(FoldAdaptor adaptor,
SmallVectorImpl<OpFoldResult> &results) {
/// prefetch(memrefcast) -> prefetch
return memref::foldMemRefCast(*this);
@@ -3705,7 +3705,7 @@ static LogicalResult canonicalizeLoopBounds(AffineParallelOp op) {
return success();
}
-LogicalResult AffineParallelOp::fold(ArrayRef<Attribute> operands,
+LogicalResult AffineParallelOp::fold(FoldAdaptor adaptor,
SmallVectorImpl<OpFoldResult> &results) {
return canonicalizeLoopBounds(*this);
}
diff --git a/mlir/lib/Dialect/Bufferization/IR/BufferizationOps.cpp b/mlir/lib/Dialect/Bufferization/IR/BufferizationOps.cpp
index f021a593f2415..59263f699032b 100644
--- a/mlir/lib/Dialect/Bufferization/IR/BufferizationOps.cpp
+++ b/mlir/lib/Dialect/Bufferization/IR/BufferizationOps.cpp
@@ -458,7 +458,7 @@ void CloneOp::getEffects(
SideEffects::DefaultResource::get());
}
-OpFoldResult CloneOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult CloneOp::fold(FoldAdaptor adaptor) {
return succeeded(memref::foldMemRefCast(*this)) ? getResult() : Value();
}
@@ -560,7 +560,7 @@ LogicalResult DeallocTensorOp::bufferize(RewriterBase &rewriter,
// ToTensorOp
//===----------------------------------------------------------------------===//
-OpFoldResult ToTensorOp::fold(ArrayRef<Attribute>) {
+OpFoldResult ToTensorOp::fold(FoldAdaptor) {
if (auto toMemref = getMemref().getDefiningOp<ToMemrefOp>())
// Approximate alias analysis by conservatively folding only when no there
// is no interleaved operation.
@@ -596,7 +596,7 @@ void ToTensorOp::getCanonicalizationPatterns(RewritePatternSet &results,
// ToMemrefOp
//===----------------------------------------------------------------------===//
-OpFoldResult ToMemrefOp::fold(ArrayRef<Attribute>) {
+OpFoldResult ToMemrefOp::fold(FoldAdaptor) {
if (auto memrefToTensor = getTensor().getDefiningOp<ToTensorOp>())
if (memrefToTensor.getMemref().getType() == getType())
return memrefToTensor.getMemref();
diff --git a/mlir/lib/Dialect/Complex/IR/ComplexOps.cpp b/mlir/lib/Dialect/Complex/IR/ComplexOps.cpp
index 44ebf69a65199..c71e2e08b6ffb 100644
--- a/mlir/lib/Dialect/Complex/IR/ComplexOps.cpp
+++ b/mlir/lib/Dialect/Complex/IR/ComplexOps.cpp
@@ -17,8 +17,7 @@ using namespace mlir::complex;
// ConstantOp
//===----------------------------------------------------------------------===//
-OpFoldResult ConstantOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.empty() && "constant has no operands");
+OpFoldResult ConstantOp::fold(FoldAdaptor adaptor) {
return getValue();
}
@@ -68,8 +67,7 @@ LogicalResult ConstantOp::verify() {
// CreateOp
//===----------------------------------------------------------------------===//
-OpFoldResult CreateOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 2 && "binary op takes two operands");
+OpFoldResult CreateOp::fold(FoldAdaptor adaptor) {
// Fold complex.create(complex.re(op), complex.im(op)).
if (auto reOp = getOperand(0).getDefiningOp<ReOp>()) {
if (auto imOp = getOperand(1).getDefiningOp<ImOp>()) {
@@ -85,9 +83,8 @@ OpFoldResult CreateOp::fold(ArrayRef<Attribute> operands) {
// ImOp
//===----------------------------------------------------------------------===//
-OpFoldResult ImOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 1 && "unary op takes 1 operand");
- ArrayAttr arrayAttr = operands[0].dyn_cast_or_null<ArrayAttr>();
+OpFoldResult ImOp::fold(FoldAdaptor adaptor) {
+ ArrayAttr arrayAttr = adaptor.getComplex().dyn_cast_or_null<ArrayAttr>();
if (arrayAttr && arrayAttr.size() == 2)
return arrayAttr[1];
if (auto createOp = getOperand().getDefiningOp<CreateOp>())
@@ -99,9 +96,8 @@ OpFoldResult ImOp::fold(ArrayRef<Attribute> operands) {
// ReOp
//===----------------------------------------------------------------------===//
-OpFoldResult ReOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 1 && "unary op takes 1 operand");
- ArrayAttr arrayAttr = operands[0].dyn_cast_or_null<ArrayAttr>();
+OpFoldResult ReOp::fold(FoldAdaptor adaptor) {
+ ArrayAttr arrayAttr = adaptor.getComplex().dyn_cast_or_null<ArrayAttr>();
if (arrayAttr && arrayAttr.size() == 2)
return arrayAttr[0];
if (auto createOp = getOperand().getDefiningOp<CreateOp>())
@@ -113,9 +109,7 @@ OpFoldResult ReOp::fold(ArrayRef<Attribute> operands) {
// AddOp
//===----------------------------------------------------------------------===//
-OpFoldResult AddOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 2 && "binary op takes 2 operands");
-
+OpFoldResult AddOp::fold(FoldAdaptor adaptor) {
// complex.add(complex.sub(a, b), b) -> a
if (auto sub = getLhs().getDefiningOp<SubOp>())
if (getRhs() == sub.getRhs())
@@ -142,9 +136,7 @@ OpFoldResult AddOp::fold(ArrayRef<Attribute> operands) {
// SubOp
//===----------------------------------------------------------------------===//
-OpFoldResult SubOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 2 && "binary op takes 2 operands");
-
+OpFoldResult SubOp::fold(FoldAdaptor adaptor) {
// complex.sub(complex.add(a, b), b) -> a
if (auto add = getLhs().getDefiningOp<AddOp>())
if (getRhs() == add.getRhs())
@@ -166,9 +158,7 @@ OpFoldResult SubOp::fold(ArrayRef<Attribute> operands) {
// NegOp
//===----------------------------------------------------------------------===//
-OpFoldResult NegOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 1 && "unary op takes 1 operand");
-
+OpFoldResult NegOp::fold(FoldAdaptor adaptor) {
// complex.neg(complex.neg(a)) -> a
if (auto negOp = getOperand().getDefiningOp<NegOp>())
return negOp.getOperand();
@@ -180,9 +170,7 @@ OpFoldResult NegOp::fold(ArrayRef<Attribute> operands) {
// LogOp
//===----------------------------------------------------------------------===//
-OpFoldResult LogOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 1 && "unary op takes 1 operand");
-
+OpFoldResult LogOp::fold(FoldAdaptor adaptor) {
// complex.log(complex.exp(a)) -> a
if (auto expOp = getOperand().getDefiningOp<ExpOp>())
return expOp.getOperand();
@@ -194,9 +182,7 @@ OpFoldResult LogOp::fold(ArrayRef<Attribute> operands) {
// ExpOp
//===----------------------------------------------------------------------===//
-OpFoldResult ExpOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 1 && "unary op takes 1 operand");
-
+OpFoldResult ExpOp::fold(FoldAdaptor adaptor) {
// complex.exp(complex.log(a)) -> a
if (auto logOp = getOperand().getDefiningOp<LogOp>())
return logOp.getOperand();
@@ -208,9 +194,7 @@ OpFoldResult ExpOp::fold(ArrayRef<Attribute> operands) {
// ConjOp
//===----------------------------------------------------------------------===//
-OpFoldResult ConjOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 1 && "unary op takes 1 operand");
-
+OpFoldResult ConjOp::fold(FoldAdaptor adaptor) {
// complex.conj(complex.conj(a)) -> a
if (auto conjOp = getOperand().getDefiningOp<ConjOp>())
return conjOp.getOperand();
diff --git a/mlir/lib/Dialect/EmitC/IR/EmitC.cpp b/mlir/lib/Dialect/EmitC/IR/EmitC.cpp
index 91046c7f70f21..75d92229c287b 100644
--- a/mlir/lib/Dialect/EmitC/IR/EmitC.cpp
+++ b/mlir/lib/Dialect/EmitC/IR/EmitC.cpp
@@ -129,8 +129,7 @@ LogicalResult emitc::ConstantOp::verify() {
return success();
}
-OpFoldResult emitc::ConstantOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.empty() && "constant has no operands");
+OpFoldResult emitc::ConstantOp::fold(FoldAdaptor adaptor) {
return getValue();
}
diff --git a/mlir/lib/Dialect/Func/IR/FuncOps.cpp b/mlir/lib/Dialect/Func/IR/FuncOps.cpp
index 7c26e6ae4fe3a..de96e46c5376f 100644
--- a/mlir/lib/Dialect/Func/IR/FuncOps.cpp
+++ b/mlir/lib/Dialect/Func/IR/FuncOps.cpp
@@ -201,8 +201,7 @@ LogicalResult ConstantOp::verify() {
return success();
}
-OpFoldResult ConstantOp::fold(ArrayRef<Attribute> operands) {
- assert(operands.empty() && "constant has no operands");
+OpFoldResult ConstantOp::fold(FoldAdaptor adaptor) {
return getValueAttr();
}
diff --git a/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp b/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp
index 8ef5483fb7ca9..e55e9a60fa67d 100644
--- a/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp
+++ b/mlir/lib/Dialect/GPU/IR/GPUDialect.cpp
@@ -1286,12 +1286,12 @@ LogicalResult SubgroupMmaComputeOp::verify() {
return success();
}
-LogicalResult MemcpyOp::fold(ArrayRef<Attribute> operands,
+LogicalResult MemcpyOp::fold(FoldAdaptor adaptor,
SmallVectorImpl<::mlir::OpFoldResult> &results) {
return memref::foldMemRefCast(*this);
}
-LogicalResult MemsetOp::fold(ArrayRef<Attribute> operands,
+LogicalResult MemsetOp::fold(FoldAdaptor adaptor,
SmallVectorImpl<::mlir::OpFoldResult> &results) {
return memref::foldMemRefCast(*this);
}
diff --git a/mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp b/mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp
index e00c6887844ef..bb47aeb040662 100644
--- a/mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp
+++ b/mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp
@@ -1441,7 +1441,7 @@ static Type getInsertExtractValueElementType(Type llvmType,
return llvmType;
}
-OpFoldResult LLVM::ExtractValueOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult LLVM::ExtractValueOp::fold(FoldAdaptor adaptor) {
auto insertValueOp = getContainer().getDefiningOp<InsertValueOp>();
OpFoldResult result = {};
while (insertValueOp) {
@@ -2275,7 +2275,7 @@ LogicalResult LLVM::ConstantOp::verify() {
}
// Constant op constant-folds to its value.
-OpFoldResult LLVM::ConstantOp::fold(ArrayRef<Attribute>) { return getValue(); }
+OpFoldResult LLVM::ConstantOp::fold(FoldAdaptor) { return getValue(); }
//===----------------------------------------------------------------------===//
// Utility functions for parsing atomic ops
@@ -2513,7 +2513,7 @@ LogicalResult FenceOp::verify() {
// Folder for LLVM::BitcastOp
//===----------------------------------------------------------------------===//
-OpFoldResult LLVM::BitcastOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult LLVM::BitcastOp::fold(FoldAdaptor adaptor) {
// bitcast(x : T0, T0) -> x
if (getArg().getType() == getType())
return getArg();
@@ -2528,7 +2528,7 @@ OpFoldResult LLVM::BitcastOp::fold(ArrayRef<Attribute> operands) {
// Folder for LLVM::AddrSpaceCastOp
//===----------------------------------------------------------------------===//
-OpFoldResult LLVM::AddrSpaceCastOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult LLVM::AddrSpaceCastOp::fold(FoldAdaptor adaptor) {
// addrcast(x : T0, T0) -> x
if (getArg().getType() == getType())
return getArg();
@@ -2543,9 +2543,9 @@ OpFoldResult LLVM::AddrSpaceCastOp::fold(ArrayRef<Attribute> operands) {
// Folder for LLVM::GEPOp
//===----------------------------------------------------------------------===//
-OpFoldResult LLVM::GEPOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult LLVM::GEPOp::fold(FoldAdaptor adaptor) {
GEPIndicesAdaptor<ArrayRef<Attribute>> indices(getRawConstantIndicesAttr(),
- operands.drop_front());
+ adaptor.getDynamicIndices());
// gep %x:T, 0 -> %x
if (getBase().getType() == getType() && indices.size() == 1)
diff --git a/mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp b/mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
index 308d003f4b9e3..7f5bccecd3707 100644
--- a/mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
+++ b/mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
@@ -980,8 +980,7 @@ void GenericOp::getCanonicalizationPatterns(RewritePatternSet &results,
results.add<EraseIdentityGenericOp>(context);
}
-LogicalResult GenericOp::fold(ArrayRef<Attribute>,
- SmallVectorImpl<OpFoldResult> &) {
+LogicalResult GenericOp::fold(FoldAdaptor, SmallVectorImpl<OpFoldResult> &) {
return memref::foldMemRefCast(*this);
}
diff --git a/mlir/lib/Dialect/Quant/IR/QuantOps.cpp b/mlir/lib/Dialect/Quant/IR/QuantOps.cpp
index fcb97ae603429..c9a6bbc9ceeea 100644
--- a/mlir/lib/Dialect/Quant/IR/QuantOps.cpp
+++ b/mlir/lib/Dialect/Quant/IR/QuantOps.cpp
@@ -36,7 +36,7 @@ void QuantizationDialect::initialize() {
addBytecodeInterface(this);
}
-OpFoldResult StorageCastOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult StorageCastOp::fold(FoldAdaptor adaptor) {
// Matches x -> [scast -> scast] -> y, replacing the second scast with the
// value of x if the casts invert each other.
auto srcScastOp = getArg().getDefiningOp<StorageCastOp>();
diff --git a/mlir/lib/Dialect/SCF/IR/SCF.cpp b/mlir/lib/Dialect/SCF/IR/SCF.cpp
index 25f953c22ac36..b98f9df6ad987 100644
--- a/mlir/lib/Dialect/SCF/IR/SCF.cpp
+++ b/mlir/lib/Dialect/SCF/IR/SCF.cpp
@@ -1598,7 +1598,7 @@ void IfOp::getSuccessorRegions(std::optional<unsigned> index,
regions.push_back(RegionSuccessor(condition ? &getThenRegion() : elseRegion));
}
-LogicalResult IfOp::fold(ArrayRef<Attribute> operands,
+LogicalResult IfOp::fold(FoldAdaptor adaptor,
SmallVectorImpl<OpFoldResult> &results) {
// if (!c) then A() else B() -> if c then B() else A()
if (getElseRegion().empty())
diff --git a/mlir/lib/Dialect/SparseTensor/IR/SparseTensorDialect.cpp b/mlir/lib/Dialect/SparseTensor/IR/SparseTensorDialect.cpp
index 4d2c4fb22b58d..59a4b1360b473 100644
--- a/mlir/lib/Dialect/SparseTensor/IR/SparseTensorDialect.cpp
+++ b/mlir/lib/Dialect/SparseTensor/IR/SparseTensorDialect.cpp
@@ -467,7 +467,7 @@ LogicalResult ConvertOp::verify() {
return emitError("unexpected type in convert");
}
-OpFoldResult ConvertOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult ConvertOp::fold(FoldAdaptor adaptor) {
Type dstType = getType();
// Fold trivial dense-to-dense convert and leave trivial sparse-to-sparse
// convert for codegen to remove. This is because we use trivial
@@ -531,7 +531,7 @@ static SetStorageSpecifierOp getSpecifierSetDef(SpecifierOp op) {
return op.getSpecifier().template getDefiningOp<SetStorageSpecifierOp>();
}
-OpFoldResult GetStorageSpecifierOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult GetStorageSpecifierOp::fold(FoldAdaptor adaptor) {
StorageSpecifierKind kind = getSpecifierKind();
std::optional<APInt> dim = getDim();
for (auto op = getSpecifierSetDef(*this); op; op = getSpecifierSetDef(op))
diff --git a/mlir/lib/Dialect/Transform/IR/TransformOps.cpp b/mlir/lib/Dialect/Transform/IR/TransformOps.cpp
index 20052bd3ac15a..0b3391ea448e2 100644
--- a/mlir/lib/Dialect/Transform/IR/TransformOps.cpp
+++ b/mlir/lib/Dialect/Transform/IR/TransformOps.cpp
@@ -463,7 +463,7 @@ void transform::MergeHandlesOp::getEffects(
// manipulation.
}
-OpFoldResult transform::MergeHandlesOp::fold(ArrayRef<Attribute> operands) {
+OpFoldResult transform::MergeHandlesOp::fold(FoldAdaptor adaptor) {
if (getDeduplicate() || getHandles().size() != 1)
return {};
diff --git a/mlir/lib/IR/BuiltinDialect.cpp b/mlir/lib/IR/BuiltinDialect.cpp
index b66346ffa39c0..0668459d622b1 100644
--- a/mlir/lib/IR/BuiltinDialect.cpp
+++ b/mlir/lib/IR/BuiltinDialect.cpp
@@ -190,7 +190,7 @@ LogicalResult ModuleOp::verify() {
//===----------------------------------------------------------------------===//
LogicalResult
-UnrealizedConversionCastOp::fold(ArrayRef<Attribute> attrOperands,
+UnrealizedConversionCastOp::fold(FoldAdaptor adaptor,
SmallVectorImpl<OpFoldResult> &foldResults) {
OperandRange operands = getInputs();
ResultRange results = getOutputs();
diff --git a/mlir/test/lib/Dialect/Test/TestDialect.cpp b/mlir/test/lib/Dialect/Test/TestDialect.cpp
index 0f48bfcf6b229..7f0ad3b69c5b9 100644
--- a/mlir/test/lib/Dialect/Test/TestDialect.cpp
+++ b/mlir/test/lib/Dialect/Test/TestDialect.cpp
@@ -1099,32 +1099,31 @@ void TestOpWithRegionPattern::getCanonicalizationPatterns(
results.add<TestRemoveOpWithInnerOps>(context);
}
-OpFoldResult TestOpWithRegionFold::fold(ArrayRef<Attribute> operands) {
+OpFoldResult TestOpWithRegionFold::fold(FoldAdaptor adaptor) {
return getOperand();
}
-OpFoldResult TestOpConstant::fold(ArrayRef<Attribute> operands) {
+OpFoldResult TestOpConstant::fold(FoldAdaptor adaptor) {
return getValue();
}
LogicalResult TestOpWithVariadicResultsAndFolder::fold(
- ArrayRef<Attribute> operands, SmallVectorImpl<OpFoldResult> &results) {
+ FoldAdaptor adaptor, SmallVectorImpl<OpFoldResult> &results) {
for (Value input : this->getOperands()) {
results.push_back(input);
}
return success();
}
-OpFoldResult TestOpInPlaceFold::fold(ArrayRef<Attribute> operands) {
- assert(operands.size() == 1);
- if (operands.front()) {
- (*this)->setAttr("attr", operands.front());
+OpFoldResult TestOpInPlaceFold::fold(FoldAdaptor adaptor) {
+ if (adaptor.getOp()) {
+ (*this)->setAttr("attr", adaptor.getOp());
return getResult();
}
return {};
}
-OpFoldResult TestPassthroughFold::fold(ArrayRef<Attribute> operands) {
+OpFoldResult TestPassthroughFold::fold(FoldAdaptor adaptor) {
return getOperand();
}
diff --git a/mlir/test/lib/Dialect/Test/TestDialect.td b/mlir/test/lib/Dialect/Test/TestDialect.td
index 9ec12749bfcc3..9dc4203915392 100644
--- a/mlir/test/lib/Dialect/Test/TestDialect.td
+++ b/mlir/test/lib/Dialect/Test/TestDialect.td
@@ -23,6 +23,7 @@ def Test_Dialect : Dialect {
let hasNonDefaultDestructor = 1;
let useDefaultTypePrinterParser = 0;
let useDefaultAttributePrinterParser = 1;
+ let useFoldAPI = kEmitFoldAdaptorFolder;
let isExtensible = 1;
let dependentDialects = ["::mlir::DLTIDialect"];
diff --git a/mlir/test/lib/Dialect/Test/TestOps.td b/mlir/test/lib/Dialect/Test/TestOps.td
index 7d661368c0e82..e9816ebcc13e8 100644
--- a/mlir/test/lib/Dialect/Test/TestOps.td
+++ b/mlir/test/lib/Dialect/Test/TestOps.td
@@ -1290,7 +1290,7 @@ def TestOpInPlaceFoldSuccess : TEST_Op<"op_in_place_fold_success"> {
let results = (outs Variadic<I1>);
let hasFolder = 1;
let extraClassDefinition = [{
- ::mlir::LogicalResult $cppClass::fold(ArrayRef<Attribute> operands,
+ ::mlir::LogicalResult $cppClass::fold(FoldAdaptor adaptor,
SmallVectorImpl<OpFoldResult> &results) {
return success();
}
@@ -1315,11 +1315,7 @@ def TestOpFoldWithFoldAdaptor
$op `,` `[` $variadic `]` `,` `{` $var_of_var `}` $body attr-dict-with-keyword
}];
- let hasFolder = 0;
-
- let extraClassDeclaration = [{
- ::mlir::OpFoldResult fold(FoldAdaptor adaptor);
- }];
+ let hasFolder = 1;
}
// An op that always fold itself.
diff --git a/mlir/test/lib/Dialect/Test/TestTraits.cpp b/mlir/test/lib/Dialect/Test/TestTraits.cpp
index 0ccffc7a47a02..d9b67ef95ace8 100644
--- a/mlir/test/lib/Dialect/Test/TestTraits.cpp
+++ b/mlir/test/lib/Dialect/Test/TestTraits.cpp
@@ -18,13 +18,13 @@ using namespace test;
//===----------------------------------------------------------------------===//
OpFoldResult TestInvolutionTraitFailingOperationFolderOp::fold(
- ArrayRef<Attribute> operands) {
+ FoldAdaptor adaptor) {
// This failure should cause the trait fold to run instead.
return {};
}
OpFoldResult TestInvolutionTraitSuccesfulOperationFolderOp::fold(
- ArrayRef<Attribute> operands) {
+ FoldAdaptor adaptor) {
auto argumentOp = getOperand();
// The success case should cause the trait fold to be supressed.
return argumentOp.getDefiningOp() ? argumentOp : OpFoldResult{};
diff --git a/mlir/tools/mlir-linalg-ods-gen/mlir-linalg-ods-yaml-gen.cpp b/mlir/tools/mlir-linalg-ods-gen/mlir-linalg-ods-yaml-gen.cpp
index 9b7816c952278..d92c36795b2bd 100644
--- a/mlir/tools/mlir-linalg-ods-gen/mlir-linalg-ods-yaml-gen.cpp
+++ b/mlir/tools/mlir-linalg-ods-gen/mlir-linalg-ods-yaml-gen.cpp
@@ -654,7 +654,7 @@ ArrayAttr {0}::getIndexingMaps() {{
// Parameters:
// {0}: Class name
const char structuredOpFoldersFormat[] = R"FMT(
-LogicalResult {0}::fold(ArrayRef<Attribute>,
+LogicalResult {0}::fold(FoldAdaptor,
SmallVectorImpl<OpFoldResult> &) {{
return memref::foldMemRefCast(*this);
}
More information about the Mlir-commits
mailing list