Files
llvm/mlir/test/lib/Dialect/Test/TestPatterns.cpp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

771 lines
30 KiB
C++
Raw Normal View History

//===- TestPatterns.cpp - Test dialect pattern driver ---------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
#include "TestDialect.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/StandardOps/Transforms/FuncConversions.h"
#include "mlir/IR/PatternMatch.h"
#include "mlir/Pass/Pass.h"
#include "mlir/Transforms/DialectConversion.h"
#include "mlir/Transforms/FoldUtils.h"
using namespace mlir;
// Native function for testing NativeCodeCall
static Value chooseOperand(Value input1, Value input2, BoolAttr choice) {
return choice.getValue() ? input1 : input2;
}
static void createOpI(PatternRewriter &rewriter, Value input) {
rewriter.create<OpI>(rewriter.getUnknownLoc(), input);
}
static void handleNoResultOp(PatternRewriter &rewriter,
OpSymbolBindingNoResult op) {
// Turn the no result op to a one-result op.
rewriter.create<OpSymbolBindingB>(op.getLoc(), op.operand().getType(),
op.operand());
}
// Test that natives calls are only called once during rewrites.
// OpM_Test will return Pi, increased by 1 for each subsequent calls.
// This let us check the number of times OpM_Test was called by inspecting
// the returned value in the MLIR output.
static int64_t opMIncreasingValue = 314159265;
static Attribute OpMTest(PatternRewriter &rewriter, Value val) {
int64_t i = opMIncreasingValue++;
return rewriter.getIntegerAttr(rewriter.getIntegerType(32), i);
}
namespace {
#include "TestPatterns.inc"
} // end anonymous namespace
//===----------------------------------------------------------------------===//
// Canonicalizer Driver.
//===----------------------------------------------------------------------===//
namespace {
struct FoldingPattern : public RewritePattern {
public:
FoldingPattern(MLIRContext *context)
: RewritePattern(TestOpInPlaceFoldAnchor::getOperationName(),
/*benefit=*/1, context) {}
LogicalResult matchAndRewrite(Operation *op,
PatternRewriter &rewriter) const override {
// Exercice OperationFolder API for a single-result operation that is folded
// upon construction. The operation being created through the folder has an
// in-place folder, and it should be still present in the output.
// Furthermore, the folder should not crash when attempting to recover the
// (unchanged) opeation result.
OperationFolder folder(op->getContext());
Value result = folder.create<TestOpInPlaceFold>(
rewriter, op->getLoc(), rewriter.getIntegerType(32), op->getOperand(0),
rewriter.getI32IntegerAttr(0));
assert(result);
rewriter.replaceOp(op, result);
return success();
}
};
struct TestPatternDriver : public PassWrapper<TestPatternDriver, FunctionPass> {
void runOnFunction() override {
mlir::OwningRewritePatternList patterns;
populateWithGenerated(&getContext(), &patterns);
// Verify named pattern is generated with expected name.
patterns.insert<FoldingPattern, TestNamedPatternRule>(&getContext());
applyPatternsAndFoldGreedily(getFunction(), patterns);
}
};
} // end anonymous namespace
//===----------------------------------------------------------------------===//
// ReturnType Driver.
//===----------------------------------------------------------------------===//
namespace {
2020-02-28 10:59:34 -08:00
// Generate ops for each instance where the type can be successfully inferred.
template <typename OpTy>
2020-02-28 10:59:34 -08:00
static void invokeCreateWithInferredReturnType(Operation *op) {
auto *context = op->getContext();
auto fop = op->getParentOfType<FuncOp>();
auto location = UnknownLoc::get(context);
OpBuilder b(op);
b.setInsertionPointAfter(op);
// Use permutations of 2 args as operands.
assert(fop.getNumArguments() >= 2);
for (int i = 0, e = fop.getNumArguments(); i < e; ++i) {
for (int j = 0; j < e; ++j) {
std::array<Value, 2> values = {{fop.getArgument(i), fop.getArgument(j)}};
2020-02-28 10:59:34 -08:00
SmallVector<Type, 2> inferredReturnTypes;
if (succeeded(OpTy::inferReturnTypes(
context, llvm::None, values, op->getAttrDictionary(),
op->getRegions(), inferredReturnTypes))) {
OperationState state(location, OpTy::getOperationName());
// TODO: Expand to regions.
OpTy::build(b, state, values, op->getAttrs());
(void)b.createOperation(state);
}
}
}
}
static void reifyReturnShape(Operation *op) {
OpBuilder b(op);
// Use permutations of 2 args as operands.
auto shapedOp = cast<OpWithShapedTypeInferTypeInterfaceOp>(op);
SmallVector<Value, 2> shapes;
if (failed(shapedOp.reifyReturnTypeShapes(b, shapes)))
return;
for (auto it : llvm::enumerate(shapes))
op->emitRemark() << "value " << it.index() << ": "
<< it.value().getDefiningOp();
}
struct TestReturnTypeDriver
: public PassWrapper<TestReturnTypeDriver, FunctionPass> {
void runOnFunction() override {
if (getFunction().getName() == "testCreateFunctions") {
std::vector<Operation *> ops;
// Collect ops to avoid triggering on inserted ops.
for (auto &op : getFunction().getBody().front())
ops.push_back(&op);
// Generate test patterns for each, but skip terminator.
for (auto *op : llvm::makeArrayRef(ops).drop_back()) {
// Test create method of each of the Op classes below. The resultant
// output would be in reverse order underneath `op` from which
// the attributes and regions are used.
2020-02-28 10:59:34 -08:00
invokeCreateWithInferredReturnType<OpWithInferTypeInterfaceOp>(op);
invokeCreateWithInferredReturnType<
OpWithShapedTypeInferTypeInterfaceOp>(op);
};
return;
}
if (getFunction().getName() == "testReifyFunctions") {
std::vector<Operation *> ops;
// Collect ops to avoid triggering on inserted ops.
for (auto &op : getFunction().getBody().front())
if (isa<OpWithShapedTypeInferTypeInterfaceOp>(op))
ops.push_back(&op);
// Generate test patterns for each, but skip terminator.
for (auto *op : ops)
reifyReturnShape(op);
}
}
};
} // end anonymous namespace
namespace {
struct TestDerivedAttributeDriver
: public PassWrapper<TestDerivedAttributeDriver, FunctionPass> {
void runOnFunction() override;
};
} // end anonymous namespace
void TestDerivedAttributeDriver::runOnFunction() {
getFunction().walk([](DerivedAttributeOpInterface dOp) {
auto dAttr = dOp.materializeDerivedAttributes();
if (!dAttr)
return;
for (auto d : dAttr)
dOp.emitRemark() << d.first << " = " << d.second;
});
}
//===----------------------------------------------------------------------===//
// Legalization Driver.
//===----------------------------------------------------------------------===//
namespace {
//===----------------------------------------------------------------------===//
// Region-Block Rewrite Testing
/// This pattern is a simple pattern that inlines the first region of a given
/// operation into the parent region.
struct TestRegionRewriteBlockMovement : public ConversionPattern {
TestRegionRewriteBlockMovement(MLIRContext *ctx)
: ConversionPattern("test.region", 1, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
// Inline this region into the parent region.
auto &parentRegion = *op->getParentRegion();
if (op->getAttr("legalizer.should_clone"))
rewriter.cloneRegionBefore(op->getRegion(0), parentRegion,
parentRegion.end());
else
rewriter.inlineRegionBefore(op->getRegion(0), parentRegion,
parentRegion.end());
// Drop this operation.
rewriter.eraseOp(op);
return success();
}
};
/// This pattern is a simple pattern that generates a region containing an
/// illegal operation.
struct TestRegionRewriteUndo : public RewritePattern {
TestRegionRewriteUndo(MLIRContext *ctx)
: RewritePattern("test.region_builder", 1, ctx) {}
LogicalResult matchAndRewrite(Operation *op,
PatternRewriter &rewriter) const final {
// Create the region operation with an entry block containing arguments.
OperationState newRegion(op->getLoc(), "test.region");
newRegion.addRegion();
auto *regionOp = rewriter.createOperation(newRegion);
auto *entryBlock = rewriter.createBlock(&regionOp->getRegion(0));
entryBlock->addArgument(rewriter.getIntegerType(64));
// Add an explicitly illegal operation to ensure the conversion fails.
rewriter.create<ILLegalOpF>(op->getLoc(), rewriter.getIntegerType(32));
rewriter.create<TestValidOp>(op->getLoc(), ArrayRef<Value>());
// Drop this operation.
rewriter.eraseOp(op);
return success();
}
};
[mlir] DialectConversion: support block creation in ConversionPatternRewriter PatternRewriter and derived classes provide a set of virtual methods to manipulate blocks, which ConversionPatternRewriter overrides to keep track of the manipulations and undo them in case the conversion fails. However, one can currently create a block only by splitting another block into two. This not only makes the API inconsistent (`splitBlock` is allowed in conversion patterns, but `createBlock` is not), but it also make it impossible for one to create blocks with argument lists different from those of already existing blocks since in-place block updates are not supported either. Such functionality precludes dialect conversion infrastructure from being used more extensively on region-containing ops, for example, for value-returning "if" operations. At the same time, ConversionPatternRewriter already allows one to undo block creation as block creation is one of the primitive operations in already supported region inlining. Support block creation in conversion patterns by hooking `createBlock` on the block action undo mechanism. This requires to make `Builder::createBlock` virtual, similarly to Op insertion. This is a minimal change to the Builder infrastructure that will later help support additional use cases such as block signature changes. `createBlock` now additionally takes the types of the block arguments that are added immediately so as to avoid in-place argument list manipulation that would be illegal in conversion patterns.
2020-04-03 19:53:13 +02:00
/// A simple pattern that creates a block at the end of the parent region of the
/// matched operation.
struct TestCreateBlock : public RewritePattern {
TestCreateBlock(MLIRContext *ctx)
: RewritePattern("test.create_block", /*benefit=*/1, ctx) {}
LogicalResult matchAndRewrite(Operation *op,
PatternRewriter &rewriter) const final {
Region &region = *op->getParentRegion();
Type i32Type = rewriter.getIntegerType(32);
rewriter.createBlock(&region, region.end(), {i32Type, i32Type});
rewriter.create<TerminatorOp>(op->getLoc());
rewriter.replaceOp(op, {});
return success();
}
};
/// A simple pattern that creates a block containing an invalid operaiton in
/// order to trigger the block creation undo mechanism.
struct TestCreateIllegalBlock : public RewritePattern {
TestCreateIllegalBlock(MLIRContext *ctx)
: RewritePattern("test.create_illegal_block", /*benefit=*/1, ctx) {}
LogicalResult matchAndRewrite(Operation *op,
PatternRewriter &rewriter) const final {
Region &region = *op->getParentRegion();
Type i32Type = rewriter.getIntegerType(32);
rewriter.createBlock(&region, region.end(), {i32Type, i32Type});
// Create an illegal op to ensure the conversion fails.
rewriter.create<ILLegalOpF>(op->getLoc(), i32Type);
rewriter.create<TerminatorOp>(op->getLoc());
rewriter.replaceOp(op, {});
return success();
}
};
/// A simple pattern that tests the undo mechanism when replacing the uses of a
/// block argument.
struct TestUndoBlockArgReplace : public ConversionPattern {
TestUndoBlockArgReplace(MLIRContext *ctx)
: ConversionPattern("test.undo_block_arg_replace", /*benefit=*/1, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
auto illegalOp =
rewriter.create<ILLegalOpF>(op->getLoc(), rewriter.getF32Type());
rewriter.replaceUsesOfBlockArgument(op->getRegion(0).getArgument(0),
illegalOp);
rewriter.updateRootInPlace(op, [] {});
return success();
}
};
/// A rewrite pattern that tests the undo mechanism when erasing a block.
struct TestUndoBlockErase : public ConversionPattern {
TestUndoBlockErase(MLIRContext *ctx)
: ConversionPattern("test.undo_block_erase", /*benefit=*/1, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
Block *secondBlock = &*std::next(op->getRegion(0).begin());
rewriter.setInsertionPointToStart(secondBlock);
rewriter.create<ILLegalOpF>(op->getLoc(), rewriter.getF32Type());
rewriter.eraseBlock(secondBlock);
rewriter.updateRootInPlace(op, [] {});
return success();
}
};
//===----------------------------------------------------------------------===//
// Type-Conversion Rewrite Testing
/// This patterns erases a region operation that has had a type conversion.
struct TestDropOpSignatureConversion : public ConversionPattern {
TestDropOpSignatureConversion(MLIRContext *ctx, TypeConverter &converter)
: ConversionPattern("test.drop_region_op", 1, converter, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
Region &region = op->getRegion(0);
Block *entry = &region.front();
// Convert the original entry arguments.
TypeConverter &converter = *getTypeConverter();
TypeConverter::SignatureConversion result(entry->getNumArguments());
if (failed(converter.convertSignatureArgs(entry->getArgumentTypes(),
result)) ||
failed(rewriter.convertRegionTypes(&region, converter, &result)))
return failure();
// Convert the region signature and just drop the operation.
rewriter.eraseOp(op);
return success();
}
};
/// This pattern simply updates the operands of the given operation.
struct TestPassthroughInvalidOp : public ConversionPattern {
TestPassthroughInvalidOp(MLIRContext *ctx)
: ConversionPattern("test.invalid", 1, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
rewriter.replaceOpWithNewOp<TestValidOp>(op, llvm::None, operands,
llvm::None);
return success();
}
};
/// This pattern handles the case of a split return value.
struct TestSplitReturnType : public ConversionPattern {
TestSplitReturnType(MLIRContext *ctx)
: ConversionPattern("test.return", 1, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
// Check for a return of F32.
if (op->getNumOperands() != 1 || !op->getOperand(0).getType().isF32())
return failure();
// Check if the first operation is a cast operation, if it is we use the
// results directly.
auto *defOp = operands[0].getDefiningOp();
if (auto packerOp = llvm::dyn_cast_or_null<TestCastOp>(defOp)) {
rewriter.replaceOpWithNewOp<TestReturnOp>(op, packerOp.getOperands());
return success();
}
// Otherwise, fail to match.
return failure();
}
};
//===----------------------------------------------------------------------===//
// Multi-Level Type-Conversion Rewrite Testing
struct TestChangeProducerTypeI32ToF32 : public ConversionPattern {
TestChangeProducerTypeI32ToF32(MLIRContext *ctx)
: ConversionPattern("test.type_producer", 1, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
// If the type is I32, change the type to F32.
[mlir] Add a signedness semantics bit to IntegerType Thus far IntegerType has been signless: a value of IntegerType does not have a sign intrinsically and it's up to the specific operation to decide how to interpret those bits. For example, std.addi does two's complement arithmetic, and std.divis/std.diviu treats the first bit as a sign. This design choice was made some time ago when we did't have lots of dialects and dialects were more rigid. Today we have much more extensible infrastructure and different dialect may want different modelling over integer signedness. So while we can say we want signless integers in the standard dialect, we cannot dictate for others. Requiring each dialect to model the signedness semantics with another set of custom types is duplicating the functionality everywhere, considering the fundamental role integer types play. This CL extends the IntegerType with a signedness semantics bit. This gives each dialect an option to opt in signedness semantics if that's what they want and helps code sharing. The parser is modified to recognize `si[1-9][0-9]*` and `ui[1-9][0-9]*` as signed and unsigned integer types, respectively, leaving the original `i[1-9][0-9]*` to continue to mean no indication over signedness semantics. All existing dialects are not affected (yet) as this is a feature to opt in. More discussions can be found at: https://groups.google.com/a/tensorflow.org/d/msg/mlir/XmkV8HOPWpo/7O4X0Nb_AQAJ Differential Revision: https://reviews.llvm.org/D72533
2020-01-10 14:48:24 -05:00
if (!Type(*op->result_type_begin()).isSignlessInteger(32))
return failure();
rewriter.replaceOpWithNewOp<TestTypeProducerOp>(op, rewriter.getF32Type());
return success();
}
};
struct TestChangeProducerTypeF32ToF64 : public ConversionPattern {
TestChangeProducerTypeF32ToF64(MLIRContext *ctx)
: ConversionPattern("test.type_producer", 1, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
// If the type is F32, change the type to F64.
if (!Type(*op->result_type_begin()).isF32())
return rewriter.notifyMatchFailure(op, "expected single f32 operand");
rewriter.replaceOpWithNewOp<TestTypeProducerOp>(op, rewriter.getF64Type());
return success();
}
};
struct TestChangeProducerTypeF32ToInvalid : public ConversionPattern {
TestChangeProducerTypeF32ToInvalid(MLIRContext *ctx)
: ConversionPattern("test.type_producer", 10, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
// Always convert to B16, even though it is not a legal type. This tests
// that values are unmapped correctly.
rewriter.replaceOpWithNewOp<TestTypeProducerOp>(op, rewriter.getBF16Type());
return success();
}
};
struct TestUpdateConsumerType : public ConversionPattern {
TestUpdateConsumerType(MLIRContext *ctx)
: ConversionPattern("test.type_consumer", 1, ctx) {}
LogicalResult
matchAndRewrite(Operation *op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
// Verify that the incoming operand has been successfully remapped to F64.
if (!operands[0].getType().isF64())
return failure();
rewriter.replaceOpWithNewOp<TestTypeConsumerOp>(op, operands[0]);
return success();
}
};
//===----------------------------------------------------------------------===//
// Non-Root Replacement Rewrite Testing
/// This pattern generates an invalid operation, but replaces it before the
/// pattern is finished. This checks that we don't need to legalize the
/// temporary op.
struct TestNonRootReplacement : public RewritePattern {
TestNonRootReplacement(MLIRContext *ctx)
: RewritePattern("test.replace_non_root", 1, ctx) {}
LogicalResult matchAndRewrite(Operation *op,
PatternRewriter &rewriter) const final {
auto resultType = *op->result_type_begin();
auto illegalOp = rewriter.create<ILLegalOpF>(op->getLoc(), resultType);
auto legalOp = rewriter.create<LegalOpB>(op->getLoc(), resultType);
rewriter.replaceOp(illegalOp, {legalOp});
rewriter.replaceOp(op, {illegalOp});
return success();
}
};
//===----------------------------------------------------------------------===//
// Recursive Rewrite Testing
/// This pattern is applied to the same operation multiple times, but has a
/// bounded recursion.
struct TestBoundedRecursiveRewrite
: public OpRewritePattern<TestRecursiveRewriteOp> {
using OpRewritePattern<TestRecursiveRewriteOp>::OpRewritePattern;
LogicalResult matchAndRewrite(TestRecursiveRewriteOp op,
PatternRewriter &rewriter) const final {
// Decrement the depth of the op in-place.
rewriter.updateRootInPlace(op, [&] {
op.setAttr("depth",
rewriter.getI64IntegerAttr(op.depth().getSExtValue() - 1));
});
return success();
}
/// The conversion target handles bounding the recursion of this pattern.
bool hasBoundedRewriteRecursion() const final { return true; }
};
struct TestNestedOpCreationUndoRewrite
: public OpRewritePattern<IllegalOpWithRegionAnchor> {
using OpRewritePattern<IllegalOpWithRegionAnchor>::OpRewritePattern;
LogicalResult matchAndRewrite(IllegalOpWithRegionAnchor op,
PatternRewriter &rewriter) const final {
// rewriter.replaceOpWithNewOp<IllegalOpWithRegion>(op);
rewriter.replaceOpWithNewOp<IllegalOpWithRegion>(op);
return success();
};
};
} // namespace
namespace {
struct TestTypeConverter : public TypeConverter {
using TypeConverter::TypeConverter;
TestTypeConverter() {
addConversion(convertType);
addMaterialization(materializeCast);
addMaterialization(materializeOneToOneCast);
}
static LogicalResult convertType(Type t, SmallVectorImpl<Type> &results) {
// Drop I16 types.
[mlir] Add a signedness semantics bit to IntegerType Thus far IntegerType has been signless: a value of IntegerType does not have a sign intrinsically and it's up to the specific operation to decide how to interpret those bits. For example, std.addi does two's complement arithmetic, and std.divis/std.diviu treats the first bit as a sign. This design choice was made some time ago when we did't have lots of dialects and dialects were more rigid. Today we have much more extensible infrastructure and different dialect may want different modelling over integer signedness. So while we can say we want signless integers in the standard dialect, we cannot dictate for others. Requiring each dialect to model the signedness semantics with another set of custom types is duplicating the functionality everywhere, considering the fundamental role integer types play. This CL extends the IntegerType with a signedness semantics bit. This gives each dialect an option to opt in signedness semantics if that's what they want and helps code sharing. The parser is modified to recognize `si[1-9][0-9]*` and `ui[1-9][0-9]*` as signed and unsigned integer types, respectively, leaving the original `i[1-9][0-9]*` to continue to mean no indication over signedness semantics. All existing dialects are not affected (yet) as this is a feature to opt in. More discussions can be found at: https://groups.google.com/a/tensorflow.org/d/msg/mlir/XmkV8HOPWpo/7O4X0Nb_AQAJ Differential Revision: https://reviews.llvm.org/D72533
2020-01-10 14:48:24 -05:00
if (t.isSignlessInteger(16))
return success();
// Convert I64 to F64.
[mlir] Add a signedness semantics bit to IntegerType Thus far IntegerType has been signless: a value of IntegerType does not have a sign intrinsically and it's up to the specific operation to decide how to interpret those bits. For example, std.addi does two's complement arithmetic, and std.divis/std.diviu treats the first bit as a sign. This design choice was made some time ago when we did't have lots of dialects and dialects were more rigid. Today we have much more extensible infrastructure and different dialect may want different modelling over integer signedness. So while we can say we want signless integers in the standard dialect, we cannot dictate for others. Requiring each dialect to model the signedness semantics with another set of custom types is duplicating the functionality everywhere, considering the fundamental role integer types play. This CL extends the IntegerType with a signedness semantics bit. This gives each dialect an option to opt in signedness semantics if that's what they want and helps code sharing. The parser is modified to recognize `si[1-9][0-9]*` and `ui[1-9][0-9]*` as signed and unsigned integer types, respectively, leaving the original `i[1-9][0-9]*` to continue to mean no indication over signedness semantics. All existing dialects are not affected (yet) as this is a feature to opt in. More discussions can be found at: https://groups.google.com/a/tensorflow.org/d/msg/mlir/XmkV8HOPWpo/7O4X0Nb_AQAJ Differential Revision: https://reviews.llvm.org/D72533
2020-01-10 14:48:24 -05:00
if (t.isSignlessInteger(64)) {
results.push_back(FloatType::getF64(t.getContext()));
return success();
}
// Convert I42 to I43.
if (t.isInteger(42)) {
results.push_back(IntegerType::get(43, t.getContext()));
return success();
}
// Split F32 into F16,F16.
if (t.isF32()) {
results.assign(2, FloatType::getF16(t.getContext()));
return success();
}
// Otherwise, convert the type directly.
results.push_back(t);
return success();
}
/// Hook for materializing a conversion. This is necessary because we generate
/// 1->N type mappings.
static Optional<Value> materializeCast(PatternRewriter &rewriter,
Type resultType, ValueRange inputs,
Location loc) {
if (inputs.size() == 1)
return inputs[0];
return rewriter.create<TestCastOp>(loc, resultType, inputs).getResult();
}
/// Materialize the cast for one-to-one conversion from i64 to f64.
static Optional<Value> materializeOneToOneCast(PatternRewriter &rewriter,
IntegerType resultType,
ValueRange inputs,
Location loc) {
if (resultType.getWidth() == 42 && inputs.size() == 1)
return rewriter.create<TestCastOp>(loc, resultType, inputs).getResult();
return llvm::None;
}
};
struct TestLegalizePatternDriver
: public PassWrapper<TestLegalizePatternDriver, OperationPass<ModuleOp>> {
/// The mode of conversion to use with the driver.
enum class ConversionMode { Analysis, Full, Partial };
TestLegalizePatternDriver(ConversionMode mode) : mode(mode) {}
void runOnOperation() override {
TestTypeConverter converter;
mlir::OwningRewritePatternList patterns;
populateWithGenerated(&getContext(), &patterns);
patterns.insert<
TestRegionRewriteBlockMovement, TestRegionRewriteUndo, TestCreateBlock,
TestCreateIllegalBlock, TestUndoBlockArgReplace, TestUndoBlockErase,
TestPassthroughInvalidOp, TestSplitReturnType,
TestChangeProducerTypeI32ToF32, TestChangeProducerTypeF32ToF64,
TestChangeProducerTypeF32ToInvalid, TestUpdateConsumerType,
TestNonRootReplacement, TestBoundedRecursiveRewrite,
TestNestedOpCreationUndoRewrite>(&getContext());
patterns.insert<TestDropOpSignatureConversion>(&getContext(), converter);
mlir::populateFuncOpTypeConversionPattern(patterns, &getContext(),
converter);
mlir::populateCallOpTypeConversionPattern(patterns, &getContext(),
converter);
// Define the conversion target used for the test.
ConversionTarget target(getContext());
target.addLegalOp<ModuleOp, ModuleTerminatorOp>();
[mlir] DialectConversion: support block creation in ConversionPatternRewriter PatternRewriter and derived classes provide a set of virtual methods to manipulate blocks, which ConversionPatternRewriter overrides to keep track of the manipulations and undo them in case the conversion fails. However, one can currently create a block only by splitting another block into two. This not only makes the API inconsistent (`splitBlock` is allowed in conversion patterns, but `createBlock` is not), but it also make it impossible for one to create blocks with argument lists different from those of already existing blocks since in-place block updates are not supported either. Such functionality precludes dialect conversion infrastructure from being used more extensively on region-containing ops, for example, for value-returning "if" operations. At the same time, ConversionPatternRewriter already allows one to undo block creation as block creation is one of the primitive operations in already supported region inlining. Support block creation in conversion patterns by hooking `createBlock` on the block action undo mechanism. This requires to make `Builder::createBlock` virtual, similarly to Op insertion. This is a minimal change to the Builder infrastructure that will later help support additional use cases such as block signature changes. `createBlock` now additionally takes the types of the block arguments that are added immediately so as to avoid in-place argument list manipulation that would be illegal in conversion patterns.
2020-04-03 19:53:13 +02:00
target.addLegalOp<LegalOpA, LegalOpB, TestCastOp, TestValidOp,
TerminatorOp>();
target
.addIllegalOp<ILLegalOpF, TestRegionBuilderOp, TestOpWithRegionFold>();
target.addDynamicallyLegalOp<TestReturnOp>([](TestReturnOp op) {
// Don't allow F32 operands.
return llvm::none_of(op.getOperandTypes(),
[](Type type) { return type.isF32(); });
});
target.addDynamicallyLegalOp<FuncOp>([&](FuncOp op) {
return converter.isSignatureLegal(op.getType()) &&
converter.isLegal(&op.getBody());
});
// Expect the type_producer/type_consumer operations to only operate on f64.
target.addDynamicallyLegalOp<TestTypeProducerOp>(
[](TestTypeProducerOp op) { return op.getType().isF64(); });
target.addDynamicallyLegalOp<TestTypeConsumerOp>([](TestTypeConsumerOp op) {
return op.getOperand().getType().isF64();
});
// Check support for marking certain operations as recursively legal.
target.markOpRecursivelyLegal<FuncOp, ModuleOp>([](Operation *op) {
return static_cast<bool>(
op->getAttrOfType<UnitAttr>("test.recursively_legal"));
});
// Mark the bound recursion operation as dynamically legal.
target.addDynamicallyLegalOp<TestRecursiveRewriteOp>(
[](TestRecursiveRewriteOp op) { return op.depth() == 0; });
// Handle a partial conversion.
if (mode == ConversionMode::Partial) {
DenseSet<Operation *> unlegalizedOps;
(void)applyPartialConversion(getOperation(), target, patterns,
&unlegalizedOps);
// Emit remarks for each legalizable operation.
for (auto *op : unlegalizedOps)
op->emitRemark() << "op '" << op->getName() << "' is not legalizable";
return;
}
// Handle a full conversion.
if (mode == ConversionMode::Full) {
// Check support for marking unknown operations as dynamically legal.
target.markUnknownOpDynamicallyLegal([](Operation *op) {
return (bool)op->getAttrOfType<UnitAttr>("test.dynamically_legal");
});
(void)applyFullConversion(getOperation(), target, patterns);
return;
}
// Otherwise, handle an analysis conversion.
assert(mode == ConversionMode::Analysis);
// Analyze the convertible operations.
DenseSet<Operation *> legalizedOps;
if (failed(applyAnalysisConversion(getOperation(), target, patterns,
legalizedOps)))
return signalPassFailure();
// Emit remarks for each legalizable operation.
for (auto *op : legalizedOps)
op->emitRemark() << "op '" << op->getName() << "' is legalizable";
}
/// The mode of conversion to use.
ConversionMode mode;
};
} // end anonymous namespace
static llvm::cl::opt<TestLegalizePatternDriver::ConversionMode>
legalizerConversionMode(
"test-legalize-mode",
llvm::cl::desc("The legalization mode to use with the test driver"),
llvm::cl::init(TestLegalizePatternDriver::ConversionMode::Partial),
llvm::cl::values(
clEnumValN(TestLegalizePatternDriver::ConversionMode::Analysis,
"analysis", "Perform an analysis conversion"),
clEnumValN(TestLegalizePatternDriver::ConversionMode::Full, "full",
"Perform a full conversion"),
clEnumValN(TestLegalizePatternDriver::ConversionMode::Partial,
"partial", "Perform a partial conversion")));
//===----------------------------------------------------------------------===//
// ConversionPatternRewriter::getRemappedValue testing. This method is used
// to get the remapped value of an original value that was replaced using
// ConversionPatternRewriter.
namespace {
/// Converter that replaces a one-result one-operand OneVResOneVOperandOp1 with
/// a one-operand two-result OneVResOneVOperandOp1 by replicating its original
/// operand twice.
///
/// Example:
/// %1 = test.one_variadic_out_one_variadic_in1"(%0)
/// is replaced with:
/// %1 = test.one_variadic_out_one_variadic_in1"(%0, %0)
struct OneVResOneVOperandOp1Converter
: public OpConversionPattern<OneVResOneVOperandOp1> {
using OpConversionPattern<OneVResOneVOperandOp1>::OpConversionPattern;
LogicalResult
matchAndRewrite(OneVResOneVOperandOp1 op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
auto origOps = op.getOperands();
assert(std::distance(origOps.begin(), origOps.end()) == 1 &&
"One operand expected");
Value origOp = *origOps.begin();
SmallVector<Value, 2> remappedOperands;
// Replicate the remapped original operand twice. Note that we don't used
// the remapped 'operand' since the goal is testing 'getRemappedValue'.
remappedOperands.push_back(rewriter.getRemappedValue(origOp));
remappedOperands.push_back(rewriter.getRemappedValue(origOp));
rewriter.replaceOpWithNewOp<OneVResOneVOperandOp1>(op, op.getResultTypes(),
remappedOperands);
return success();
}
};
struct TestRemappedValue
: public mlir::PassWrapper<TestRemappedValue, FunctionPass> {
void runOnFunction() override {
mlir::OwningRewritePatternList patterns;
patterns.insert<OneVResOneVOperandOp1Converter>(&getContext());
mlir::ConversionTarget target(getContext());
target.addLegalOp<ModuleOp, ModuleTerminatorOp, FuncOp, TestReturnOp>();
// We make OneVResOneVOperandOp1 legal only when it has more that one
// operand. This will trigger the conversion that will replace one-operand
// OneVResOneVOperandOp1 with two-operand OneVResOneVOperandOp1.
target.addDynamicallyLegalOp<OneVResOneVOperandOp1>(
[](Operation *op) -> bool {
return std::distance(op->operand_begin(), op->operand_end()) > 1;
});
if (failed(mlir::applyFullConversion(getFunction(), target, patterns))) {
signalPassFailure();
}
}
};
} // end anonymous namespace
//===----------------------------------------------------------------------===//
// Test patterns without a specific root operation kind
//===----------------------------------------------------------------------===//
namespace {
/// This pattern matches and removes any operation in the test dialect.
struct RemoveTestDialectOps : public RewritePattern {
RemoveTestDialectOps() : RewritePattern(/*benefit=*/1, MatchAnyOpTypeTag()) {}
LogicalResult matchAndRewrite(Operation *op,
PatternRewriter &rewriter) const override {
if (!isa<TestDialect>(op->getDialect()))
return failure();
rewriter.eraseOp(op);
return success();
}
};
struct TestUnknownRootOpDriver
: public mlir::PassWrapper<TestUnknownRootOpDriver, FunctionPass> {
void runOnFunction() override {
mlir::OwningRewritePatternList patterns;
patterns.insert<RemoveTestDialectOps>();
mlir::ConversionTarget target(getContext());
target.addIllegalDialect<TestDialect>();
if (failed(applyPartialConversion(getFunction(), target, patterns)))
signalPassFailure();
}
};
} // end anonymous namespace
namespace mlir {
void registerPatternsTestPass() {
PassRegistration<TestReturnTypeDriver>("test-return-type",
"Run return type functions");
PassRegistration<TestDerivedAttributeDriver>("test-derived-attr",
"Run test derived attributes");
PassRegistration<TestPatternDriver>("test-patterns",
"Run test dialect patterns");
PassRegistration<TestLegalizePatternDriver>(
"test-legalize-patterns", "Run test dialect legalization patterns", [] {
return std::make_unique<TestLegalizePatternDriver>(
legalizerConversionMode);
});
PassRegistration<TestRemappedValue>(
"test-remapped-value",
"Test public remapped value mechanism in ConversionPatternRewriter");
PassRegistration<TestUnknownRootOpDriver>(
"test-legalize-unknown-root-patterns",
"Test public remapped value mechanism in ConversionPatternRewriter");
}
} // namespace mlir