The MLIR classes Type/Attribute/Operation/Op/Value support cast/dyn_cast/isa/dyn_cast_or_null functionality through llvm's doCast functionality in addition to defining methods with the same name. This change begins the migration of uses of the method to the corresponding function call as has been decided as more consistent. Note that there still exist classes that only define methods directly, such as AffineExpr, and this does not include work currently to support a functional cast/isa call. Caveats include: - This clang-tidy script probably has more problems. - This only touches C++ code, so nothing that is being generated. Context: - https://mlir.llvm.org/deprecation/ at "Use the free function variants for dyn_cast/cast/isa/…" - Original discussion at https://discourse.llvm.org/t/preferred-casting-style-going-forward/68443 Implementation: This first patch was created with the following steps. The intention is to only do automated changes at first, so I waste less time if it's reverted, and so the first mass change is more clear as an example to other teams that will need to follow similar steps. Steps are described per line, as comments are removed by git: 0. Retrieve the change from the following to build clang-tidy with an additional check: https://github.com/llvm/llvm-project/compare/main...tpopp:llvm-project:tidy-cast-check 1. Build clang-tidy 2. Run clang-tidy over your entire codebase while disabling all checks and enabling the one relevant one. Run on all header files also. 3. Delete .inc files that were also modified, so the next build rebuilds them to a pure state. 4. Some changes have been deleted for the following reasons: - Some files had a variable also named cast - Some files had not included a header file that defines the cast functions - Some files are definitions of the classes that have the casting methods, so the code still refers to the method instead of the function without adding a prefix or removing the method declaration at the same time. ``` ninja -C $BUILD_DIR clang-tidy run-clang-tidy -clang-tidy-binary=$BUILD_DIR/bin/clang-tidy -checks='-*,misc-cast-functions'\ -header-filter=mlir/ mlir/* -fix rm -rf $BUILD_DIR/tools/mlir/**/*.inc git restore mlir/lib/IR mlir/lib/Dialect/DLTI/DLTI.cpp\ mlir/lib/Dialect/Complex/IR/ComplexDialect.cpp\ mlir/lib/**/IR/\ mlir/lib/Dialect/SparseTensor/Transforms/SparseVectorization.cpp\ mlir/lib/Dialect/Vector/Transforms/LowerVectorMultiReduction.cpp\ mlir/test/lib/Dialect/Test/TestTypes.cpp\ mlir/test/lib/Dialect/Transform/TestTransformDialectExtension.cpp\ mlir/test/lib/Dialect/Test/TestAttributes.cpp\ mlir/unittests/TableGen/EnumsGenTest.cpp\ mlir/test/python/lib/PythonTestCAPI.cpp\ mlir/include/mlir/IR/ ``` Differential Revision: https://reviews.llvm.org/D150123
204 lines
8.3 KiB
C++
204 lines
8.3 KiB
C++
//===- BufferUtils.cpp - buffer transformation utilities ------------------===//
|
|
//
|
|
// 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
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
//
|
|
// This file implements utilities for buffer optimization passes.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#include "mlir/Dialect/Bufferization/Transforms/BufferUtils.h"
|
|
#include "mlir/Dialect/Bufferization/Transforms/Bufferize.h"
|
|
#include "mlir/Dialect/MemRef/IR/MemRef.h"
|
|
#include "mlir/Dialect/MemRef/Utils/MemRefUtils.h"
|
|
#include "mlir/IR/Operation.h"
|
|
#include "mlir/Interfaces/ControlFlowInterfaces.h"
|
|
#include "mlir/Interfaces/LoopLikeInterface.h"
|
|
#include "mlir/Pass/Pass.h"
|
|
#include "llvm/ADT/SetOperations.h"
|
|
#include "llvm/ADT/SmallString.h"
|
|
#include <optional>
|
|
|
|
using namespace mlir;
|
|
using namespace mlir::bufferization;
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
// BufferPlacementAllocs
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
/// Get the start operation to place the given alloc value withing the
|
|
// specified placement block.
|
|
Operation *BufferPlacementAllocs::getStartOperation(Value allocValue,
|
|
Block *placementBlock,
|
|
const Liveness &liveness) {
|
|
// We have to ensure that we place the alloc before its first use in this
|
|
// block.
|
|
const LivenessBlockInfo &livenessInfo = *liveness.getLiveness(placementBlock);
|
|
Operation *startOperation = livenessInfo.getStartOperation(allocValue);
|
|
// Check whether the start operation lies in the desired placement block.
|
|
// If not, we will use the terminator as this is the last operation in
|
|
// this block.
|
|
if (startOperation->getBlock() != placementBlock) {
|
|
Operation *opInPlacementBlock =
|
|
placementBlock->findAncestorOpInBlock(*startOperation);
|
|
startOperation = opInPlacementBlock ? opInPlacementBlock
|
|
: placementBlock->getTerminator();
|
|
}
|
|
|
|
return startOperation;
|
|
}
|
|
|
|
/// Initializes the internal list by discovering all supported allocation
|
|
/// nodes.
|
|
BufferPlacementAllocs::BufferPlacementAllocs(Operation *op) { build(op); }
|
|
|
|
/// Searches for and registers all supported allocation entries.
|
|
void BufferPlacementAllocs::build(Operation *op) {
|
|
op->walk([&](MemoryEffectOpInterface opInterface) {
|
|
// Try to find a single allocation result.
|
|
SmallVector<MemoryEffects::EffectInstance, 2> effects;
|
|
opInterface.getEffects(effects);
|
|
|
|
SmallVector<MemoryEffects::EffectInstance, 2> allocateResultEffects;
|
|
llvm::copy_if(
|
|
effects, std::back_inserter(allocateResultEffects),
|
|
[=](MemoryEffects::EffectInstance &it) {
|
|
Value value = it.getValue();
|
|
return isa<MemoryEffects::Allocate>(it.getEffect()) && value &&
|
|
isa<OpResult>(value) &&
|
|
it.getResource() !=
|
|
SideEffects::AutomaticAllocationScopeResource::get();
|
|
});
|
|
// If there is one result only, we will be able to move the allocation and
|
|
// (possibly existing) deallocation ops.
|
|
if (allocateResultEffects.size() != 1)
|
|
return;
|
|
// Get allocation result.
|
|
Value allocValue = allocateResultEffects[0].getValue();
|
|
// Find the associated dealloc value and register the allocation entry.
|
|
std::optional<Operation *> dealloc = memref::findDealloc(allocValue);
|
|
// If the allocation has > 1 dealloc associated with it, skip handling it.
|
|
if (!dealloc)
|
|
return;
|
|
allocs.push_back(std::make_tuple(allocValue, *dealloc));
|
|
});
|
|
}
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
// BufferPlacementTransformationBase
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
/// Constructs a new transformation base using the given root operation.
|
|
BufferPlacementTransformationBase::BufferPlacementTransformationBase(
|
|
Operation *op)
|
|
: aliases(op), allocs(op), liveness(op) {}
|
|
|
|
/// Returns true if the given operation represents a loop by testing whether it
|
|
/// implements the `LoopLikeOpInterface` or the `RegionBranchOpInterface`. In
|
|
/// the case of a `RegionBranchOpInterface`, it checks all region-based control-
|
|
/// flow edges for cycles.
|
|
bool BufferPlacementTransformationBase::isLoop(Operation *op) {
|
|
// If the operation implements the `LoopLikeOpInterface` it can be considered
|
|
// a loop.
|
|
if (isa<LoopLikeOpInterface>(op))
|
|
return true;
|
|
|
|
// If the operation does not implement the `RegionBranchOpInterface`, it is
|
|
// (currently) not possible to detect a loop.
|
|
RegionBranchOpInterface regionInterface;
|
|
if (!(regionInterface = dyn_cast<RegionBranchOpInterface>(op)))
|
|
return false;
|
|
|
|
// Recurses into a region using the current region interface to find potential
|
|
// cycles.
|
|
SmallPtrSet<Region *, 4> visitedRegions;
|
|
std::function<bool(Region *)> recurse = [&](Region *current) {
|
|
if (!current)
|
|
return false;
|
|
// If we have found a back edge, the parent operation induces a loop.
|
|
if (!visitedRegions.insert(current).second)
|
|
return true;
|
|
// Recurses into all region successors.
|
|
SmallVector<RegionSuccessor, 2> successors;
|
|
regionInterface.getSuccessorRegions(current->getRegionNumber(), successors);
|
|
for (RegionSuccessor ®ionEntry : successors)
|
|
if (recurse(regionEntry.getSuccessor()))
|
|
return true;
|
|
return false;
|
|
};
|
|
|
|
// Start with all entry regions and test whether they induce a loop.
|
|
SmallVector<RegionSuccessor, 2> successorRegions;
|
|
regionInterface.getSuccessorRegions(/*index=*/std::nullopt, successorRegions);
|
|
for (RegionSuccessor ®ionEntry : successorRegions) {
|
|
if (recurse(regionEntry.getSuccessor()))
|
|
return true;
|
|
visitedRegions.clear();
|
|
}
|
|
|
|
return false;
|
|
}
|
|
|
|
//===----------------------------------------------------------------------===//
|
|
// BufferPlacementTransformationBase
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
FailureOr<memref::GlobalOp>
|
|
bufferization::getGlobalFor(arith::ConstantOp constantOp, uint64_t alignment,
|
|
Attribute memorySpace) {
|
|
auto type = cast<RankedTensorType>(constantOp.getType());
|
|
auto moduleOp = constantOp->getParentOfType<ModuleOp>();
|
|
if (!moduleOp)
|
|
return failure();
|
|
|
|
// If we already have a global for this constant value, no need to do
|
|
// anything else.
|
|
for (Operation &op : moduleOp.getRegion().getOps()) {
|
|
auto globalOp = dyn_cast<memref::GlobalOp>(&op);
|
|
if (!globalOp)
|
|
continue;
|
|
if (!globalOp.getInitialValue().has_value())
|
|
continue;
|
|
uint64_t opAlignment = globalOp.getAlignment().value_or(0);
|
|
Attribute initialValue = globalOp.getInitialValue().value();
|
|
if (opAlignment == alignment && initialValue == constantOp.getValue())
|
|
return globalOp;
|
|
}
|
|
|
|
// Create a builder without an insertion point. We will insert using the
|
|
// symbol table to guarantee unique names.
|
|
OpBuilder globalBuilder(moduleOp.getContext());
|
|
SymbolTable symbolTable(moduleOp);
|
|
|
|
// Create a pretty name.
|
|
SmallString<64> buf;
|
|
llvm::raw_svector_ostream os(buf);
|
|
interleave(type.getShape(), os, "x");
|
|
os << "x" << type.getElementType();
|
|
|
|
// Add an optional alignment to the global memref.
|
|
IntegerAttr memrefAlignment =
|
|
alignment > 0 ? IntegerAttr::get(globalBuilder.getI64Type(), alignment)
|
|
: IntegerAttr();
|
|
|
|
BufferizeTypeConverter typeConverter;
|
|
auto memrefType = cast<MemRefType>(typeConverter.convertType(type));
|
|
if (memorySpace)
|
|
memrefType = MemRefType::Builder(memrefType).setMemorySpace(memorySpace);
|
|
auto global = globalBuilder.create<memref::GlobalOp>(
|
|
constantOp.getLoc(), (Twine("__constant_") + os.str()).str(),
|
|
/*sym_visibility=*/globalBuilder.getStringAttr("private"),
|
|
/*type=*/memrefType,
|
|
/*initial_value=*/cast<ElementsAttr>(constantOp.getValue()),
|
|
/*constant=*/true,
|
|
/*alignment=*/memrefAlignment);
|
|
symbolTable.insert(global);
|
|
// The symbol table inserts at the end of the module, but globals are a bit
|
|
// nicer if they are at the beginning.
|
|
global->moveBefore(&moduleOp.front());
|
|
return global;
|
|
}
|