llvm-project/mlir/lib/Dialect/Vector/Transforms/VectorDistribute.cpp

1286 lines
55 KiB
C++

//===- VectorDistribute.cpp - patterns to do vector distribution ----------===//
//
// 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 "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/Arith/IR/Arith.h"
#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/IR/SCF.h"
#include "mlir/Dialect/Vector/IR/VectorOps.h"
#include "mlir/Dialect/Vector/Transforms/VectorDistribution.h"
#include "mlir/IR/AffineExpr.h"
#include "mlir/Interfaces/SideEffectInterfaces.h"
#include "mlir/Transforms/RegionUtils.h"
#include "llvm/ADT/SetVector.h"
#include <utility>
using namespace mlir;
using namespace mlir::vector;
/// Currently the distribution map is implicit based on the vector shape. In the
/// future it will be part of the op.
/// Example:
/// ```
/// %0 = vector.warp_execute_on_lane_0(%arg0) -> (vector<1x16x2xf32>) {
/// ...
/// vector.yield %3 : vector<32x16x64xf32>
/// }
/// ```
/// Would have an implicit map of:
/// `(d0, d1, d2) -> (d0, d2)`
static AffineMap calculateImplicitMap(VectorType sequentialType,
VectorType distributedType) {
SmallVector<AffineExpr> perm;
perm.reserve(1);
// Check which dimensions of the sequential type are different than the
// dimensions of the distributed type to know the distributed dimensions. Then
// associate each distributed dimension to an ID in order.
for (unsigned i = 0, e = sequentialType.getRank(); i < e; i++) {
if (sequentialType.getDimSize(i) != distributedType.getDimSize(i))
perm.push_back(getAffineDimExpr(i, distributedType.getContext()));
}
auto map = AffineMap::get(sequentialType.getRank(), 0, perm,
distributedType.getContext());
assert(map.getNumResults() <= 1 &&
"only support distribution along one dimension for now.");
return map;
}
namespace {
/// Helper struct to create the load / store operations that permit transit
/// through the parallel / sequential and the sequential / parallel boundaries
/// when performing `rewriteWarpOpToScfFor`.
///
/// The vector distribution dimension is inferred from the vector types.
struct DistributedLoadStoreHelper {
DistributedLoadStoreHelper(Value sequentialVal, Value distributedVal,
Value laneId, Value zero)
: sequentialVal(sequentialVal), distributedVal(distributedVal),
laneId(laneId), zero(zero) {
sequentialVectorType = sequentialVal.getType().dyn_cast<VectorType>();
distributedVectorType = distributedVal.getType().dyn_cast<VectorType>();
if (sequentialVectorType && distributedVectorType)
distributionMap =
calculateImplicitMap(sequentialVectorType, distributedVectorType);
}
Value buildDistributedOffset(RewriterBase &b, Location loc, int64_t index) {
int64_t distributedSize = distributedVectorType.getDimSize(index);
AffineExpr tid = getAffineSymbolExpr(0, b.getContext());
return b.createOrFold<AffineApplyOp>(loc, tid * distributedSize,
ArrayRef<Value>{laneId});
}
/// Create a store during the process of distributing the
/// `vector.warp_execute_on_thread_0` op.
/// Vector distribution assumes the following convention regarding the
/// temporary buffers that are created to transition values. This **must**
/// be properly specified in the `options.warpAllocationFn`:
/// 1. scalars of type T transit through a memref<1xT>.
/// 2. vectors of type V<shapexT> transit through a memref<shapexT>
Operation *buildStore(RewriterBase &b, Location loc, Value val,
Value buffer) {
assert((val == distributedVal || val == sequentialVal) &&
"Must store either the preregistered distributed or the "
"preregistered sequential value.");
// Scalar case can directly use memref.store.
if (!val.getType().isa<VectorType>())
return b.create<memref::StoreOp>(loc, val, buffer, zero);
// Vector case must use vector::TransferWriteOp which will later lower to
// vector.store of memref.store depending on further lowerings.
int64_t rank = sequentialVectorType.getRank();
SmallVector<Value> indices(rank, zero);
if (val == distributedVal) {
for (auto dimExpr : distributionMap.getResults()) {
int64_t index = dimExpr.cast<AffineDimExpr>().getPosition();
indices[index] = buildDistributedOffset(b, loc, index);
}
}
SmallVector<bool> inBounds(indices.size(), true);
return b.create<vector::TransferWriteOp>(
loc, val, buffer, indices,
ArrayRef<bool>(inBounds.begin(), inBounds.end()));
}
/// Create a load during the process of distributing the
/// `vector.warp_execute_on_thread_0` op.
/// Vector distribution assumes the following convention regarding the
/// temporary buffers that are created to transition values. This **must**
/// be properly specified in the `options.warpAllocationFn`:
/// 1. scalars of type T transit through a memref<1xT>.
/// 2. vectors of type V<shapexT> transit through a memref<shapexT>
///
/// When broadcastMode is true, the load is not distributed to account for
/// the broadcast semantics of the `vector.warp_execute_on_lane_0` op.
///
/// Example:
///
/// ```
/// %r = vector.warp_execute_on_lane_0(...) -> (f32) {
/// vector.yield %cst : f32
/// }
/// // Both types are f32. The constant %cst is broadcasted to all lanes.
/// ```
/// This behavior described in more detail in the documentation of the op.
Value buildLoad(RewriterBase &b, Location loc, Type type, Value buffer) {
// Scalar case can directly use memref.store.
if (!type.isa<VectorType>())
return b.create<memref::LoadOp>(loc, buffer, zero);
// Other cases must be vector atm.
// Vector case must use vector::TransferReadOp which will later lower to
// vector.read of memref.read depending on further lowerings.
assert((type == distributedVectorType || type == sequentialVectorType) &&
"Must store either the preregistered distributed or the "
"preregistered sequential type.");
SmallVector<Value> indices(sequentialVectorType.getRank(), zero);
if (type == distributedVectorType) {
for (auto dimExpr : distributionMap.getResults()) {
int64_t index = dimExpr.cast<AffineDimExpr>().getPosition();
indices[index] = buildDistributedOffset(b, loc, index);
}
}
SmallVector<bool> inBounds(indices.size(), true);
return b.create<vector::TransferReadOp>(
loc, type.cast<VectorType>(), buffer, indices,
ArrayRef<bool>(inBounds.begin(), inBounds.end()));
}
Value sequentialVal, distributedVal, laneId, zero;
VectorType sequentialVectorType, distributedVectorType;
AffineMap distributionMap;
};
} // namespace
/// Helper to create a new WarpExecuteOnLane0Op with different signature.
static WarpExecuteOnLane0Op moveRegionToNewWarpOpAndReplaceReturns(
RewriterBase &rewriter, WarpExecuteOnLane0Op warpOp,
ValueRange newYieldedValues, TypeRange newReturnTypes) {
// Create a new op before the existing one, with the extra operands.
OpBuilder::InsertionGuard g(rewriter);
rewriter.setInsertionPoint(warpOp);
auto newWarpOp = rewriter.create<WarpExecuteOnLane0Op>(
warpOp.getLoc(), newReturnTypes, warpOp.getLaneid(), warpOp.getWarpSize(),
warpOp.getArgs(), warpOp.getBody()->getArgumentTypes());
Region &opBody = warpOp.getBodyRegion();
Region &newOpBody = newWarpOp.getBodyRegion();
Block &newOpFirstBlock = newOpBody.front();
rewriter.inlineRegionBefore(opBody, newOpBody, newOpBody.begin());
rewriter.eraseBlock(&newOpFirstBlock);
assert(newWarpOp.getWarpRegion().hasOneBlock() &&
"expected WarpOp with single block");
auto yield =
cast<vector::YieldOp>(newOpBody.getBlocks().begin()->getTerminator());
rewriter.updateRootInPlace(
yield, [&]() { yield.getOperandsMutable().assign(newYieldedValues); });
return newWarpOp;
}
/// Helper to create a new WarpExecuteOnLane0Op region with extra outputs.
/// `indices` return the index of each new output.
static WarpExecuteOnLane0Op moveRegionToNewWarpOpAndAppendReturns(
RewriterBase &rewriter, WarpExecuteOnLane0Op warpOp,
ValueRange newYieldedValues, TypeRange newReturnTypes,
llvm::SmallVector<size_t> &indices) {
SmallVector<Type> types(warpOp.getResultTypes().begin(),
warpOp.getResultTypes().end());
auto yield = cast<vector::YieldOp>(
warpOp.getBodyRegion().getBlocks().begin()->getTerminator());
llvm::SmallSetVector<Value, 32> yieldValues(yield.getOperands().begin(),
yield.getOperands().end());
for (auto newRet : llvm::zip(newYieldedValues, newReturnTypes)) {
if (yieldValues.insert(std::get<0>(newRet))) {
types.push_back(std::get<1>(newRet));
indices.push_back(yieldValues.size() - 1);
} else {
// If the value already exit the region don't create a new output.
for (auto &yieldOperand : llvm::enumerate(yieldValues.getArrayRef())) {
if (yieldOperand.value() == std::get<0>(newRet)) {
indices.push_back(yieldOperand.index());
break;
}
}
}
}
yieldValues.insert(newYieldedValues.begin(), newYieldedValues.end());
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndReplaceReturns(
rewriter, warpOp, yieldValues.getArrayRef(), types);
rewriter.replaceOp(warpOp,
newWarpOp.getResults().take_front(warpOp.getNumResults()));
return newWarpOp;
}
/// Helper to know if an op can be hoisted out of the region.
static bool canBeHoisted(Operation *op,
function_ref<bool(Value)> definedOutside) {
return llvm::all_of(op->getOperands(), definedOutside) &&
isMemoryEffectFree(op) && op->getNumRegions() == 0;
}
/// Return a value yielded by `warpOp` which statifies the filter lamdba
/// condition and is not dead.
static OpOperand *getWarpResult(WarpExecuteOnLane0Op warpOp,
const std::function<bool(Operation *)> &fn) {
auto yield = cast<vector::YieldOp>(
warpOp.getBodyRegion().getBlocks().begin()->getTerminator());
for (OpOperand &yieldOperand : yield->getOpOperands()) {
Value yieldValues = yieldOperand.get();
Operation *definedOp = yieldValues.getDefiningOp();
if (definedOp && fn(definedOp)) {
if (!warpOp.getResult(yieldOperand.getOperandNumber()).use_empty())
return &yieldOperand;
}
}
return {};
}
// Clones `op` into a new operation that takes `operands` and returns
// `resultTypes`.
static Operation *cloneOpWithOperandsAndTypes(RewriterBase &rewriter,
Location loc, Operation *op,
ArrayRef<Value> operands,
ArrayRef<Type> resultTypes) {
OperationState res(loc, op->getName().getStringRef(), operands, resultTypes,
op->getAttrs());
return rewriter.create(res);
}
namespace {
/// Rewrite a WarpExecuteOnLane0Op into a predicated scf.if op where the single
/// thread `laneId` executes the entirety of the computation.
///
/// After the transformation:
/// - the IR within the scf.if op can be thought of as executing sequentially
/// (from the point of view of threads along `laneId`).
/// - the IR outside of the scf.if op can be thought of as executing in
/// parallel (from the point of view of threads along `laneId`).
///
/// Values that need to transit through the parallel / sequential and the
/// sequential / parallel boundaries do so via reads and writes to a temporary
/// memory location.
///
/// The transformation proceeds in multiple steps:
/// 1. Create the scf.if op.
/// 2. Insert appropriate (alloc, write)-pairs before the scf.if and reads
/// within the scf.if to transit the values captured from above.
/// 3. Synchronize before the scf.if to ensure all writes inserted in 2. are
/// consistent within the scf.if.
/// 4. Move the body of the WarpExecuteOnLane0Op inside the scf.if.
/// 5. Insert appropriate writes within scf.if and reads after the scf.if to
/// transit the values returned by the op.
/// 6. Synchronize after the scf.if to ensure all writes inserted in 5. are
/// consistent after the scf.if.
/// 7. Perform late cleanups.
///
/// All this assumes the vector distribution occurs along the most minor
/// distributed vector dimension.
struct WarpOpToScfIfPattern : public OpRewritePattern<WarpExecuteOnLane0Op> {
WarpOpToScfIfPattern(MLIRContext *context,
const WarpExecuteOnLane0LoweringOptions &options,
PatternBenefit benefit = 1)
: OpRewritePattern<WarpExecuteOnLane0Op>(context, benefit),
options(options) {}
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
assert(warpOp.getBodyRegion().hasOneBlock() &&
"expected WarpOp with single block");
Block *warpOpBody = &warpOp.getBodyRegion().front();
Location loc = warpOp.getLoc();
// Passed all checks. Start rewriting.
OpBuilder::InsertionGuard g(rewriter);
rewriter.setInsertionPoint(warpOp);
// Step 1: Create scf.if op.
Value c0 = rewriter.create<arith::ConstantIndexOp>(loc, 0);
Value isLane0 = rewriter.create<arith::CmpIOp>(
loc, arith::CmpIPredicate::eq, warpOp.getLaneid(), c0);
auto ifOp = rewriter.create<scf::IfOp>(loc, isLane0,
/*withElseRegion=*/false);
rewriter.eraseOp(ifOp.thenBlock()->getTerminator());
// Step 2: insert appropriate (alloc, write)-pairs before the scf.if and
// reads within the scf.if to transit the values captured from above.
SmallVector<Value> bbArgReplacements;
for (const auto &it : llvm::enumerate(warpOp.getArgs())) {
Value sequentialVal = warpOpBody->getArgument(it.index());
Value distributedVal = it.value();
DistributedLoadStoreHelper helper(sequentialVal, distributedVal,
warpOp.getLaneid(), c0);
// Create buffer before the ifOp.
rewriter.setInsertionPoint(ifOp);
Value buffer = options.warpAllocationFn(loc, rewriter, warpOp,
sequentialVal.getType());
// Store distributed vector into buffer, before the ifOp.
helper.buildStore(rewriter, loc, distributedVal, buffer);
// Load sequential vector from buffer, inside the ifOp.
rewriter.setInsertionPointToStart(ifOp.thenBlock());
bbArgReplacements.push_back(
helper.buildLoad(rewriter, loc, sequentialVal.getType(), buffer));
}
// Step 3. Insert sync after all the stores and before all the loads.
if (!warpOp.getArgs().empty()) {
rewriter.setInsertionPoint(ifOp);
options.warpSyncronizationFn(loc, rewriter, warpOp);
}
// Step 4. Move body of warpOp to ifOp.
rewriter.mergeBlocks(warpOpBody, ifOp.thenBlock(), bbArgReplacements);
// Step 5. Insert appropriate writes within scf.if and reads after the
// scf.if to transit the values returned by the op.
// TODO: at this point, we can reuse the shared memory from previous
// buffers.
SmallVector<Value> replacements;
auto yieldOp = cast<vector::YieldOp>(ifOp.thenBlock()->getTerminator());
Location yieldLoc = yieldOp.getLoc();
for (const auto &it : llvm::enumerate(yieldOp.getOperands())) {
Value sequentialVal = it.value();
Value distributedVal = warpOp->getResult(it.index());
DistributedLoadStoreHelper helper(sequentialVal, distributedVal,
warpOp.getLaneid(), c0);
// Create buffer before the ifOp.
rewriter.setInsertionPoint(ifOp);
Value buffer = options.warpAllocationFn(loc, rewriter, warpOp,
sequentialVal.getType());
// Store yielded value into buffer, inside the ifOp, before the
// terminator.
rewriter.setInsertionPoint(yieldOp);
helper.buildStore(rewriter, loc, sequentialVal, buffer);
// Load distributed value from buffer, after the warpOp.
rewriter.setInsertionPointAfter(ifOp);
// Result type and yielded value type are the same. This is a broadcast.
// E.g.:
// %r = vector.warp_execute_on_lane_0(...) -> (f32) {
// vector.yield %cst : f32
// }
// Both types are f32. The constant %cst is broadcasted to all lanes.
// This is described in more detail in the documentation of the op.
replacements.push_back(
helper.buildLoad(rewriter, loc, distributedVal.getType(), buffer));
}
// Step 6. Insert sync after all the stores and before all the loads.
if (!yieldOp.getOperands().empty()) {
rewriter.setInsertionPointAfter(ifOp);
options.warpSyncronizationFn(loc, rewriter, warpOp);
}
// Step 7. Delete terminator and add empty scf.yield.
rewriter.eraseOp(yieldOp);
rewriter.setInsertionPointToEnd(ifOp.thenBlock());
rewriter.create<scf::YieldOp>(yieldLoc);
// Compute replacements for WarpOp results.
rewriter.replaceOp(warpOp, replacements);
return success();
}
private:
const WarpExecuteOnLane0LoweringOptions &options;
};
/// Clone `writeOp` assumed to be nested under `warpOp` into a new warp execute
/// op with the proper return type.
/// The new write op is updated to write the result of the new warp execute op.
/// The old `writeOp` is deleted.
static vector::TransferWriteOp cloneWriteOp(RewriterBase &rewriter,
WarpExecuteOnLane0Op warpOp,
vector::TransferWriteOp writeOp,
VectorType targetType) {
assert(writeOp->getParentOp() == warpOp &&
"write must be nested immediately under warp");
OpBuilder::InsertionGuard g(rewriter);
SmallVector<size_t> newRetIndices;
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndAppendReturns(
rewriter, warpOp, ValueRange{{writeOp.getVector()}},
TypeRange{targetType}, newRetIndices);
rewriter.setInsertionPointAfter(newWarpOp);
auto newWriteOp =
cast<vector::TransferWriteOp>(rewriter.clone(*writeOp.getOperation()));
rewriter.eraseOp(writeOp);
newWriteOp.getVectorMutable().assign(newWarpOp.getResult(newRetIndices[0]));
return newWriteOp;
}
/// Return the distributed vector type based on the original type and the
/// distribution map. The map is expected to have a dimension equal to the
/// original type rank and should be a projection where the results are the
/// distributed dimensions. The number of results should be equal to the number
/// of warp sizes which is currently limited to 1.
/// Example: For a vector<16x32x64> distributed with a map(d0, d1, d2) -> (d1)
/// and a warp size of 16 would distribute the second dimension (associated to
/// d1) and return vector<16x2x64>
static VectorType getDistributedType(VectorType originalType, AffineMap map,
int64_t warpSize) {
if (map.getNumResults() != 1)
return VectorType();
SmallVector<int64_t> targetShape(originalType.getShape().begin(),
originalType.getShape().end());
for (unsigned i = 0, e = map.getNumResults(); i < e; i++) {
unsigned position = map.getDimPosition(i);
if (targetShape[position] % warpSize != 0)
return VectorType();
targetShape[position] = targetShape[position] / warpSize;
}
VectorType targetType =
VectorType::get(targetShape, originalType.getElementType());
return targetType;
}
/// Distribute transfer_write ops based on the affine map returned by
/// `distributionMapFn`.
/// Example:
/// ```
/// %0 = vector.warp_execute_on_lane_0(%id){
/// ...
/// vector.transfer_write %v, %A[%c0] : vector<32xf32>, memref<128xf32>
/// vector.yield
/// }
/// ```
/// To
/// ```
/// %r:3 = vector.warp_execute_on_lane_0(%id) -> (vector<1xf32>) {
/// ...
/// vector.yield %v : vector<32xf32>
/// }
/// vector.transfer_write %v, %A[%id] : vector<1xf32>, memref<128xf32>
struct WarpOpTransferWrite : public OpRewritePattern<vector::TransferWriteOp> {
WarpOpTransferWrite(MLIRContext *ctx, DistributionMapFn fn,
PatternBenefit b = 1)
: OpRewritePattern<vector::TransferWriteOp>(ctx, b),
distributionMapFn(std::move(fn)) {}
/// Distribute the TransferWriteOp. Only 1D distributions and vector dims that
/// are multiples of the distribution ratio are supported at the moment.
LogicalResult tryDistributeOp(RewriterBase &rewriter,
vector::TransferWriteOp writeOp,
WarpExecuteOnLane0Op warpOp) const {
VectorType writtenVectorType = writeOp.getVectorType();
// 1. If the write is 0-D, we just clone it into a new WarpExecuteOnLane0Op
// to separate it from the rest.
if (writtenVectorType.getRank() == 0)
return failure();
// 2. Compute the distributed type.
AffineMap map = distributionMapFn(writeOp.getVector());
VectorType targetType =
getDistributedType(writtenVectorType, map, warpOp.getWarpSize());
if (!targetType)
return failure();
// 3. clone the write into a new WarpExecuteOnLane0Op to separate it from
// the rest.
vector::TransferWriteOp newWriteOp =
cloneWriteOp(rewriter, warpOp, writeOp, targetType);
// 4. Reindex the write using the distribution map.
auto newWarpOp =
newWriteOp.getVector().getDefiningOp<WarpExecuteOnLane0Op>();
rewriter.setInsertionPoint(newWriteOp);
AffineMap indexMap = map.compose(newWriteOp.getPermutationMap());
Location loc = newWriteOp.getLoc();
SmallVector<Value> indices(newWriteOp.getIndices().begin(),
newWriteOp.getIndices().end());
for (auto it : llvm::zip(indexMap.getResults(), map.getResults())) {
AffineExpr d0, d1;
bindDims(newWarpOp.getContext(), d0, d1);
auto indexExpr = std::get<0>(it).dyn_cast<AffineDimExpr>();
if (!indexExpr)
continue;
unsigned indexPos = indexExpr.getPosition();
unsigned vectorPos = std::get<1>(it).cast<AffineDimExpr>().getPosition();
auto scale =
rewriter.getAffineConstantExpr(targetType.getDimSize(vectorPos));
indices[indexPos] =
makeComposedAffineApply(rewriter, loc, d0 + scale * d1,
{indices[indexPos], newWarpOp.getLaneid()});
}
newWriteOp.getIndicesMutable().assign(indices);
return success();
}
/// Extract TransferWriteOps of vector<1x> into a separate warp op.
LogicalResult tryExtractOp(RewriterBase &rewriter,
vector::TransferWriteOp writeOp,
WarpExecuteOnLane0Op warpOp) const {
Location loc = writeOp.getLoc();
VectorType vecType = writeOp.getVectorType();
// Only sink out vector of 1 element for now to not serialize large vector
// store. This can later be controlled by user.
if (vecType.getNumElements() != 1)
return failure();
// Do not process warp ops that contain only TransferWriteOps.
if (llvm::all_of(warpOp.getOps(), [](Operation &op) {
return isa<vector::TransferWriteOp, vector::YieldOp>(&op);
}))
return failure();
SmallVector<Value> yieldValues = {writeOp.getVector()};
SmallVector<Type> retTypes = {vecType};
SmallVector<size_t> newRetIndices;
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndAppendReturns(
rewriter, warpOp, yieldValues, retTypes, newRetIndices);
rewriter.setInsertionPointAfter(newWarpOp);
// Create a second warp op that contains only writeOp.
auto secondWarpOp = rewriter.create<WarpExecuteOnLane0Op>(
loc, TypeRange(), newWarpOp.getLaneid(), newWarpOp.getWarpSize());
Block &body = secondWarpOp.getBodyRegion().front();
rewriter.setInsertionPointToStart(&body);
auto newWriteOp =
cast<vector::TransferWriteOp>(rewriter.clone(*writeOp.getOperation()));
newWriteOp.getVectorMutable().assign(newWarpOp.getResult(newRetIndices[0]));
rewriter.eraseOp(writeOp);
rewriter.create<vector::YieldOp>(newWarpOp.getLoc());
return success();
}
LogicalResult matchAndRewrite(vector::TransferWriteOp writeOp,
PatternRewriter &rewriter) const override {
// Ops with mask not supported yet.
if (writeOp.getMask())
return failure();
auto warpOp = dyn_cast<WarpExecuteOnLane0Op>(writeOp->getParentOp());
if (!warpOp)
return failure();
// There must be no op with a side effect after writeOp.
Operation *nextOp = writeOp.getOperation();
while ((nextOp = nextOp->getNextNode()))
if (!isMemoryEffectFree(nextOp))
return failure();
if (!llvm::all_of(writeOp->getOperands(), [&](Value value) {
return writeOp.getVector() == value ||
warpOp.isDefinedOutsideOfRegion(value);
}))
return failure();
if (succeeded(tryDistributeOp(rewriter, writeOp, warpOp)))
return success();
if (succeeded(tryExtractOp(rewriter, writeOp, warpOp)))
return success();
return failure();
}
private:
DistributionMapFn distributionMapFn;
};
/// Sink out elementwise op feeding into a warp op yield.
/// ```
/// %0 = vector.warp_execute_on_lane_0(%arg0) -> (vector<1xf32>) {
/// ...
/// %3 = arith.addf %1, %2 : vector<32xf32>
/// vector.yield %3 : vector<32xf32>
/// }
/// ```
/// To
/// ```
/// %r:3 = vector.warp_execute_on_lane_0(%arg0) -> (vector<1xf32>,
/// vector<1xf32>, vector<1xf32>) {
/// ...
/// %4 = arith.addf %2, %3 : vector<32xf32>
/// vector.yield %4, %2, %3 : vector<32xf32>, vector<32xf32>,
/// vector<32xf32>
/// }
/// %0 = arith.addf %r#1, %r#2 : vector<1xf32>
struct WarpOpElementwise : public OpRewritePattern<WarpExecuteOnLane0Op> {
using OpRewritePattern<WarpExecuteOnLane0Op>::OpRewritePattern;
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
OpOperand *yieldOperand = getWarpResult(warpOp, [](Operation *op) {
return OpTrait::hasElementwiseMappableTraits(op);
});
if (!yieldOperand)
return failure();
Operation *elementWise = yieldOperand->get().getDefiningOp();
unsigned operandIndex = yieldOperand->getOperandNumber();
Value distributedVal = warpOp.getResult(operandIndex);
SmallVector<Value> yieldValues;
SmallVector<Type> retTypes;
Location loc = warpOp.getLoc();
for (OpOperand &operand : elementWise->getOpOperands()) {
Type targetType;
if (auto vecType = distributedVal.getType().dyn_cast<VectorType>()) {
// If the result type is a vector, the operands must also be vectors.
auto operandType = operand.get().getType().cast<VectorType>();
targetType =
VectorType::get(vecType.getShape(), operandType.getElementType());
} else {
auto operandType = operand.get().getType();
assert(!operandType.isa<VectorType>() &&
"unexpected yield of vector from op with scalar result type");
targetType = operandType;
}
retTypes.push_back(targetType);
yieldValues.push_back(operand.get());
}
SmallVector<size_t> newRetIndices;
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndAppendReturns(
rewriter, warpOp, yieldValues, retTypes, newRetIndices);
rewriter.setInsertionPointAfter(newWarpOp);
SmallVector<Value> newOperands(elementWise->getOperands().begin(),
elementWise->getOperands().end());
for (unsigned i : llvm::seq(unsigned(0), elementWise->getNumOperands())) {
newOperands[i] = newWarpOp.getResult(newRetIndices[i]);
}
OpBuilder::InsertionGuard g(rewriter);
rewriter.setInsertionPointAfter(newWarpOp);
Operation *newOp = cloneOpWithOperandsAndTypes(
rewriter, loc, elementWise, newOperands,
{newWarpOp.getResult(operandIndex).getType()});
newWarpOp.getResult(operandIndex).replaceAllUsesWith(newOp->getResult(0));
return success();
}
};
/// Sink out splat constant op feeding into a warp op yield.
/// ```
/// %0 = vector.warp_execute_on_lane_0(%arg0) -> (vector<1xf32>) {
/// ...
/// %cst = arith.constant dense<2.0> : vector<32xf32>
/// vector.yield %cst : vector<32xf32>
/// }
/// ```
/// To
/// ```
/// vector.warp_execute_on_lane_0(%arg0 {
/// ...
/// }
/// %0 = arith.constant dense<2.0> : vector<1xf32>
struct WarpOpConstant : public OpRewritePattern<WarpExecuteOnLane0Op> {
using OpRewritePattern<WarpExecuteOnLane0Op>::OpRewritePattern;
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
OpOperand *yieldOperand = getWarpResult(
warpOp, [](Operation *op) { return isa<arith::ConstantOp>(op); });
if (!yieldOperand)
return failure();
auto constantOp = yieldOperand->get().getDefiningOp<arith::ConstantOp>();
auto dense = constantOp.getValue().dyn_cast<SplatElementsAttr>();
if (!dense)
return failure();
unsigned operandIndex = yieldOperand->getOperandNumber();
Attribute scalarAttr = dense.getSplatValue<Attribute>();
Attribute newAttr = DenseElementsAttr::get(
warpOp.getResult(operandIndex).getType(), scalarAttr);
Location loc = warpOp.getLoc();
rewriter.setInsertionPointAfter(warpOp);
Value distConstant = rewriter.create<arith::ConstantOp>(loc, newAttr);
warpOp.getResult(operandIndex).replaceAllUsesWith(distConstant);
return success();
}
};
/// Sink out transfer_read op feeding into a warp op yield.
/// ```
/// %0 = vector.warp_execute_on_lane_0(%arg0) -> (vector<1xf32>) {
/// ...
// %2 = vector.transfer_read %src[%c0], %cst : memref<1024xf32>,
// vector<32xf32>
/// vector.yield %2 : vector<32xf32>
/// }
/// ```
/// To
/// ```
/// %dead = vector.warp_execute_on_lane_0(%arg0) -> (vector<1xf32>,
/// vector<1xf32>, vector<1xf32>) {
/// ...
/// %2 = vector.transfer_read %src[%c0], %cst : memref<1024xf32>,
/// vector<32xf32> vector.yield %2 : vector<32xf32>
/// }
/// %0 = vector.transfer_read %src[%c0], %cst : memref<1024xf32>, vector<1xf32>
struct WarpOpTransferRead : public OpRewritePattern<WarpExecuteOnLane0Op> {
using OpRewritePattern<WarpExecuteOnLane0Op>::OpRewritePattern;
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
OpOperand *operand = getWarpResult(
warpOp, [](Operation *op) { return isa<vector::TransferReadOp>(op); });
if (!operand)
return failure();
auto read = operand->get().getDefiningOp<vector::TransferReadOp>();
// Don't duplicate transfer_read ops when distributing.
if (!read.getResult().hasOneUse())
return failure();
unsigned operandIndex = operand->getOperandNumber();
Value distributedVal = warpOp.getResult(operandIndex);
SmallVector<Value, 4> indices(read.getIndices().begin(),
read.getIndices().end());
auto sequentialType = read.getResult().getType().cast<VectorType>();
auto distributedType = distributedVal.getType().cast<VectorType>();
AffineMap map = calculateImplicitMap(sequentialType, distributedType);
AffineMap indexMap = map.compose(read.getPermutationMap());
OpBuilder::InsertionGuard g(rewriter);
rewriter.setInsertionPointAfter(warpOp);
for (auto it : llvm::zip(indexMap.getResults(), map.getResults())) {
AffineExpr d0, d1;
bindDims(read.getContext(), d0, d1);
auto indexExpr = std::get<0>(it).dyn_cast<AffineDimExpr>();
if (!indexExpr)
continue;
unsigned indexPos = indexExpr.getPosition();
unsigned vectorPos = std::get<1>(it).cast<AffineDimExpr>().getPosition();
int64_t scale =
distributedVal.getType().cast<VectorType>().getDimSize(vectorPos);
indices[indexPos] =
makeComposedAffineApply(rewriter, read.getLoc(), d0 + scale * d1,
{indices[indexPos], warpOp.getLaneid()});
}
Value newRead = rewriter.create<vector::TransferReadOp>(
read.getLoc(), distributedVal.getType(), read.getSource(), indices,
read.getPermutationMapAttr(), read.getPadding(), read.getMask(),
read.getInBoundsAttr());
distributedVal.replaceAllUsesWith(newRead);
return success();
}
};
/// Remove any result that has no use along with the matching yieldOp operand.
// TODO: Move this in WarpExecuteOnLane0Op canonicalization.
struct WarpOpDeadResult : public OpRewritePattern<WarpExecuteOnLane0Op> {
using OpRewritePattern<WarpExecuteOnLane0Op>::OpRewritePattern;
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
SmallVector<Type> newResultTypes;
newResultTypes.reserve(warpOp->getNumResults());
SmallVector<Value> newYieldValues;
newYieldValues.reserve(warpOp->getNumResults());
DenseMap<Value, int64_t> dedupYieldOperandPositionMap;
DenseMap<OpResult, int64_t> dedupResultPositionMap;
auto yield = cast<vector::YieldOp>(
warpOp.getBodyRegion().getBlocks().begin()->getTerminator());
// Some values may be yielded multiple times and correspond to multiple
// results. Deduplicating occurs by taking each result with its matching
// yielded value, and:
// 1. recording the unique first position at which the value is yielded.
// 2. recording for the result, the first position at which the dedup'ed
// value is yielded.
// 3. skipping from the new result types / new yielded values any result
// that has no use or whose yielded value has already been seen.
for (OpResult result : warpOp.getResults()) {
Value yieldOperand = yield.getOperand(result.getResultNumber());
auto it = dedupYieldOperandPositionMap.insert(
std::make_pair(yieldOperand, newResultTypes.size()));
dedupResultPositionMap.insert(std::make_pair(result, it.first->second));
if (result.use_empty() || !it.second)
continue;
newResultTypes.push_back(result.getType());
newYieldValues.push_back(yieldOperand);
}
// No modification, exit early.
if (yield.getNumOperands() == newYieldValues.size())
return failure();
// Move the body of the old warpOp to a new warpOp.
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndReplaceReturns(
rewriter, warpOp, newYieldValues, newResultTypes);
// Replace results of the old warpOp by the new, deduplicated results.
SmallVector<Value> newValues;
newValues.reserve(warpOp->getNumResults());
for (OpResult result : warpOp.getResults()) {
if (result.use_empty())
newValues.push_back(Value());
else
newValues.push_back(
newWarpOp.getResult(dedupResultPositionMap.lookup(result)));
}
rewriter.replaceOp(warpOp, newValues);
return success();
}
};
// If an operand is directly yielded out of the region we can forward it
// directly and it doesn't need to go through the region.
struct WarpOpForwardOperand : public OpRewritePattern<WarpExecuteOnLane0Op> {
using OpRewritePattern<WarpExecuteOnLane0Op>::OpRewritePattern;
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
SmallVector<Type> resultTypes;
SmallVector<Value> yieldValues;
auto yield = cast<vector::YieldOp>(
warpOp.getBodyRegion().getBlocks().begin()->getTerminator());
Value valForwarded;
unsigned resultIndex;
for (OpOperand &operand : yield->getOpOperands()) {
Value result = warpOp.getResult(operand.getOperandNumber());
if (result.use_empty())
continue;
// Assume all the values coming from above are uniform.
if (!warpOp.getBodyRegion().isAncestor(operand.get().getParentRegion())) {
if (result.getType() != operand.get().getType())
continue;
valForwarded = operand.get();
resultIndex = operand.getOperandNumber();
break;
}
auto arg = operand.get().dyn_cast<BlockArgument>();
if (!arg || arg.getOwner()->getParentOp() != warpOp.getOperation())
continue;
Value warpOperand = warpOp.getArgs()[arg.getArgNumber()];
if (result.getType() != warpOperand.getType())
continue;
valForwarded = warpOperand;
resultIndex = operand.getOperandNumber();
break;
}
if (!valForwarded)
return failure();
warpOp.getResult(resultIndex).replaceAllUsesWith(valForwarded);
return success();
}
};
struct WarpOpBroadcast : public OpRewritePattern<WarpExecuteOnLane0Op> {
using OpRewritePattern<WarpExecuteOnLane0Op>::OpRewritePattern;
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
OpOperand *operand = getWarpResult(
warpOp, [](Operation *op) { return isa<vector::BroadcastOp>(op); });
if (!operand)
return failure();
unsigned int operandNumber = operand->getOperandNumber();
auto broadcastOp = operand->get().getDefiningOp<vector::BroadcastOp>();
Location loc = broadcastOp.getLoc();
auto destVecType =
warpOp->getResultTypes()[operandNumber].cast<VectorType>();
SmallVector<size_t> newRetIndices;
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndAppendReturns(
rewriter, warpOp, {broadcastOp.getSource()},
{broadcastOp.getSource().getType()}, newRetIndices);
rewriter.setInsertionPointAfter(newWarpOp);
Value broadcasted = rewriter.create<vector::BroadcastOp>(
loc, destVecType, newWarpOp->getResult(newRetIndices[0]));
newWarpOp->getResult(operandNumber).replaceAllUsesWith(broadcasted);
return success();
}
};
/// Pattern to move out vector.extract of single element vector. Those don't
/// need to be distributed and can just be propagated outside of the region.
struct WarpOpExtract : public OpRewritePattern<WarpExecuteOnLane0Op> {
using OpRewritePattern<WarpExecuteOnLane0Op>::OpRewritePattern;
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
OpOperand *operand = getWarpResult(
warpOp, [](Operation *op) { return isa<vector::ExtractOp>(op); });
if (!operand)
return failure();
unsigned int operandNumber = operand->getOperandNumber();
auto extractOp = operand->get().getDefiningOp<vector::ExtractOp>();
if (extractOp.getVectorType().getNumElements() != 1)
return failure();
Location loc = extractOp.getLoc();
SmallVector<size_t> newRetIndices;
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndAppendReturns(
rewriter, warpOp, {extractOp.getVector()}, {extractOp.getVectorType()},
newRetIndices);
rewriter.setInsertionPointAfter(newWarpOp);
Value newExtract = rewriter.create<vector::ExtractOp>(
loc, newWarpOp->getResult(newRetIndices[0]), extractOp.getPosition());
newWarpOp->getResult(operandNumber).replaceAllUsesWith(newExtract);
return success();
}
};
/// Pattern to move out vector.extractelement of 0-D tensors. Those don't
/// need to be distributed and can just be propagated outside of the region.
struct WarpOpExtractElement : public OpRewritePattern<WarpExecuteOnLane0Op> {
WarpOpExtractElement(MLIRContext *ctx, WarpShuffleFromIdxFn fn,
PatternBenefit b = 1)
: OpRewritePattern<WarpExecuteOnLane0Op>(ctx, b),
warpShuffleFromIdxFn(std::move(fn)) {}
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
OpOperand *operand = getWarpResult(warpOp, [](Operation *op) {
return isa<vector::ExtractElementOp>(op);
});
if (!operand)
return failure();
unsigned int operandNumber = operand->getOperandNumber();
auto extractOp = operand->get().getDefiningOp<vector::ExtractElementOp>();
VectorType extractSrcType = extractOp.getVectorType();
bool is0dExtract = extractSrcType.getRank() == 0;
Type elType = extractSrcType.getElementType();
VectorType distributedVecType;
if (!is0dExtract) {
assert(extractSrcType.getRank() == 1 &&
"expected that extractelement src rank is 0 or 1");
int64_t elementsPerLane =
extractSrcType.getShape()[0] / warpOp.getWarpSize();
distributedVecType = VectorType::get({elementsPerLane}, elType);
} else {
distributedVecType = extractSrcType;
}
// Yield source vector from warp op.
Location loc = extractOp.getLoc();
SmallVector<size_t> newRetIndices;
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndAppendReturns(
rewriter, warpOp, {extractOp.getVector()}, {distributedVecType},
newRetIndices);
rewriter.setInsertionPointAfter(newWarpOp);
Value distributedVec = newWarpOp->getResult(newRetIndices[0]);
// 0d extract: The new warp op broadcasts the source vector to all lanes.
// All lanes extract the scalar.
if (is0dExtract) {
Value newExtract =
rewriter.create<vector::ExtractElementOp>(loc, distributedVec);
newWarpOp->getResult(operandNumber).replaceAllUsesWith(newExtract);
return success();
}
// 1d extract: Distribute the source vector. One lane extracts and shuffles
// the value to all other lanes.
int64_t elementsPerLane = distributedVecType.getShape()[0];
AffineExpr sym0 = getAffineSymbolExpr(0, rewriter.getContext());
// tid of extracting thread: pos / elementsPerLane
Value broadcastFromTid = rewriter.create<AffineApplyOp>(
loc, sym0.ceilDiv(elementsPerLane), extractOp.getPosition());
// Extract at position: pos % elementsPerLane
Value pos = rewriter.create<AffineApplyOp>(loc, sym0 % elementsPerLane,
extractOp.getPosition());
Value extracted =
rewriter.create<vector::ExtractElementOp>(loc, distributedVec, pos);
// Shuffle the extracted value to all lanes.
Value shuffled = warpShuffleFromIdxFn(
loc, rewriter, extracted, broadcastFromTid, newWarpOp.getWarpSize());
newWarpOp->getResult(operandNumber).replaceAllUsesWith(shuffled);
return success();
}
private:
WarpShuffleFromIdxFn warpShuffleFromIdxFn;
};
/// Sink scf.for region out of WarpExecuteOnLane0Op. This can be done only if
/// the scf.ForOp is the last operation in the region so that it doesn't change
/// the order of execution. This creates a new scf.for region after the
/// WarpExecuteOnLane0Op. The new scf.for region will contain a new
/// WarpExecuteOnLane0Op region. Example:
/// ```
/// %w = vector.warp_execute_on_lane_0(%laneid) -> (vector<4xf32>) {
/// ...
/// %v1 = scf.for %arg3 = %c0 to %c128 step %c1 iter_args(%arg4 = %v)
/// -> (vector<128xf32>) {
/// ...
/// scf.yield %r : vector<128xf32>
/// }
/// vector.yield %v1 : vector<128xf32>
/// }
/// ```
/// To:
/// %w0 = vector.warp_execute_on_lane_0(%arg0) -> (vector<4xf32>) {
/// ...
/// vector.yield %v : vector<128xf32>
/// }
/// %w = scf.for %arg3 = %c0 to %c128 step %c1 iter_args(%varg = %q0)
/// -> (vector<4xf32>) {
/// %iw = vector.warp_execute_on_lane_0(%laneid)
/// args(%varg : vector<4xf32>) -> (vector<4xf32>) {
/// ^bb0(%arg: vector<128xf32>):
/// ...
/// vector.yield %ir : vector<128xf32>
/// }
/// scf.yield %iw : vector<4xf32>
/// }
/// ```
struct WarpOpScfForOp : public OpRewritePattern<WarpExecuteOnLane0Op> {
WarpOpScfForOp(MLIRContext *ctx, DistributionMapFn fn, PatternBenefit b = 1)
: OpRewritePattern<WarpExecuteOnLane0Op>(ctx, b),
distributionMapFn(std::move(fn)) {}
using OpRewritePattern<WarpExecuteOnLane0Op>::OpRewritePattern;
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
auto yield = cast<vector::YieldOp>(
warpOp.getBodyRegion().getBlocks().begin()->getTerminator());
// Only pick up forOp if it is the last op in the region.
Operation *lastNode = yield->getPrevNode();
auto forOp = dyn_cast_or_null<scf::ForOp>(lastNode);
if (!forOp)
return failure();
// Collect Values that come from the warp op but are outside the forOp.
// Those Value needs to be returned by the original warpOp and passed to the
// new op.
llvm::SmallSetVector<Value, 32> escapingValues;
SmallVector<Type> inputTypes;
SmallVector<Type> distTypes;
mlir::visitUsedValuesDefinedAbove(
forOp.getBodyRegion(), [&](OpOperand *operand) {
Operation *parent = operand->get().getParentRegion()->getParentOp();
if (warpOp->isAncestor(parent)) {
if (!escapingValues.insert(operand->get()))
return;
Type distType = operand->get().getType();
if (auto vecType = distType.cast<VectorType>()) {
AffineMap map = distributionMapFn(operand->get());
distType = getDistributedType(vecType, map, warpOp.getWarpSize());
}
inputTypes.push_back(operand->get().getType());
distTypes.push_back(distType);
}
});
SmallVector<size_t> newRetIndices;
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndAppendReturns(
rewriter, warpOp, escapingValues.getArrayRef(), distTypes,
newRetIndices);
yield = cast<vector::YieldOp>(
newWarpOp.getBodyRegion().getBlocks().begin()->getTerminator());
SmallVector<Value> newOperands;
SmallVector<unsigned> resultIdx;
// Collect all the outputs coming from the forOp.
for (OpOperand &yieldOperand : yield->getOpOperands()) {
if (yieldOperand.get().getDefiningOp() != forOp.getOperation())
continue;
auto forResult = yieldOperand.get().cast<OpResult>();
newOperands.push_back(
newWarpOp.getResult(yieldOperand.getOperandNumber()));
yieldOperand.set(forOp.getIterOperands()[forResult.getResultNumber()]);
resultIdx.push_back(yieldOperand.getOperandNumber());
}
OpBuilder::InsertionGuard g(rewriter);
rewriter.setInsertionPointAfter(newWarpOp);
// Create a new for op outside the region with a WarpExecuteOnLane0Op region
// inside.
auto newForOp = rewriter.create<scf::ForOp>(
forOp.getLoc(), forOp.getLowerBound(), forOp.getUpperBound(),
forOp.getStep(), newOperands);
rewriter.setInsertionPoint(newForOp.getBody(), newForOp.getBody()->begin());
SmallVector<Value> warpInput(newForOp.getRegionIterArgs().begin(),
newForOp.getRegionIterArgs().end());
SmallVector<Type> warpInputType(forOp.getResultTypes().begin(),
forOp.getResultTypes().end());
llvm::SmallDenseMap<Value, int64_t> argIndexMapping;
for (auto [i, retIdx] : llvm::enumerate(newRetIndices)) {
warpInput.push_back(newWarpOp.getResult(retIdx));
argIndexMapping[escapingValues[i]] = warpInputType.size();
warpInputType.push_back(inputTypes[i]);
}
auto innerWarp = rewriter.create<WarpExecuteOnLane0Op>(
newWarpOp.getLoc(), newForOp.getResultTypes(), newWarpOp.getLaneid(),
newWarpOp.getWarpSize(), warpInput, warpInputType);
SmallVector<Value> argMapping;
argMapping.push_back(newForOp.getInductionVar());
for (Value args : innerWarp.getBody()->getArguments()) {
argMapping.push_back(args);
}
argMapping.resize(forOp.getBody()->getNumArguments());
SmallVector<Value> yieldOperands;
for (Value operand : forOp.getBody()->getTerminator()->getOperands())
yieldOperands.push_back(operand);
rewriter.eraseOp(forOp.getBody()->getTerminator());
rewriter.mergeBlocks(forOp.getBody(), innerWarp.getBody(), argMapping);
rewriter.setInsertionPoint(innerWarp.getBody(), innerWarp.getBody()->end());
rewriter.create<vector::YieldOp>(innerWarp.getLoc(), yieldOperands);
rewriter.setInsertionPointAfter(innerWarp);
if (!innerWarp.getResults().empty())
rewriter.create<scf::YieldOp>(forOp.getLoc(), innerWarp.getResults());
rewriter.eraseOp(forOp);
// Replace the warpOp result coming from the original ForOp.
for (const auto &res : llvm::enumerate(resultIdx)) {
newWarpOp.getResult(res.value())
.replaceAllUsesWith(newForOp.getResult(res.index()));
newForOp->setOperand(res.index() + 3, newWarpOp.getResult(res.value()));
}
newForOp.walk([&](Operation *op) {
for (OpOperand &operand : op->getOpOperands()) {
auto it = argIndexMapping.find(operand.get());
if (it == argIndexMapping.end())
continue;
operand.set(innerWarp.getBodyRegion().getArgument(it->second));
}
});
return success();
}
private:
DistributionMapFn distributionMapFn;
};
/// A pattern that extracts vector.reduction ops from a WarpExecuteOnLane0Op.
/// The vector is reduced in parallel. Currently limited to vector size matching
/// the warpOp size. E.g.:
/// ```
/// %r = vector_ext.warp_execute_on_lane_0(%laneid)[32] -> (f32) {
/// %0 = "some_def"() : () -> (vector<32xf32>)
/// %1 = vector.reduction "add", %0 : vector<32xf32> into f32
/// vector_ext.yield %1 : f32
/// }
/// ```
/// is lowered to:
/// ```
/// %0 = vector_ext.warp_execute_on_lane_0(%laneid)[32] -> (vector<1xf32>) {
/// %1 = "some_def"() : () -> (vector<32xf32>)
/// vector_ext.yield %1 : vector<32xf32>
/// }
/// %a = vector.extract %0[0] : vector<1xf32>
/// %r = ("warp.reduction %a")
/// ```
struct WarpOpReduction : public OpRewritePattern<WarpExecuteOnLane0Op> {
WarpOpReduction(MLIRContext *context,
DistributedReductionFn distributedReductionFn,
PatternBenefit benefit = 1)
: OpRewritePattern<WarpExecuteOnLane0Op>(context, benefit),
distributedReductionFn(std::move(distributedReductionFn)) {}
LogicalResult matchAndRewrite(WarpExecuteOnLane0Op warpOp,
PatternRewriter &rewriter) const override {
OpOperand *yieldOperand = getWarpResult(
warpOp, [](Operation *op) { return isa<vector::ReductionOp>(op); });
if (!yieldOperand)
return failure();
auto reductionOp =
cast<vector::ReductionOp>(yieldOperand->get().getDefiningOp());
auto vectorType = reductionOp.getVector().getType().cast<VectorType>();
// Only rank 1 vectors supported.
if (vectorType.getRank() != 1)
return rewriter.notifyMatchFailure(
warpOp, "Only rank 1 reductions can be distributed.");
// Only warp_size-sized vectors supported.
if (vectorType.getShape()[0] % warpOp.getWarpSize() != 0)
return rewriter.notifyMatchFailure(
warpOp, "Reduction vector dimension must match was size.");
// Only f32, i32, f16, i8 element types are supported.
if (!reductionOp.getType().isF32() &&
!reductionOp.getType().isSignlessInteger(32) &&
!reductionOp.getType().isF16() && !reductionOp.getType().isInteger(8))
return rewriter.notifyMatchFailure(
warpOp, "Reduction distribution currently only supports 32bits, f16, "
"and i8 types.");
int64_t numElements = vectorType.getShape()[0] / warpOp.getWarpSize();
// Return vector that will be reduced from the WarpExecuteOnLane0Op.
unsigned operandIndex = yieldOperand->getOperandNumber();
SmallVector<Value> yieldValues = {reductionOp.getVector()};
SmallVector<Type> retTypes = {
VectorType::get({numElements}, reductionOp.getType())};
if (reductionOp.getAcc()) {
yieldValues.push_back(reductionOp.getAcc());
retTypes.push_back(reductionOp.getAcc().getType());
}
SmallVector<size_t> newRetIndices;
WarpExecuteOnLane0Op newWarpOp = moveRegionToNewWarpOpAndAppendReturns(
rewriter, warpOp, yieldValues, retTypes, newRetIndices);
rewriter.setInsertionPointAfter(newWarpOp);
// Obtain data to reduce for a single lane.
Value laneValVec = newWarpOp.getResult(newRetIndices[0]);
// Distribute and reduce across threads.
Value fullReduce =
distributedReductionFn(reductionOp.getLoc(), rewriter, laneValVec,
reductionOp.getKind(), newWarpOp.getWarpSize());
if (reductionOp.getAcc()) {
fullReduce = vector::makeArithReduction(
rewriter, reductionOp.getLoc(), reductionOp.getKind(), fullReduce,
newWarpOp.getResult(newRetIndices[1]));
}
newWarpOp.getResult(operandIndex).replaceAllUsesWith(fullReduce);
return success();
}
private:
DistributedReductionFn distributedReductionFn;
};
} // namespace
void mlir::vector::populateWarpExecuteOnLane0OpToScfForPattern(
RewritePatternSet &patterns,
const WarpExecuteOnLane0LoweringOptions &options, PatternBenefit benefit) {
patterns.add<WarpOpToScfIfPattern>(patterns.getContext(), options, benefit);
}
void mlir::vector::populateDistributeTransferWriteOpPatterns(
RewritePatternSet &patterns, const DistributionMapFn &distributionMapFn,
PatternBenefit benefit) {
patterns.add<WarpOpTransferWrite>(patterns.getContext(), distributionMapFn,
benefit);
}
void mlir::vector::populatePropagateWarpVectorDistributionPatterns(
RewritePatternSet &patterns, const DistributionMapFn &distributionMapFn,
const WarpShuffleFromIdxFn &warpShuffleFromIdxFn, PatternBenefit benefit) {
patterns.add<WarpOpElementwise, WarpOpTransferRead, WarpOpDeadResult,
WarpOpBroadcast, WarpOpExtract, WarpOpForwardOperand,
WarpOpConstant>(patterns.getContext(), benefit);
patterns.add<WarpOpExtractElement>(patterns.getContext(),
warpShuffleFromIdxFn, benefit);
patterns.add<WarpOpScfForOp>(patterns.getContext(), distributionMapFn,
benefit);
}
void mlir::vector::populateDistributeReduction(
RewritePatternSet &patterns,
const DistributedReductionFn &distributedReductionFn,
PatternBenefit benefit) {
patterns.add<WarpOpReduction>(patterns.getContext(), distributedReductionFn,
benefit);
}
void mlir::vector::moveScalarUniformCode(WarpExecuteOnLane0Op warpOp) {
Block *body = warpOp.getBody();
// Keep track of the ops we want to hoist.
llvm::SmallSetVector<Operation *, 8> opsToMove;
// Helper to check if a value is or will be defined outside of the region.
auto isDefinedOutsideOfBody = [&](Value value) {
auto *definingOp = value.getDefiningOp();
return (definingOp && opsToMove.count(definingOp)) ||
warpOp.isDefinedOutsideOfRegion(value);
};
// Do not use walk here, as we do not want to go into nested regions and hoist
// operations from there.
for (auto &op : body->without_terminator()) {
bool hasVectorResult = llvm::any_of(op.getResults(), [](Value result) {
return result.getType().isa<VectorType>();
});
if (!hasVectorResult && canBeHoisted(&op, isDefinedOutsideOfBody))
opsToMove.insert(&op);
}
// Move all the ops marked as uniform outside of the region.
for (Operation *op : opsToMove)
op->moveBefore(warpOp);
}