Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Brgemm pattern matching #996

Closed
wants to merge 1 commit into from
+813 −58
Closed

Brgemm pattern matching #996

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
48 changes: 48 additions & 0 deletions include/TPP/Dialect/Xsmm/XsmmUtils.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,12 +22,33 @@ class Operation;
class PatternRewriter;
class VectorType;
class MemRefType;
class ModuleOp;

namespace func {
class CallOp;
}

namespace vector {
class ContractionOp;
}

namespace xsmm {

struct BrgemmInfo {
int64_t m;
int64_t n;
int64_t k;
int64_t batch;

int64_t lda;
int64_t ldb;
int64_t ldc;
int64_t strideA;
int64_t strideB;

bool isVnni = false;
};

class UnaryKindAttr;

struct UnaryInfo {
Expand Down Expand Up @@ -89,19 +110,41 @@ FailureOr<UnaryFlags> getUnaryFlags(Type inputType, Type outputType);

// Compute the broadcasting flags for 'operandType' based on 'outputType'.
enum class OperandPos { LHS = 0, RHS = 1 };
FailureOr<BinaryFlags> getBinFlags(ArrayRef<int64_t> shapeOutput,
ArrayRef<int64_t> shapeOperand,
OperandPos operandNumber);
FailureOr<BinaryFlags> getBinaryFlags(Type operandType, Type outputType,
OperandPos operandNumber);
FailureOr<BinaryFlags> getBinaryFlagsVectorType(Type operandType,
Type outputType,
OperandPos operandNumber);

FailureOr<FusedMatch> getFusedBrgemmSequenceFromProducer(Operation *op);

ArrayAttr getUnaryDispatchFlags(UnaryOp op);

ArrayAttr getBinaryDispatchFlags(BinaryOp op);

int64_t getOredFlags(ArrayAttr flags);

func::CallOp buildInvokeCall(RewriterBase &rewriter, Operation *parentOp,
ModuleOp module, SmallVector<Value> inputOperands,
SmallVector<Value> prependValues, int prependIndex,
SmallVector<Value> operands, StringRef invokeName,
DataTypeAttr dtype, bool getResults = false);

template <typename DispatchOpTy>
FailureOr<SmallVector<Attribute>> getBrgemmFlags(PatternRewriter &rewriter,
DispatchOpTy dispatchOpTy,
bool returnNone);
std::optional<unsigned>
getPosInCodomain(unsigned dim, vector::ContractionOp contractOp, AffineMap map);
FailureOr<xsmm::BrgemmInfo> checkAccess(PatternRewriter &rewriter,
vector::ContractionOp contractOp,
unsigned m, unsigned n, unsigned k,
std::optional<unsigned> batchPos,
SmallVector<Value> inputs,
ArrayRef<AffineMap> indexingMap);

SmallVector<Type> extractOperandTypes(OpBuilder &builder,
ArrayRef<Value> operands);
Expand All @@ -122,6 +165,11 @@ func::CallOp buildXsmmCall(RewriterBase &rewriter, XsmmCallType callType,
SmallVector<XsmmOperand> operands, TypeRange results,
FlatSymbolRefAttr fnName, Operation *parentOp,
Operation *insertBefore);
FailureOr<BrgemmInfo> isMappableToBrgemm(PatternRewriter &rewriter,
vector::ContractionOp contractOp,
SmallVector<Value> &inputs,
SmallVector<Value> &output,
ArrayRef<AffineMap> indexingMap);
} // namespace utils
} // namespace xsmm
} // namespace mlir
Expand Down
4 changes: 3 additions & 1 deletion include/TPP/Transforms/Utils/VNNIUtils.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@
#ifndef TPP_TRANSFORMS_UTILS_VNNIUTILS_H
#define TPP_TRANSFORMS_UTILS_VNNIUTILS_H

#include "mlir/Dialect/Vector/IR/VectorOps.h"
#include "mlir/Support/LogicalResult.h"
#include <cstdint>
#include <optional>
Expand Down Expand Up @@ -51,7 +52,8 @@ bool isInVnniLayout(int64_t expectedRank, VectorType vector);
FailureOr<AffineDimExpr> isInVnniLayout(linalg::GenericOp linalgOp,
AffineMap affineMap,
int64_t blockingFactor);

FailureOr<AffineDimExpr> isInVnniLayout(mlir::vector::ContractionOp contractOp,
int64_t blockingFactor);
} // namespace utils
} // namespace vnni
} // namespace mlir
Expand Down
178 changes: 163 additions & 15 deletions lib/TPP/Conversion/ConvertVectorToXsmm/ConvertVectorToXsmm.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,8 @@
//===----------------------------------------------------------------------===//
#include "TPP/Conversion/ConvertVectorToXsmm/ConvertVectorToXsmm.h"
#include "TPP/Dialect/Xsmm/XsmmUtils.h"
#include "TPP/Transforms/Transforms.h"
#include "TPP/Transforms/Utils/TransformUtils.h"
#include "TPP/Transforms/Utils/VNNIUtils.h"
#include "TPP/Transforms/Utils/ValueUtils.h"
#include "mlir/Dialect/Func/IR/FuncOps.h"
Expand Down Expand Up @@ -68,9 +70,9 @@ getUnaryXSMMCalls(PatternRewriter &rewriter, xsmm::UnaryInfo &unaryInfo,
}

static std::pair<Operation *, Operation *>
convertTransposeOp(PatternRewriter &rewriter, Operation *transposeOp,
Operation *input, Operation *output, Type outputType) {
LLVM_DEBUG(llvm::dbgs() << "convertTransposeOp\n");
convertTranspose(PatternRewriter &rewriter, Operation *transposeOp,
Operation *input, Operation *output, Type outputType) {
LLVM_DEBUG(llvm::dbgs() << "convertTranspose\n");
VectorType outType = cast<VectorType>(outputType);
xsmm::UnaryKind opType;
xsmm::UnaryInfo unaryInfo;
Expand Down Expand Up @@ -111,11 +113,10 @@ convertTransposeOp(PatternRewriter &rewriter, Operation *transposeOp,
output, transposeOp, unaryFlags.getInt());
}

static LogicalResult validateTransposeOp(PatternRewriter &rewriter,
Operation *transposeOp,
Operation *input, Operation *output,
Type outputType) {
LLVM_DEBUG(llvm::dbgs() << "validateTransposeOp\n");
static LogicalResult validateTranspose(PatternRewriter &rewriter,
Operation *transposeOp, Operation *input,
Operation *output, Type outputType) {
LLVM_DEBUG(llvm::dbgs() << "validateTranspose\n");
Value result = input->getResult(0);
Value source = input->getOperand(0);
VectorType outType = cast<VectorType>(outputType);
Expand Down Expand Up @@ -187,10 +188,10 @@ convertBroadcast(PatternRewriter &rewriter, Operation *broadcastOp,
broadcastOp, unaryFlags.getInt());
}

static LogicalResult validateBroadcastOp(PatternRewriter &rewriter,
Operation *broadcastOp,
Operation *input, Operation *output) {
LLVM_DEBUG(llvm::dbgs() << "validateBroadcastOp\n");
static LogicalResult validateBroadcast(PatternRewriter &rewriter,
Operation *broadcastOp, Operation *input,
Operation *output) {
LLVM_DEBUG(llvm::dbgs() << "validateBroadcast\n");
auto unaryFlag = xsmm::utils::getUnaryFlags(input->getOperand(0).getType(),
output->getOperand(1).getType());
auto inputMemRefType = dyn_cast<MemRefType>(input->getOperand(0).getType());
Expand All @@ -205,15 +206,162 @@ static LogicalResult validateBroadcastOp(PatternRewriter &rewriter,
return success();
}

FailureOr<xsmm::BrgemmInfo>
computeBrgemmInfo(PatternRewriter &rewriter, Operation *contractOp,
Operation *input0, Operation *input1, Operation *input2) {
SmallVector<Value> inputs;
LLVM_DEBUG(llvm::dbgs() << "computeBrgemminfo\n");

inputs.push_back(input0->getResult(0));
inputs.push_back(input1->getResult(0));
inputs.push_back(input2->getResult(0));

SmallVector<Value> outputs;
outputs.push_back(nullptr);
auto failedOrbrgemmInfo = mlir::xsmm::utils::isMappableToBrgemm(
rewriter, dyn_cast<mlir::vector::ContractionOp>(contractOp), inputs,
outputs,
dyn_cast<mlir::vector::ContractionOp>(contractOp).getIndexingMapsArray());
if (failed(failedOrbrgemmInfo))
return failure();
xsmm::BrgemmInfo brgemmInfo = *failedOrbrgemmInfo;
return brgemmInfo;
}

static std::pair<Operation *, Operation *>
createBrgemmImpl(PatternRewriter &rewriter, Operation *contractOp, Value input0,
Value input1, Value input2, xsmm::BrgemmInfo brgemmInfo,
SmallVector<Attribute> flags) {
SmallVector<Value> inputs;
inputs.push_back(input0);
inputs.push_back(input1);
inputs.push_back(input2);
auto m = brgemmInfo.m;
auto n = brgemmInfo.n;
auto k = brgemmInfo.k;
auto batch = brgemmInfo.batch;
int64_t lda = brgemmInfo.lda;
int64_t ldb = brgemmInfo.ldb;
int64_t ldc = brgemmInfo.ldc;
int64_t strideA = brgemmInfo.strideA;
int64_t strideB = brgemmInfo.strideB;
auto loc = contractOp->getLoc();
auto dtype =
xsmm::utils::getDataType(rewriter, contractOp->getOperand(0).getType());
SmallVector<xsmm::utils::XsmmOperand> dispatchOperands;
// Dispatch the data type.
dispatchOperands.push_back(dyn_cast<DataTypeAttr>(dtype).getInt());

ArrayAttr brgemmFlags = rewriter.getArrayAttr(flags);
SmallVector<Value, 10> invokeOperands;
std::string dispatchName = "xsmm_gemm_dispatch";
std::string invokeName = "xsmm_gemm_invoke";

if (batch != 0) {
dispatchName = "xsmm_brgemm_dispatch";
invokeName = "xsmm_brgemm_invoke";
}

dispatchOperands.append(
SmallVector<xsmm::utils::XsmmOperand>{m, n, k, lda, ldb, ldc});
if (batch != 0) {
dispatchOperands.push_back(strideA);
dispatchOperands.push_back(strideB);
}
int64_t oredFlag = xsmm::utils::getOredFlags(brgemmFlags);
dispatchOperands.push_back(oredFlag);

auto dispatchCall = xsmm::utils::buildXsmmCall(
rewriter, xsmm::utils::XsmmCallType::DISPATCH, loc, dtype,
dispatchOperands, IntegerType::get(rewriter.getContext(), 64),
SymbolRefAttr::get(contractOp->getContext(), dispatchName), contractOp,
nullptr);

SmallVector<xsmm::utils::XsmmOperand> operandRange{
dyn_cast<DataTypeAttr>(dtype).getInt(),
xsmm::utils::XsmmCall{xsmm::utils::XsmmCallType::DISPATCH,
dispatchCall.getResult(0)},
input0.getDefiningOp()->getOperand(0),
input1.getDefiningOp()->getOperand(0),
input2.getDefiningOp()->getOperand(0)};

if (batch != 0) {
operandRange.push_back(batch);
}
auto invokeCall = xsmm::utils::buildXsmmCall(
rewriter, xsmm::utils::XsmmCallType::INVOKE, loc, dtype, operandRange,
TypeRange(), SymbolRefAttr::get(contractOp->getContext(), invokeName),
contractOp, input2.getDefiningOp());
return std::make_pair(&*dispatchCall, &*invokeCall);
}

static std::pair<Operation *, Operation *>
createBrgemmWithBetaZero(PatternRewriter &rewriter, Operation *contractOp,
Operation *input0, Operation *input1,
Operation *input2, Operation *betaZero) {
LLVM_DEBUG(llvm::dbgs() << "createBrgemmWithBetaZero\n");
auto brgemmInfo =
computeBrgemmInfo(rewriter, contractOp, input0, input1, input2);
SmallVector<Attribute> flags;
if (brgemmInfo->isVnni) {
flags.push_back(xsmm::GemmFlagsAttr::get(rewriter.getContext(),
xsmm::GemmFlags::VNNI_B));
}
flags.push_back(
xsmm::GemmFlagsAttr::get(rewriter.getContext(), xsmm::GemmFlags::BETA_0));

return createBrgemmImpl(rewriter, contractOp, input0->getResult(0),
input1->getResult(0), input2->getResult(0),
*brgemmInfo, flags);
}

static std::pair<Operation *, Operation *>
createBrgemm(PatternRewriter &rewriter, Operation *contractOp,
Operation *input0, Operation *input1, Operation *input2) {
LLVM_DEBUG(llvm::dbgs() << "createBrgemm\n");
FailureOr<mlir::xsmm::BrgemmInfo> brgemmInfo;
brgemmInfo = computeBrgemmInfo(rewriter, contractOp, input0, input1, input2);
SmallVector<Attribute> flags;
if (brgemmInfo->isVnni) {
flags.push_back(xsmm::GemmFlagsAttr::get(rewriter.getContext(),
xsmm::GemmFlags::VNNI_B));
}

return createBrgemmImpl(rewriter, contractOp, input0->getResult(0),
input1->getResult(0), input2->getResult(0),
*brgemmInfo, flags);
}

static LogicalResult validateBrgemm(PatternRewriter &rewriter,
Operation *contractOp, Operation *input0,
Operation *input1, Operation *input2,
Operation *result) {
LLVM_DEBUG(llvm::dbgs() << "validateBrgemm\n");
FailureOr<xsmm::BrgemmInfo> brgemmInfo =
computeBrgemmInfo(rewriter, contractOp, input0, input1, input2);

if (failed(brgemmInfo)) {
return failure(contractOp);
}

return success(contractOp);
}

static void registerNativeRewrite(RewritePatternSet &patterns) {
patterns.getPDLPatterns().registerRewriteFunction("ConvertTranspose",
convertTransposeOp);
convertTranspose);
patterns.getPDLPatterns().registerConstraintFunction("ValidateTranspose",
validateTransposeOp);
validateTranspose);
patterns.getPDLPatterns().registerRewriteFunction("ConvertBroadcast",
convertBroadcast);
patterns.getPDLPatterns().registerConstraintFunction("ValidateBroadcast",
validateBroadcastOp);
validateBroadcast);
patterns.getPDLPatterns().registerRewriteFunction("CreateBrgemmWithBetaZero",
createBrgemmWithBetaZero);
patterns.getPDLPatterns().registerRewriteFunction("CreateBrgemm",
createBrgemm);
patterns.getPDLPatterns().registerConstraintFunction("ValidateBrgemm",
validateBrgemm);
}

namespace mlir {
Expand Down
Loading