AMDGPUToROCDL.cpp File Reference

#include "mlir/Conversion/AMDGPUToROCDL/AMDGPUToROCDL.h"
#include "mlir/Conversion/LLVMCommon/ConversionTarget.h"
#include "mlir/Conversion/LLVMCommon/Pattern.h"
#include "mlir/Conversion/LLVMCommon/TypeConverter.h"
#include "mlir/Dialect/AMDGPU/IR/AMDGPUDialect.h"
#include "mlir/Dialect/AMDGPU/Utils/Chipset.h"
#include "mlir/Dialect/LLVMIR/LLVMDialect.h"
#include "mlir/Dialect/LLVMIR/ROCDLDialect.h"
#include "mlir/IR/BuiltinTypes.h"
#include "mlir/IR/TypeUtilities.h"
#include "mlir/Pass/Pass.h"
#include "llvm/ADT/STLExtras.h"
#include <optional>
#include "mlir/Conversion/Passes.h.inc"

Go to the source code of this file.

Namespaces
	mlir
	Include the generated interface declarations.

Macros
#define	GEN_PASS_DEF_CONVERTAMDGPUTOROCDL

Functions
static Value	createI32Constant (ConversionPatternRewriter &rewriter, Location loc, int32_t value)
static Value	createI1Constant (ConversionPatternRewriter &rewriter, Location loc, bool value)
static Value	convertMFMAVectorOperand (ConversionPatternRewriter &rewriter, Location loc, Value input)
	Converts a MFMA vector operand from MLIR AMDGPU dialect convention to ROCDL and LLVM AMDGPU intrinsics convention. More...
static void	wmmaPushInputOperand (ConversionPatternRewriter &rewriter, Location loc, const TypeConverter *typeConverter, bool isUnsigned, Value llvmInput, Value mlirInput, SmallVector< Value, 4 > &operands)
	Push an input operand. More...
static void	wmmaPushOutputOperand (ConversionPatternRewriter &rewriter, Location loc, const TypeConverter *typeConverter, Value output, int32_t subwordOffset, bool clamp, SmallVector< Value, 4 > &operands)
	Push the output operand. More...
static std::optional< StringRef >	mfmaOpToIntrinsic (MFMAOp mfma, Chipset chipset)
	Return the rocdl intrinsic corresponding to a MFMA operation mfma if one exists. More...
static std::optional< StringRef >	wmmaOpToIntrinsic (WMMAOp wmma, Chipset chipset)
	Return the rocdl intrinsic corresponding to a WMMA operation wmma if one exists. More...

Macro Definition Documentation

GEN_PASS_DEF_CONVERTAMDGPUTOROCDL

#define GEN_PASS_DEF_CONVERTAMDGPUTOROCDL

Definition at line 26 of file AMDGPUToROCDL.cpp.

Function Documentation

convertMFMAVectorOperand()

static Value convertMFMAVectorOperand ( ConversionPatternRewriter &  rewriter, Location  loc, Value  input  )

static

Converts a MFMA vector operand from MLIR AMDGPU dialect convention to ROCDL and LLVM AMDGPU intrinsics convention.

Specifically:

1. If input is a vector of N bytes, bitcast it to a (N * 8)-bit integer.
2. If the element type is bfloat16, bitcast it to i16.

Definition at line 360 of file AMDGPUToROCDL.cpp.

References mlir::OpBuilder::create(), mlir::Builder::getI16Type(), mlir::Builder::getIntegerType(), and mlir::Value::getType().

createI1Constant()

static Value createI1Constant ( ConversionPatternRewriter &  rewriter, Location  loc, bool  value  )

static

Definition at line 39 of file AMDGPUToROCDL.cpp.

References mlir::OpBuilder::create(), and mlir::Builder::getI1Type().

Referenced by wmmaPushInputOperand(), and wmmaPushOutputOperand().

createI32Constant()

static Value createI32Constant ( ConversionPatternRewriter &  rewriter, Location  loc, int32_t  value  )

static

Definition at line 33 of file AMDGPUToROCDL.cpp.

References mlir::OpBuilder::create(), and mlir::Builder::getI32Type().

mfmaOpToIntrinsic()

static std::optional<StringRef> mfmaOpToIntrinsic ( MFMAOp  mfma, Chipset  chipset  )

static

Return the rocdl intrinsic corresponding to a MFMA operation mfma if one exists.

This includes checking to ensure the intrinsic is supported on the architecture you are compiling for.

Definition at line 452 of file AMDGPUToROCDL.cpp.

wmmaOpToIntrinsic()

static std::optional<StringRef> wmmaOpToIntrinsic ( WMMAOp  wmma, Chipset  chipset  )

static

Return the rocdl intrinsic corresponding to a WMMA operation wmma if one exists.

This includes checking to ensure the intrinsic is supported on the architecture you are compiling for.

Definition at line 587 of file AMDGPUToROCDL.cpp.

wmmaPushInputOperand()

static void wmmaPushInputOperand ( ConversionPatternRewriter &  rewriter, Location  loc, const TypeConverter *  typeConverter, bool  isUnsigned, Value  llvmInput, Value  mlirInput, SmallVector< Value, 4 > &  operands  )

static

Push an input operand.

If it is a float type, nothing to do. If it is an integer type, then we need to also push its signdness (1 for signed, 0 for unsigned) and we need to pack the input 16xi8 vector into a 4xi32 vector. We also need to convert bfloat inputs to i16 to account for the lack of bfloat support in the WMMA intrinsics themselves.

Definition at line 380 of file AMDGPUToROCDL.cpp.

References mlir::TypeConverter::convertType(), mlir::OpBuilder::create(), createI1Constant(), mlir::OpBuilder::createOrFold(), mlir::get(), mlir::Builder::getI16Type(), mlir::Builder::getI32Type(), mlir::Value::getType(), mlir::Type::isBF16(), mlir::Type::isInteger(), mlir::Type::isSignedInteger(), and mlir::Type::isUnsignedInteger().

wmmaPushOutputOperand()

static void wmmaPushOutputOperand ( ConversionPatternRewriter &  rewriter, Location  loc, const TypeConverter *  typeConverter, Value  output, int32_t  subwordOffset, bool  clamp, SmallVector< Value, 4 > &  operands  )

static

Push the output operand.

For many cases this is only pushing the output in the operand list. But when we have f16 -> f16 or bf16 -> bf16 intrinsics, since the same numbers of VGPRs is used, we need to decide if to store the result in the upper 16 bits of the VGPRs or in the lower part. To store the result in the lower 16 bits, set subwordOffset to 1, otherwise result will be stored it in the upper part

Definition at line 430 of file AMDGPUToROCDL.cpp.

References clamp(), mlir::OpBuilder::create(), createI1Constant(), mlir::Builder::getI16Type(), mlir::Value::getType(), mlir::Type::isBF16(), mlir::Type::isF16(), and mlir::Type::isInteger().