MLIR
15.0.0git
|
Namespaces | |
amd | |
Classes | |
class | AsyncTokenType |
struct | KernelDim3 |
Utility class for the GPU dialect to represent triples of Value s accessible through .x , .y , and .z similarly to CUDA notation. More... | |
struct | MMAMatrixStorageType |
MMAMatrixType storage and uniquing. More... | |
class | MMAMatrixType |
MMAMatrix represents a matrix held by a subgroup for matrix-matrix multiply accumulate operations. More... | |
class | SerializeToBlobPass |
Base pass class to serialize kernel functions through LLVM into user-specified IR and add the resulting blob as module attribute. More... | |
Functions | |
void | addAsyncDependency (Operation *op, Value token) |
StringRef | getMappingAttrName () |
Name of the mapping attribute produced by loop mappers. More... | |
Processor | getProcessor (ParallelLoopDimMapping attr) |
Get the value of the processor in the ParallelLoopDimMapping attribute. More... | |
ParallelLoopDimMapping | getParallelLoopDimMappingAttr (Processor processor, AffineMap map, AffineMap bound) |
Helper function to create a ParallelDimMapperAttr. More... | |
LogicalResult | setMappingAttr (scf::ParallelOp ploopOp, ArrayRef< ParallelLoopDimMapping > mapping) |
Sets the mapping attribute of a scf.parallel operation. More... | |
std::string | getDefaultGpuBinaryAnnotation () |
Returns the default annotation name for GPU binary blobs. More... | |
Definition at line 371 of file GPUDialect.cpp.
References mlir::Block::addArgument(), mlir::OperationState::addAttribute(), mlir::OperationState::addOperands(), mlir::OperationState::addRegion(), mlir::Operation::getContext(), mlir::Builder::getI32VectorAttr(), mlir::Builder::getIndexType(), mlir::OpTrait::AttrSizedOperandSegments< ConcreteType >::getOperandSegmentSizeAttr(), mlir::Builder::getType(), mlir::Operation::insertOperands(), mlir::OperationState::location, mlir::Region::push_back(), mlir::Operation::setAttr(), mlir::success(), and mlir::OperationState::types.
std::string mlir::gpu::getDefaultGpuBinaryAnnotation | ( | ) |
Returns the default annotation name for GPU binary blobs.
Definition at line 30 of file SerializeToBlob.cpp.
Referenced by mlir::populateGpuRewritePatterns().
StringRef mlir::gpu::getMappingAttrName | ( | ) |
Name of the mapping attribute produced by loop mappers.
Definition at line 31 of file ParallelLoopMapper.cpp.
Referenced by mlir::configureParallelLoopToGPULegality(), mapParallelOp(), processParallelLoop(), and setMappingAttr().
ParallelLoopDimMapping mlir::gpu::getParallelLoopDimMappingAttr | ( | Processor | processor, |
AffineMap | map, | ||
AffineMap | bound | ||
) |
Helper function to create a ParallelDimMapperAttr.
TODO: Replace its uses with an auto-gened method.
Definition at line 33 of file ParallelLoopMapper.cpp.
References mlir::Builder::getContext(), and mlir::AffineMap::getContext().
Referenced by getProcessor(), and mapParallelOp().
|
inline |
Get the value of the processor in the ParallelLoopDimMapping attribute.
Definition at line 45 of file ParallelLoopMapper.h.
References getParallelLoopDimMappingAttr(), mlir::greedilyMapParallelSCFToGPU(), and setMappingAttr().
Referenced by processParallelLoop(), and setMappingAttr().
LogicalResult mlir::gpu::setMappingAttr | ( | scf::ParallelOp | ploopOp, |
ArrayRef< ParallelLoopDimMapping > | mapping | ||
) |
Sets the mapping attribute of a scf.parallel operation.
Verifies that the mapping passed is valid.
ploopOp
.Definition at line 43 of file ParallelLoopMapper.cpp.
References getMappingAttrName(), getProcessor(), and mlir::success().
Referenced by getProcessor(), and mapParallelOp().