MCPcopy Create free account

hub / github.com/NVIDIA/cuda-tile / functions

Functions974 in github.com/NVIDIA/cuda-tile

MethodinferReturnTypes
lib/Dialect/CudaTile/IR/CudaTile.cpp:4943
Methodinfo
tools/cuda-tile-tblgen/Emitter.h:168
MethodinfoLine
tools/cuda-tile-tblgen/Emitter.h:172
Methodinitialize
This method initializes the debug info reader after construction.
lib/Bytecode/Reader/BytecodeReader.cpp:832
Methodinitialize
lib/Dialect/CudaTile/IR/CudaTile.cpp:5643
Methodint_tile
tools/cuda-tile-tblgen/CudaTileType.cpp:156
Functionint_to_ptr
(el_type, src: Tile, *, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:2335
MethodinvalidLocError
lib/Bytecode/Writer/BytecodeWriter.cpp:751
Functioniota
(n: int, el_type, *, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:3647
MethodisLegalToInline
lib/Dialect/CudaTile/IR/CudaTile.cpp:5545
MethodisOpcodeAvailableInVersion
lib/Bytecode/Common/Version.cpp:70
MethodisPointerLike
lib/Dialect/CudaTile/IR/Types.cpp:40
MethodisTileIRBytecode
lib/Bytecode/Reader/BytecodeReader.cpp:42
Functionitof
( el_type, from_, *, rounding_mode: RoundingMode = RoundingMode.NEAREST_EVEN, signedness:
python/cuda_tile/dialects/cuda_tile_ops.py:3692
Functionjoin_tokens
Join multiple tokens into a single token. Args: *tokens: Variable number of Token objects to join loc: Source location ip
python/cuda_tile/dialects/cuda_tile_ops.py:3328
Functionload_ptr_tko
Load data from memory with specified ordering and optional masking. Loads data from the given source pointer(s) using the specified memory sy
python/cuda_tile/dialects/cuda_tile_ops.py:3048
Functionload_view_tko
Load data from a tile view with specified memory ordering and scope.
python/cuda_tile/dialects/cuda_tile_ops.py:3192
Methodlocation
Gets an unknown location for the CUDA Tile IR.
python/cuda_tile/dialects/cuda_tile_ops.py:5094
Functionlog
(source: Tile, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:4809
Functionlog10
(source: Tile, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:4816
Functionlog1p
(source: Tile, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:4823
Functionlog2
(source: Tile, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:4830
Functionloop_break
( operands: Union[Tile, Token, Iterable[Union[Tile, Token]]], *, loc=None, ip=None )
python/cuda_tile/dialects/cuda_tile_ops.py:3905
Functionloop_continue
( operands: Union[Tile, Token, Iterable[Union[Tile, Token]]], *, loc=None, ip=None )
python/cuda_tile/dialects/cuda_tile_ops.py:3921
Functionloop_generate
( inputs: Iterable[Union[Tile, Token]], loop_body: Callable, *, loc=None, ip=None, )
python/cuda_tile/dialects/cuda_tile_ops.py:3872
Functionmain
tools/cuda-tile-opt/cuda-tile-opt.cpp:25
Functionmain
tools/cuda-tile-optimize/cuda-tile-optimize.cpp:145
Functionmain
tools/cuda-tile-tblgen/cuda-tile-tblgen.cpp:28
Functionmain
tools/cuda-tile-translate/cuda-tile-translate.cpp:18
Functionmain
test/CAPI/register.c:13
Functionmake_partition_view
( tensor_view: TensorView, tile_shape: List[int], *, dim_map: List[int] | None = None, pad
python/cuda_tile/dialects/cuda_tile_ops.py:4895
Functionmake_strided_view
( tensor_view: TensorView, tile_shape: List[int], traversal_strides: List[int], *, dim_map
python/cuda_tile/dialects/cuda_tile_ops.py:4917
Functionmake_tensor_view
( base_ptr, el_type, shape: List[int | Tile] | None = None, strides: List[int | Tile] | None =
python/cuda_tile/dialects/cuda_tile_ops.py:4050
Functionmake_token
(*, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:3323
MethodmatchAndRewrite
lib/Dialect/CudaTile/Transforms/FuseFMA.cpp:26
MethodmatchAndRewrite
lib/Dialect/CudaTile/Transforms/FuseFMA.cpp:61
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:2968
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:2993
MethodmatchAndRewrite
Porting of SCF::IfOp::RemoveUnusedResults Additional support for non-YieldOp terminator inside transferBody()
lib/Dialect/CudaTile/IR/CudaTile.cpp:3105
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:3158
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:3218
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:3368
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:3392
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:3507
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:5111
MethodmatchAndRewrite
lib/Dialect/CudaTile/IR/CudaTile.cpp:5155
Methodmatches_types
Check if the given MLIR types match this configuration
python/cuda_tile/dialects/cuda_tile_ops.py:720
FunctionmlirCudaTileOperationIsAModuleOp
lib/CAPI/Dialect/CudaTileDialect.cpp:654
FunctionmlirCudaTileOptimizationHintsAttrGetEmpty
lib/CAPI/Dialect/CudaTileDialect.cpp:580
FunctionmlirCudaTilePartitionViewTypeGet
lib/CAPI/Dialect/CudaTileDialect.cpp:179
FunctionmlirCudaTilePartitionViewTypeGetTypeID
lib/CAPI/Dialect/CudaTileDialect.cpp:175
FunctionmlirCudaTilePointerTypeGetTypeID
lib/CAPI/Dialect/CudaTileDialect.cpp:50
FunctionmlirCudaTileRegisterCSEPass
lib/CAPI/Dialect/CudaTileDialect.cpp:740
FunctionmlirCudaTileRegisterCanonicalizerPass
lib/CAPI/Dialect/CudaTileDialect.cpp:736
FunctionmlirCudaTileRegisterFuseFMAPass
lib/CAPI/Dialect/CudaTileDialect.cpp:732
FunctionmlirCudaTileRegisterLoopSplitPass
lib/CAPI/Dialect/CudaTileDialect.cpp:734
FunctionmlirCudaTileRegisterSynthesizeDebugInfoScopesPass
lib/CAPI/Dialect/CudaTileDialect.cpp:728
FunctionmlirCudaTileStridedViewTypeGet
lib/CAPI/Dialect/CudaTileDialect.cpp:249
FunctionmlirCudaTileStridedViewTypeGetTypeID
lib/CAPI/Dialect/CudaTileDialect.cpp:245
FunctionmlirCudaTileTensorViewTypeGet
lib/CAPI/Dialect/CudaTileDialect.cpp:128
FunctionmlirCudaTileTensorViewTypeGetTypeID
lib/CAPI/Dialect/CudaTileDialect.cpp:124
FunctionmlirCudaTileTileTypeGet
lib/CAPI/Dialect/CudaTileDialect.cpp:74
FunctionmlirCudaTileTileTypeGetTypeID
lib/CAPI/Dialect/CudaTileDialect.cpp:70
FunctionmlirCudaTileTokenTypeGetTypeID
lib/CAPI/Dialect/CudaTileDialect.cpp:108
FunctionmlirOperationIsAModuleOp
lib/CAPI/Dialect/CudaTileDialect.cpp:658
Functionmlir_context
Create an MLIR context for tests that need types.
test/python/test_typing.py:16
Methodmlir_type
(cls)
python/cuda_tile/dialects/cuda_tile_ops.py:35
Functionmma
Computes the mma product of lhs and rhs.
python/cuda_tile/dialects/cuda_tile_ops.py:2751
Functionmmaf_scaled
Computes scaled matrix-multiply-accumulate for low-precision types.
python/cuda_tile/dialects/cuda_tile_ops.py:2847
Functionmulhii
(lhs: Tile, rhs: Tile, *, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:3831
Methodnext
lib/Bytecode/Reader/BytecodeReader.cpp:796
Methodnum_elements
(self)
python/cuda_tile/dialects/cuda_tile_ops.py:898
Methodnumeric_tile
tools/cuda-tile-tblgen/CudaTileType.cpp:178
Methodoperator!=
include/cuda_tile/Bytecode/Common/Version.h:48
Methodoperator<
tools/cuda-tile-tblgen/BytecodeGenUtilities.cpp:139
Methodoperator<
include/cuda_tile/Bytecode/Common/Version.h:51
Methodoperator<=
include/cuda_tile/Bytecode/Common/Version.h:58
Methodoperator==
Various comparison operators for comparing versions.
include/cuda_tile/Bytecode/Common/Version.h:44
Methodoperator>
include/cuda_tile/Bytecode/Common/Version.h:61
Methodoperator>=
include/cuda_tile/Bytecode/Common/Version.h:62
Functionoptimization_barrier
( value: Tile, keep_axis_info: bool = False, *, loc=None, ip=None )
python/cuda_tile/dialects/cuda_tile_ops.py:4272
Functionpack
Pack a numeric tile into an i8 byte array, preserving bit values.
python/cuda_tile/dialects/cuda_tile_ops.py:4953
Methodparse
lib/Bytecode/Common/CommandLineOptions.cpp:28
Methodparse
lib/Dialect/CudaTile/IR/Attributes.cpp:369
Methodparse
lib/Dialect/CudaTile/IR/CudaTile.cpp:1905
Methodparse
lib/Dialect/CudaTile/IR/CudaTileTesting.cpp:26
MethodparseAPInt
Parses an APInt from the bytecode stream.
lib/Bytecode/Reader/BytecodeReader.cpp:1296
FunctionparseArgumentRegion
lib/Dialect/CudaTile/IR/CudaTile.cpp:1046
FunctionparseAssumePredicate
lib/Dialect/CudaTile/IR/CudaTile.cpp:938
MethodparseBlock
Helper function to parse a given block during deserialization.
lib/Bytecode/Reader/BytecodeReader.cpp:1196
MethodparseBytecodeTestValueAttr
Parses a BytecodeTestValueAttr attribute.
lib/Bytecode/Reader/BytecodeReader.cpp:1461
MethodparseCallSite
call-site =: DebugTag[CallSite] diCalleeIndex[varint] - LocationAttr diCallerIndex[varint] - LocationAttr
lib/Bytecode/Reader/BytecodeReader.cpp:1063
FunctionparseComparisonOrdering
lib/Dialect/CudaTile/IR/CudaTile.cpp:694
FunctionparseComparisonPredicate
lib/Dialect/CudaTile/IR/CudaTile.cpp:666
MethodparseConstantAttrIndex
Parses a DenseElementsAttr (reads an index into the constant pool). `expectedType` is the MLIR Type of the constant (e.g., TileType).
lib/Bytecode/Reader/BytecodeReader.cpp:1403
MethodparseCudaTileType
Parse a type, if type is unprefixed, assume it is from the cuda_tile dialect
lib/Dialect/CudaTile/IR/Types.cpp:90
MethodparseCudaTileTypeSplat
lib/Dialect/CudaTile/IR/Types.cpp:125
MethodparseDICompileUnit
di-compile-unit =: DebugTag[DICompileUnit] diFileIndex[varint] - DIFileAttr
lib/Bytecode/Reader/BytecodeReader.cpp:914
MethodparseDIFile
di-file =: DebugTag[DIFile] fileNameIndex[varint] - StringAttr directoryIndex[varint] - StringAttr
lib/Bytecode/Reader/BytecodeReader.cpp:929
MethodparseDILexicalBlock
di-lexical-block =: DebugTag[DILexicalBlock] diScopeIndex[varint] - DILocalScopeAttr diFileIndex[varint] - DIFileAttr lineNumber[varint] - unsigned co
lib/Bytecode/Reader/BytecodeReader.cpp:952
← previousnext →701–800 of 974, ranked by callers