MCPcopy Create free account

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

Functions974 in github.com/NVIDIA/cuda-tile

↓ 275 callersMethodsize
lib/Bytecode/Reader/BytecodeReader.cpp:530
↓ 266 callersMethodemitError
Emits an error message associated with the current reader offset. TODO: Generate a location based on the current offset instead of UnknownLoc.
lib/Bytecode/Reader/BytecodeReader.cpp:266
↓ 154 callersMethodgetType
lib/Bytecode/Reader/BytecodeReader.cpp:493
↓ 101 callersMethodget
lib/Dialect/CudaTile/IR/Types.cpp:171
↓ 95 callersFunctionreturn_results
Return op results as Tile(s), Token, or (Tile, Token) depending on context. - If the op has 1 result and it's a Token -> return Token -
python/cuda_tile/dialects/cuda_tile_ops.py:1270
↓ 77 callersMethodwriteVarInt
lib/Bytecode/Writer/BytecodeWriter.cpp:73
↓ 48 callersMethodreadVarInt
lib/Bytecode/Reader/BytecodeReader.cpp:94
↓ 31 callersMethodgetOperationName
tools/cuda-tile-tblgen/CudaTileOp.h:169
↓ 27 callersMethodisInteger
tools/cuda-tile-tblgen/BytecodeTypeAnalysis.h:124
↓ 27 callersFunctionmax
( lhs: Tile, rhs: Tile, *, propagate_nan: bool = False, flush_to_zero: bool = False, s
python/cuda_tile/dialects/cuda_tile_ops.py:4159
↓ 25 callersMethod__init__
(self, value: _ods_ir.Value, type: _ods_ir.Type)
python/cuda_tile/dialects/cuda_tile_ops.py:881
↓ 19 callersFunctioncat
Concatenates lhs and rhs along the specified dimension.
python/cuda_tile/dialects/cuda_tile_ops.py:2714
↓ 19 callersFunctionmake_tile_type
Create a TileType with a specified element type and shape. Args: el_type: Element type - can be a type wrapper (Int32, Float32, etc.) or
python/cuda_tile/dialects/cuda_tile_ops.py:1404
↓ 19 callersMethodtoString
tools/cuda-tile-tblgen/CudaTileType.cpp:258
↓ 17 callersMethodcurrentOffset
lib/Bytecode/Reader/BytecodeReader.cpp:249
↓ 16 callersMethodwrite
lib/Bytecode/Writer/BytecodeWriter.cpp:135
↓ 15 callersFunction_get_mlir_type
Extract MLIR type from element type wrapper or return as-is if already MLIR type.
python/cuda_tile/dialects/cuda_tile_ops.py:111
↓ 15 callersMethodgetElseBlock
lib/Dialect/CudaTile/IR/CudaTile.cpp:2879
↓ 15 callersMethodremaining
lib/Bytecode/Reader/BytecodeReader.cpp:197
↓ 13 callersMethodgetElseTerminator
lib/Dialect/CudaTile/IR/CudaTile.cpp:2883
↓ 13 callersMethodgetThenTerminator
lib/Dialect/CudaTile/IR/CudaTile.cpp:2877
↓ 13 callersFunctionget_rounding_mode_attr
Convert an enum value to the corresponding RoundingModeAttr. Args: mode: RoundingMode enum value context: Optional MLIR cont
python/cuda_tile/dialects/cuda_tile_ops.py:385
↓ 13 callersFunctionget_signedness_attr
Convert an enum value to the corresponding SignednessAttr.
python/cuda_tile/dialects/cuda_tile_ops.py:426
↓ 11 callersFunctionemitConfigError
Utility: emit error and return failure().
lib/Dialect/CudaTile/Optimizer/CudaTileOptimizer.cpp:142
↓ 11 callersMethodgetStringIndex
lib/Bytecode/Writer/BytecodeWriter.cpp:246
↓ 10 callersFunctionisTerminatorForParent
Return True if Terminator is ContinueOp/ReturnOp/BreakOp, so no operation from parent region will be executed after it Return False if Terminator is Y
lib/Dialect/CudaTile/IR/CudaTile.cpp:2891
↓ 10 callersMethodparse
lib/Dialect/CudaTile/IR/Types.cpp:176
↓ 10 callersMethodreadLE
lib/Bytecode/Reader/BytecodeReader.cpp:127
↓ 9 callersMethodgetThenBlock
lib/Dialect/CudaTile/IR/CudaTile.cpp:2876
↓ 9 callersMethodisValid
lib/Dialect/CudaTile/IR/Attributes.cpp:51
↓ 8 callersMethodalignTo
lib/Bytecode/Writer/BytecodeWriter.cpp:143
↓ 8 callersFunctionemitDiagnostic
lib/Dialect/CudaTile/IR/Attributes.cpp:56
↓ 8 callersMethodreadAndGetString
Reads a string index and returns the corresponding StringRef.
lib/Bytecode/Reader/BytecodeReader.cpp:237
↓ 8 callersMethodtell
lib/Bytecode/Writer/BytecodeWriter.cpp:141
↓ 8 callersFunctionverifyFtz
include/cuda_tile/Dialect/CudaTile/IR/SharedVerifiers.h:96
↓ 8 callersMethodwriteByte
lib/Bytecode/Writer/BytecodeWriter.cpp:66
↓ 7 callersMethodgetCurrentPtr
lib/Bytecode/Reader/BytecodeReader.cpp:216
↓ 7 callersMethodgetTypeDescription
tools/cuda-tile-tblgen/CudaTileOp.cpp:96
↓ 7 callersMethodgetViewIndexRank
lib/Dialect/CudaTile/IR/Types.cpp:623
↓ 7 callersFunctionget_memory_ordering_semantics_attr
Convert an enum value to the corresponding MemoryOrderingSemanticsAttr. Args: semantics: MemoryOrderingSemantics enum value
python/cuda_tile/dialects/cuda_tile_ops.py:367
↓ 7 callersFunctionget_memory_scope_attr
Convert an enum value to the corresponding MemoryScopeAttr. Args: scope: MemoryScope enum value context: Optional MLIR conte
python/cuda_tile/dialects/cuda_tile_ops.py:329
↓ 7 callersMethodreadSignedVarInt
Parse a signed variable length encoded integer from the byte stream. A signed varint is encoded as a normal varint with zigzag encoding applied, i.e.
lib/Bytecode/Reader/BytecodeReader.cpp:118
↓ 7 callersMethodskip
lib/Bytecode/Reader/BytecodeReader.cpp:190
↓ 7 callersMethodskipPadding
lib/Bytecode/Reader/BytecodeReader.cpp:251
↓ 7 callersFunctionwriteSectionHeader
lib/Bytecode/Writer/BytecodeWriter.cpp:199
↓ 7 callersMethodwriteSignedVarInt
Emit a signed variable length integer. Signed varints are encoded using a varint with zigzag encoding, meaning that we use the low bit of the value to
lib/Bytecode/Writer/BytecodeWriter.cpp:97
↓ 6 callersMethodemitDescription
tools/cuda-tile-tblgen/Emitter.h:253
↓ 6 callersMethodgetRequiredAlignment
lib/Bytecode/Writer/BytecodeWriter.cpp:154
↓ 6 callersFunctionisConstantTrueVal
lib/Dialect/CudaTile/IR/CudaTile.cpp:1324
↓ 6 callersFunctionparseRoundingModeWithModes
lib/Dialect/CudaTile/IR/CudaTile.cpp:727
↓ 6 callersMethodwriteTypeIndex
Helper function to write the index of a given type to the writer.
lib/Bytecode/Writer/BytecodeWriter.cpp:388
↓ 5 callersFunctionconstant
Helper function that builds a cuda_tile.constant op for the given value, which is either a scalar (integer/float) or a Python list. Nested li
python/cuda_tile/dialects/cuda_tile_ops.py:4640
↓ 5 callersMethodemitAnchor
tools/cuda-tile-tblgen/Emitter.h:268
↓ 5 callersMethodgetAttributes
tools/cuda-tile-tblgen/CudaTileOp.cpp:505
↓ 5 callersFunctiongetConstantBoolValue
Try to get constant bool defined by given Value tile<i1> or tile<...xi1> is expected for defining ConstantOp
lib/Dialect/CudaTile/IR/CudaTile.cpp:1310
↓ 5 callersMethodgetViewTileType
lib/Dialect/CudaTile/IR/Types.cpp:627
↓ 5 callersMethodinheritStringTableFrom
lib/Bytecode/Reader/BytecodeReader.cpp:271
↓ 5 callersMethodreadAndGetType
Reads a type index using the provided reader and retrieves the corresponding Type. Emits an error and returns a null Type on failure.
lib/Bytecode/Reader/BytecodeReader.cpp:537
↓ 5 callersMethodreadBytes
lib/Bytecode/Reader/BytecodeReader.cpp:199
↓ 4 callersFunctiongenerateParameterDeserialization
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:253
↓ 4 callersFunctiongenerateVersionCheck
Generate version check with proper indentation.
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:176
↓ 4 callersFunctiongetOperationName
tools/cuda-tile-tblgen/SpecGen.cpp:71
↓ 4 callersMethodgetParamOrResult
tools/cuda-tile-tblgen/CudaTileOp.h:145
↓ 4 callersFunctiongetType
tools/cuda-tile-tblgen/CudaTileType.cpp:448
↓ 4 callersMethodisFloat
tools/cuda-tile-tblgen/BytecodeTypeAnalysis.h:125
↓ 4 callersMethodparseRegion
Helper function to parse a region during deserialization.
lib/Bytecode/Reader/BytecodeReader.cpp:1263
↓ 4 callersFunctionprintFunctionSignatureWithCudaTileTypes
Print function signature with cuda_tile dialect type support.
lib/Dialect/CudaTile/IR/CudaTile.cpp:225
↓ 4 callersFunctionverifyIEEERoundingModes
lib/Dialect/CudaTile/IR/CudaTile.cpp:1409
↓ 4 callersFunctionverifyOptHintsCommon
include/cuda_tile/Dialect/CudaTile/IR/SharedVerifiers.h:30
↓ 3 callersFunction_check_is_rhs_tile
To allow mixing of Python values and SSA values, we generate an MLIR value using `constant` for the RHS, matching the type of the LHS tile.
python/cuda_tile/dialects/cuda_tile_ops.py:1707
↓ 3 callersFunction_flatten_constants
Helper function for cuda_tile.constant and cuda_tile.global that flattens values and determines the shape.
python/cuda_tile/dialects/cuda_tile_ops.py:4564
↓ 3 callersFunction_index_list_to_tiles
Ensures all tiles in index are scalar integer tiles of the same type, and converts constant indices to tiles of that type.
python/cuda_tile/dialects/cuda_tile_ops.py:1230
↓ 3 callersFunctioncheck_scalar
(x: int | Tile, name: str)
python/cuda_tile/dialects/cuda_tile_ops.py:3968
↓ 3 callersFunctioncleanName
tools/cuda-tile-tblgen/CudaTileAttr.cpp:31
↓ 3 callersFunctionextractFileLoc
Attempt to extract a filename for the given loc.
lib/Dialect/CudaTile/Transforms/SynthesizeDebugInfoScopes.cpp:27
↓ 3 callersFunctiongenerateParameterSerialization
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:212
↓ 3 callersMethodgetAnchor
tools/cuda-tile-tblgen/CudaTileAttr.cpp:44
↓ 3 callersMethodgetDescription
tools/cuda-tile-tblgen/CudaTileOp.cpp:85
↓ 3 callersMethodgetIterator
lib/Bytecode/Reader/BytecodeReader.cpp:829
↓ 3 callersFunctiongetLoopSplitThresholdAttr
lib/Dialect/CudaTile/Transforms/LoopSplit.cpp:342
↓ 3 callersMethodgetMajor
Returns the major version number.
include/cuda_tile/Bytecode/Common/Version.h:35
↓ 3 callersMethodgetMinor
Returns the minor version number.
include/cuda_tile/Bytecode/Common/Version.h:38
↓ 3 callersFunctiongetSerializationOrder
Get parameters in serialization order.
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:151
↓ 3 callersMethodgetString
lib/Bytecode/Reader/BytecodeReader.cpp:223
↓ 3 callersFunctionget_integer_overflow_attr
Convert an enum value to the corresponding IntegerOverflowAttr.
python/cuda_tile/dialects/cuda_tile_ops.py:404
↓ 3 callersMethodinitialize
lib/Bytecode/Reader/BytecodeReader.cpp:485
↓ 3 callersFunctionmin
( lhs: Tile, rhs: Tile, *, propagate_nan: bool = False, flush_to_zero: bool = False, s
python/cuda_tile/dialects/cuda_tile_ops.py:4241
↓ 3 callersFunctionmlirCudaTileDenseI32ArrayAttrGet
lib/CAPI/Dialect/CudaTileDialect.cpp:392
↓ 3 callersFunctionmlirCudaTileDenseI32ArrayAttrGetElement
lib/CAPI/Dialect/CudaTileDialect.cpp:404
↓ 3 callersFunctionmlirCudaTileDenseI32ArrayAttrGetNumElements
lib/CAPI/Dialect/CudaTileDialect.cpp:399
↓ 3 callersFunctionmlirCudaTileTensorViewTypeGetDynamicSize
lib/CAPI/Dialect/CudaTileDialect.cpp:154
↓ 3 callersFunctionmlirCudaTileTypeIsATensorViewType
lib/CAPI/Dialect/CudaTileDialect.cpp:120
↓ 3 callersFunctionparseIndexTypes
Parse index types that may be either: - A single type (splat) applied to all indices - Individual types for each index Returns the parsed index types.
lib/Dialect/CudaTile/IR/CudaTile.cpp:3643
↓ 3 callersFunctionparseOptionalPaddingValue
lib/Dialect/CudaTile/IR/Types.cpp:62
↓ 3 callersMethodprint
lib/Dialect/CudaTile/IR/CudaTile.cpp:1898
↓ 3 callersMethodprint
lib/Dialect/CudaTile/IR/Types.cpp:188
↓ 3 callersFunctionprintAppliedType
tools/cuda-tile-tblgen/CudaTileType.cpp:241
↓ 3 callersFunctionprintIndexTypes
Print index types (splat or individual).
lib/Dialect/CudaTile/IR/CudaTile.cpp:3747
↓ 3 callersFunctionprintRoundingModeIfNotRN
lib/Dialect/CudaTile/IR/CudaTile.cpp:718
↓ 3 callersFunctionprintViewTkoCommon
lib/Dialect/CudaTile/IR/CudaTile.cpp:3719
next →1–100 of 974, ranked by callers