MCPcopy Create free account

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

Functions974 in github.com/NVIDIA/cuda-tile

↓ 1 callersFunction_offset
(lhs: Tile, rhs: Tile, *, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:1896
↓ 1 callersFunction_subf
( lhs: Tile, rhs: Tile, *, flush_to_zero: bool, rounding_mode: RoundingMode, loc=None,
python/cuda_tile/dialects/cuda_tile_ops.py:2651
↓ 1 callersFunction_subi
(lhs: Tile, rhs: Tile, *, loc=None, ip=None)
python/cuda_tile/dialects/cuda_tile_ops.py:2642
↓ 1 callersFunctionabsf
Performs element-wise absolute value on input float tile.
python/cuda_tile/dialects/cuda_tile_ops.py:1762
↓ 1 callersFunctionabsi
Performs element-wise absolute value on input integer tile.
python/cuda_tile/dialects/cuda_tile_ops.py:1755
↓ 1 callersFunctionbuildCudaTileOptimizationPipeline
Build optimization pipeline (default or with builder/text overrides)
lib/Dialect/CudaTile/Optimizer/CudaTileOptimizer.cpp:82
↓ 1 callersFunctionbuildDefaultCudaTilePipeline
Build default optimization pipeline
lib/Dialect/CudaTile/Optimizer/CudaTileOptimizer.cpp:56
↓ 1 callersMethodbuildFunctionMap
Collect all function metadata.
lib/Bytecode/Writer/BytecodeWriter.cpp:1338
↓ 1 callersFunctioncanonicalizeAddOperands
lib/Dialect/CudaTile/IR/CudaTile.cpp:1431
↓ 1 callersFunctioncompute_shape
(val)
python/cuda_tile/dialects/cuda_tile_ops.py:4573
↓ 1 callersFunctionconvertAttribute
tools/cuda-tile-tblgen/CudaTileType.cpp:413
↓ 1 callersFunctionconvertAttributeDef
tools/cuda-tile-tblgen/CudaTileType.cpp:336
↓ 1 callersFunctioncreateCompileUnitForLoc
Returns a new compile unit based on the file location contained within `loc`.
lib/Dialect/CudaTile/Transforms/SynthesizeDebugInfoScopes.cpp:56
↓ 1 callersFunctioncreateFunction
Creates a function based on the parsed FunctionInfo.
lib/Bytecode/Reader/BytecodeReader.cpp:2125
↓ 1 callersFunctioncreateGlobal
Creates a global (cuda_tile::GlobalOp) based on the parsed GlobalInfo.
lib/Bytecode/Reader/BytecodeReader.cpp:2353
↓ 1 callersFunctiondeserializeModule
lib/Bytecode/Translation/BytecodeTranslation.cpp:25
↓ 1 callersFunctiondiv
Performs element-wise division of two tiles.
python/cuda_tile/dialects/cuda_tile_ops.py:2549
↓ 1 callersFunctiondivf
( lhs: Tile, rhs: Tile, *, flush_to_zero: bool, rounding_mode: RoundingMode, loc=None,
python/cuda_tile/dialects/cuda_tile_ops.py:2514
↓ 1 callersFunctiondivi
( lhs: Tile, rhs: Tile, *, signedness: Signedness = Signedness.SIGNED, loc=None, ip=No
python/cuda_tile/dialects/cuda_tile_ops.py:2480
↓ 1 callersFunctionelementTypeFromString
tools/cuda-tile-tblgen/CudaTileType.cpp:28
↓ 1 callersFunctionelementTypeToString
tools/cuda-tile-tblgen/CudaTileType.cpp:121
↓ 1 callersFunctionemitAllElementTypeMatch
tools/cuda-tile-tblgen/SpecGen.cpp:243
↓ 1 callersFunctionemitAllRanksMatch
tools/cuda-tile-tblgen/SpecGen.cpp:274
↓ 1 callersFunctionemitAllTypesMatch
tools/cuda-tile-tblgen/SpecGen.cpp:229
↓ 1 callersFunctionemitAnyTypeOf
tools/cuda-tile-tblgen/SpecGen.cpp:261
↓ 1 callersFunctionemitAttribute
tools/cuda-tile-tblgen/SpecGen.cpp:440
↓ 1 callersFunctionemitAttributeInterface
tools/cuda-tile-tblgen/SpecGen.cpp:419
↓ 1 callersMethodemitCodeBlock
tools/cuda-tile-tblgen/Emitter.h:234
↓ 1 callersMethodemitComment
tools/cuda-tile-tblgen/Emitter.h:274
↓ 1 callersFunctionemitEnumAttribute
tools/cuda-tile-tblgen/SpecGen.cpp:374
↓ 1 callersMethodemitLiteralInclude
tools/cuda-tile-tblgen/Emitter.cpp:153
↓ 1 callersFunctionemitOpDoc
Emit documentation for an operation of the rough form: OP_NAME SHORT_DESCRIPTION SIGNATURE ARGUMENTS RESULTS DESCRIPTION CONSTRAINTS
tools/cuda-tile-tblgen/SpecGen.cpp:538
↓ 1 callersMethodemitOpHeading
todo move impl to .cpp files
tools/cuda-tile-tblgen/Emitter.h:213
↓ 1 callersFunctionemitOperationConstraint
tools/cuda-tile-tblgen/SpecGen.cpp:350
↓ 1 callersFunctionemitOperationExample
tools/cuda-tile-tblgen/SpecGen.cpp:516
↓ 1 callersFunctionemitOperationSignature
Emit the signature of an operation.
tools/cuda-tile-tblgen/SpecGen.cpp:454
↓ 1 callersFunctionemitOperationTrait
tools/cuda-tile-tblgen/SpecGen.cpp:345
↓ 1 callersFunctionemitOutputs
lib/Dialect/CudaTile/Optimizer/CudaTileOptimizer.cpp:229
↓ 1 callersFunctionemitSameOperandsAndResultElementType
tools/cuda-tile-tblgen/SpecGen.cpp:331
↓ 1 callersFunctionemitSameOperandsAndResultShape
tools/cuda-tile-tblgen/SpecGen.cpp:313
↓ 1 callersFunctionemitSameTypeOperands
tools/cuda-tile-tblgen/SpecGen.cpp:295
↓ 1 callersMethodemitSummary
tools/cuda-tile-tblgen/Emitter.h:224
↓ 1 callersFunctionemitTypesMatchWith
tools/cuda-tile-tblgen/SpecGen.cpp:290
↓ 1 callersFunctionentry
( sym_name, function_type, *, arch=None, arg_attrs=None, num_cta=None, num_worker_
python/cuda_tile/dialects/cuda_tile_ops.py:4004
↓ 1 callersFunctioneraseRestOfBlockFrom
Erase rest of block below given uop Needed when region, that replaced the operation, contains terminator
lib/Dialect/CudaTile/IR/CudaTile.cpp:2897
↓ 1 callersFunctionexamplesAppendixFile
tools/cuda-tile-tblgen/Emitter.cpp:141
↓ 1 callersFunctionextractFormatExpression
Extract a format expression from the given string, assuming that the string begins directly with the expression.
lib/Dialect/CudaTile/IR/CudaTile.cpp:4712
↓ 1 callersFunctionfindInterfaceImplementors
tools/cuda-tile-tblgen/CudaTileAttr.cpp:116
↓ 1 callersFunctionfind_mma_config
Find a matching MMA configuration for the given MLIR types
python/cuda_tile/dialects/cuda_tile_ops.py:669
↓ 1 callersFunctionfind_mma_scaled_config
Find a matching MMA Scaled configuration for the given MLIR types
python/cuda_tile/dialects/cuda_tile_ops.py:819
↓ 1 callersFunctionflatten
(val, depth)
python/cuda_tile/dialects/cuda_tile_ops.py:4592
↓ 1 callersFunctiongenerateAttrBytecode
Generate attribute bytecode definitions including: - AttributeTag enum - Runtime version checking function
tools/cuda-tile-tblgen/BytecodeGen.cpp:990
↓ 1 callersFunctiongenerateAttributeDeserialization
Generates C++ code within the 'parse<OpName>' function to deserialize the attributes of the given operation by calling the parseOpAttribute helper.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:358
↓ 1 callersFunctiongenerateAttributeParsingLogic
Helper function to generate common attribute parsing logic.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:122
↓ 1 callersFunctiongenerateAttributeSerialization
Generates C++ code within the 'write<OpName>' function to serialize the attributes of the given operation by calling the writeOpAttribute helper.
tools/cuda-tile-tblgen/BytecodeGen.cpp:431
↓ 1 callersFunctiongenerateBuiltinTypeDeserializers
Generate deserializers for all built-in types.
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:535
↓ 1 callersFunctiongenerateBuiltinTypeSerializers
Generate serializers for all built-in types.
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:313
↓ 1 callersFunctiongenerateBytecode
The main entry point for the TableGen backend.
tools/cuda-tile-tblgen/BytecodeGen.cpp:725
↓ 1 callersFunctiongenerateBytecodeReader
The main entry point for the TableGen backend.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:708
↓ 1 callersFunctiongenerateCudaTileTypeDeserializer
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:656
↓ 1 callersFunctiongenerateCudaTileTypeSerializer
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:457
↓ 1 callersFunctiongenerateDispatchSwitch
Generates the TypeSwitch statement for dispatching to op-specific writers. Returns FailureOr<size_t> where size_t is the number of serialized results.
tools/cuda-tile-tblgen/BytecodeGen.cpp:684
↓ 1 callersFunctiongenerateFlagsFieldDeserialization
Reads the flags field that encodes the presence of optional attributes and operands using individual bits.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:206
↓ 1 callersFunctiongenerateFlagsFieldSerialization
Generates the flags field serialization for optional attributes and operands. Version checking is only done for optional attributes and operands. The
tools/cuda-tile-tblgen/BytecodeGen.cpp:263
↓ 1 callersFunctiongenerateFunctionSignature
Generates the C++ function signature for the 'write<OpName>' function, which handles serialization for a specific cuda_tile operation. Returns Failure
tools/cuda-tile-tblgen/BytecodeGen.cpp:235
↓ 1 callersFunctiongenerateFunctionSignature
Generates the C++ function signature for the 'parse<OpName>' function, which handles deserialization for a specific cuda_tile operation.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:256
↓ 1 callersFunctiongenerateOpReader
Generates the complete C++ function 'parse<OpName>'.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:657
↓ 1 callersFunctiongenerateOpReaderDispatch
Generates the C++ switch statement to dispatch based on opcode.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:689
↓ 1 callersFunctiongenerateOpReaderImplementations
Generates the implementations of the individual op reader functions.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:675
↓ 1 callersFunctiongenerateOpWriter
Generates the complete C++ function 'write<OpName>'.
tools/cuda-tile-tblgen/BytecodeGen.cpp:644
↓ 1 callersFunctiongenerateOpWriterImplementations
Generates the implementations of the individual op writer functions.
tools/cuda-tile-tblgen/BytecodeGen.cpp:659
↓ 1 callersFunctiongenerateOpcodeEnumDefinition
Generates the opcode enum definition from TableGen records.
tools/cuda-tile-tblgen/BytecodeGen.cpp:158
↓ 1 callersFunctiongenerateOpcodeMap
Generates the opcode map implementation from TableGen records
tools/cuda-tile-tblgen/BytecodeGen.cpp:198
↓ 1 callersFunctiongenerateOpcodes
Generate opcode definitions in single file with ifdef guards
tools/cuda-tile-tblgen/BytecodeGen.cpp:905
↓ 1 callersFunctiongenerateOperandDeserialization
Generates C++ code within the 'parse<OpName>' function to deserialize the operands of the given operation.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:262
↓ 1 callersFunctiongenerateOperandSerialization
Generates C++ code within the 'write<OpName>' function to serialize the operands of the given operation.
tools/cuda-tile-tblgen/BytecodeGen.cpp:497
↓ 1 callersFunctiongenerateOperandSerializationLogic
Helper function to generate common operand serialization logic.
tools/cuda-tile-tblgen/BytecodeGen.cpp:413
↓ 1 callersFunctiongenerateOperationDeserialization
Generates C++ code within the 'parse<OpName>' function to deserialize the operation.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:649
↓ 1 callersFunctiongenerateOptionalParamFlags
Generates flags field serialization for optional type parameters.
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:376
↓ 1 callersFunctiongenerateOptionalParamFlagsReader
Generates flags field deserialization for optional type parameters.
tools/cuda-tile-tblgen/BytecodeTypeCodeGen.cpp:600
↓ 1 callersFunctiongenerateRegionDeserialization
Generates C++ code within the 'parse<OpName>' function to deserialize the regions of the given operation, if it has any.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:638
↓ 1 callersFunctiongenerateRegionSerialization
Generates C++ code within the 'write<OpName>' function to serialize the regions of the given operation, if it has any.
tools/cuda-tile-tblgen/BytecodeGen.cpp:524
↓ 1 callersFunctiongenerateResultTypeDeserialization
Generates C++ code to deserialize the result types of the operation.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:631
↓ 1 callersFunctiongenerateResultTypeSerialization
Generates C++ code within the 'write<OpName>' function to serialize the result types of the given operation.
tools/cuda-tile-tblgen/BytecodeGen.cpp:630
↓ 1 callersFunctiongenerateSimpleResultDeserialization
Generate result deserialization without version checking.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:490
↓ 1 callersFunctiongenerateSimpleResultSerialization
Generate result serialization without version checking.
tools/cuda-tile-tblgen/BytecodeGen.cpp:538
↓ 1 callersFunctiongenerateSupportedVersionsList
Generate getSupportedVersions() function from SupportedVersion records.
tools/cuda-tile-tblgen/BytecodeGen.cpp:875
↓ 1 callersFunctiongenerateTypeBytecode
Generate type bytecode functions.
tools/cuda-tile-tblgen/BytecodeGen.cpp:945
↓ 1 callersFunctiongenerateTypeReaderBytecode
Generate type reader bytecode functions.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:728
↓ 1 callersFunctiongenerateVersionAwareResultDeserialization
Generate version-aware result deserialization.
tools/cuda-tile-tblgen/BytecodeReaderGen.cpp:515
↓ 1 callersFunctiongenerateVersionAwareResultSerialization
Generate version-aware result serialization.
tools/cuda-tile-tblgen/BytecodeGen.cpp:553
↓ 1 callersFunctiongenerateVersionConstants
Generate version constants based on actual opcode assignments
tools/cuda-tile-tblgen/BytecodeGen.cpp:744
↓ 1 callersFunctiongenerateVersionValidation
Generate version validation function from SupportedVersion records.
tools/cuda-tile-tblgen/BytecodeGen.cpp:827
↓ 1 callersFunctiongetCudaTileModuleOp
===----------------------------------------------------------------------===// Serialization registration ===-----------------------------------------
lib/Bytecode/Translation/BytecodeTranslation.cpp:47
↓ 1 callersMethodgetCudaTileSpecGroup
tools/cuda-tile-tblgen/CudaTileOp.cpp:407
↓ 1 callersMethodgetDescriptionTables
tools/cuda-tile-tblgen/CudaTileOp.cpp:487
↓ 1 callersFunctiongetMLIRExamples
tools/cuda-tile-tblgen/CudaTileAttr.cpp:26
↓ 1 callersFunctiongetOperationConstraints
tools/cuda-tile-tblgen/CudaTileOp.cpp:137
↓ 1 callersFunctiongetRequestedOpDefinitions
tools/cuda-tile-tblgen/SpecGen.cpp:79
↓ 1 callersFunctiongetSplitThreshold
Merge optimization hints - more precise hint (if any) gets priority Default value is splitThreshold == 1 defined in pass options Return threshold (min
lib/Dialect/CudaTile/Transforms/LoopSplit.cpp:330
← previousnext →201–300 of 974, ranked by callers