tvm
Classes | Namespaces | Macros | Enumerations | Functions
buffer.h File Reference

Symbolic n-dimensional array, to represent a memory buffer. More...

#include <tvm/ir/expr.h>
#include <tvm/node/script_printer.h>
#include <tvm/runtime/container/array.h>
#include <tvm/runtime/container/string.h>
#include <tvm/tir/var.h>
#include <string>
Include dependency graph for buffer.h:
This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Classes

class  tvm::tir::BufferNode
 Node to represent a buffer. More...
 
class  tvm::tir::Buffer
 Buffer is a symbolic n-darray structure. It is a composition of primitive symbolic types, used to specify the memory layout of the Tensor used in program input. More...
 
class  tvm::tir::DataProducerNode
 Base node for data producers. More...
 
class  tvm::tir::DataProducer
 Managed reference to DataProducerNode. More...
 

Namespaces

 tvm
 runtime implementation for LibTorch/TorchScript.
 
 tvm::tir
 

Macros

#define TVM_INDEX_DEFAULT_I64   1
 

Enumerations

enum  tvm::tir::BufferType : int { tvm::tir::kDefault = 1 , tvm::tir::kAutoBroadcast = 2 }
 buffer type More...
 

Functions

DataType tvm::tir::DefaultIndexType ()
 if TVM_INDEX_DEFAULT_I64 is set, return int64, otherwise return int32 More...
 
Buffer tvm::tir::decl_buffer (Array< PrimExpr > shape, DataType dtype=DataType::Float(32), String name="buffer", String storage_scope="", Array< IntImm > axis_separators={}, Span span=Span())
 Construct a new buffer given shape, and dtype. More...
 
tir::Buffer tvm::tir::BufferWithOffsetAlignment (Array< PrimExpr > shape, DataType dtype, std::string name, int data_alignment, int offset_factor, bool compact, std::string memory_scope="")
 Creates TIR Buffer for provided parameters. More...
 

Detailed Description

Symbolic n-dimensional array, to represent a memory buffer.

Macro Definition Documentation

◆ TVM_INDEX_DEFAULT_I64

#define TVM_INDEX_DEFAULT_I64   1