Deep Neural Network Library (DNNL)  1.1.3
Performance library for Deep Learning
Functions

A primitive to copy data between memory formats. More...

Functions

dnnl_status_t DNNL_API dnnl_reorder_primitive_desc_create (dnnl_primitive_desc_t *reorder_primitive_desc, const dnnl_memory_desc_t *src_md, dnnl_engine_t src_engine, const dnnl_memory_desc_t *dst_md, dnnl_engine_t dst_engine, const_dnnl_primitive_attr_t attr)
 Initializes a reorder_primitive_desc using the description of the source (src_engine and src_md) and destination (dst_engine and dst_md) memory, and an attr attribute. More...
 

Detailed Description

A primitive to copy data between memory formats.

See also
Reorder in developer guide
Reorder in C++ API

Function Documentation

◆ dnnl_reorder_primitive_desc_create()

dnnl_status_t DNNL_API dnnl_reorder_primitive_desc_create ( dnnl_primitive_desc_t reorder_primitive_desc,
const dnnl_memory_desc_t src_md,
dnnl_engine_t  src_engine,
const dnnl_memory_desc_t dst_md,
dnnl_engine_t  dst_engine,
const_dnnl_primitive_attr_t  attr 
)

Initializes a reorder_primitive_desc using the description of the source (src_engine and src_md) and destination (dst_engine and dst_md) memory, and an attr attribute.

Inputs:

Outputs:

Examples:
cnn_inference_f32.c, cpu_cnn_training_f32.c, and cross_engine_reorder.c.