Function: set_kernel_args

Applicability

Product	Supported (√/x)
Atlas A3 training products / Atlas A3 inference products	√
Atlas A2 training products / Atlas A2 inference products	√
Atlas training products	√
Atlas inference products	√
Atlas 200I/500 A2 inference products	√

Function Usage

Sets tiling arguments and number of concurrency times of a dynamic-shape operator.

Prototype

C Prototype

        
             aclError aclopSetKernelArgs(aclopKernelDesc *kernelDesc,
const char *kernelId,
uint32_t blockDim,
const void *args,
uint32_t argSize)

Python Function

        
             ret = acl.op.set_kernel_args(kernel_desc, kernel_id, block_dim, args, arg_size)

Parameter Description

Parameter	Description
kernel_desc	Int, kernel description buffer, pointer address of the aclopKernelDesc type.
kernel_id	Str, kernel ID. It must be the same as that passed to the acl.op.create.kernel call.
block_dim	Int, number of blocks for kernel concurrent execution.
args	Int, tiling parameter. Use the NumPy array tobytes to obtain the bytes object, and then call acl.utils.bytes_to_ptr() to obtain the tiling parameter pointer address.
arg_size	Int, memory size of the tiling parameter, in bytes. It is the object length of bytes converted from numpy.tobytes() and can be obtained by using the len function.

Return Value Description

Return Value	Description
ret	Int, error code: 0 on success; else, failure.

Reference

For details about the API call sequence and example, see Single-Operator with Dynamic Shape (Operator Selector Registered).

Parent topic: Single-Operator Model Execution