HcclCommSuspend

This API is reserved and may be changed in later versions. It is not supported for developers.

Applicability

Product	Supported
Atlas 350 Accelerator Card	☓
Atlas A3 training product/Atlas A3 inference product	√
Atlas A2 training product/Atlas A2 inference product	√
Atlas 200I/500 A2 inference product	☓
Atlas inference product	☓
Atlas training product	☓

For Atlas A2 training product/Atlas A2 inference product, only the Atlas 800T A2 training server, Atlas 900 A2 PoD cluster basic unit, and Atlas 200T A2 Box16 heterogeneous subrack are supported.

Function

Suspends a communicator when an uncorrectable error (UCE) occurs in the on-chip memory (ACL_ERROR_RT_DEVICE_MTE_ERROR is returned by an ACL API).

When this API is used to suspend a communicator, you do not need to exit the process on the host. After the fault is rectified, you can call the HcclCommResume API to restore the communicator status.

Prototype

HcclResult HcclCommSuspend(HcclComm comm)

Parameters

Parameter	Input/Output	Description
comm	Input	Communicator to be suspended. For details about the definition of the HcclComm type, see HcclComm.

Parameter

Input/Output

Description

comm

Input

Communicator to be suspended.

For details about the definition of the HcclComm type, see HcclComm.

Returns

HcclResult: HCCL_SUCCESS on success, or else failure.

Constraints

This API must be used together with HcclCommResume.
This API cannot be executed concurrently with APIs related to Collective Communication or Point-to-Point Communication.

Parent topic: C APIs