aclrtMemSetPidToShareableHandleV2

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	√

Description

Sets the trustlist of processes that can share memory.

This API is enhanced based on aclrtMemSetPidToShareableHandle. You can use the shareType parameter to specify whether to export the shared handle within an AI server or from another AI server.

The usage of this API is similar to that of aclrtMemExportToShareableHandle. However, this API must be used together with aclrtMemExportToShareableHandleV2 to export a shared handle and with aclrtMemImportFromShareableHandleV2 to import a shared handle.

Prototype

aclError aclrtMemSetPidToShareableHandleV2(void *shareableHandle, aclrtMemSharedHandleType shareType, int32_t *pid, size_t pidNum)

Parameters

Parameter	Input/Output	Description
shareableHandle	Input	A shareableHandle exported using the aclrtMemExportToShareableHandleV2 API. It is a pointer that points to the shared handle.
shareType	Input	Type of the exported shared handle. For details about the type definition, see aclrtMemSharedHandleType.
pid	Input	Array for storing trustlisted process IDs. You can call aclrtDeviceGetBareTgid to obtain the process ID. In the Docker scenario, the process ID on the physical machine is obtained. In the non-Docker scenario, the process ID is obtained.
pidNum	Input	Number of processes in the trustlist. The value must be the same as the size of the pid parameter array.

Returns

0 on success; else, failure. For details, see aclError.

Parent topic: Memory Management