Wait

Product Support

Product	Supported
Atlas A3 training products/Atlas A3 inference products	x
Atlas A2 training products/Atlas A2 inference products	√
Atlas 200I/500 A2 inference products	x
Atlas inference product's AI Core	x
Atlas inference product's Vector Core	x
Atlas training products	x

Function

Checks whether a message has been processed by the AIC after PostMessage or PostFakeMessage is called.

Prototype

template <bool sync = true>
__aicore__ inline bool Wait(uint16_t offset)

Parameters

**Table 1** Template parameters
Parameter	Description
sync	Whether the program needs to wait when messages are queried. Values: true: The program can continue to run only after the AIC has processed the message. false: The system only checks whether the AIC has processed the message.

**Table 2** API parameters
Parameter	Input/Output	Description
offset	Input	Offset of the message space address, which is obtained from the return of PostMessage or PostFakeMessage.

Returns

true: The current message has been processed by the AIC.
false: The current message has not been fully processed by AIC.

Restrictions

None

Example

auto msgPtr = handle.AllocMessage();        // A new message can be sent at the location pointed to by msgPtr.
AscendC::CubeGroupMsgHead headA = {AscendC::CubeMsgState::VALID, 0};
AscendC::CubeMsgBody msgA = {headA, 1, 0, 0, false, false, false, false, 0, 0, 0, 0, 0, 0, 0, 0};
auto offset = handle.PostMessage(msgPtr, msgA);           // Fill the custom message structure at the location pointed to by msgPtr and send the message.
bool waitState = handle.template Wait<true>(offset);      // Wait for the AIC to finish processing msgA.
// Fake message scenario
auto msgFakePtr = handle.AllocMessage();
offset = handle.PostFakeMsg(msgFakePtr);
bool waitState = handle.template Wait<true>(offset); // Wait for the AIC to finish processing msgFake.

Parent topic: CubeResGroupHandle