Message transmission among processing units using interrupt control technique

摘要

The transmission of messages among multiple processors is controlled by the use of token and message interrupts. A token interrupt must be received from the processor sought to perform message processing. Failure to receive such a token interrupt prohibits requesting processors from making processing demands, i.e. sending a message interrupt, to the processor. Message interrupts, which may only be transmitted following receipt of a token interrupt, identify commands and user messages to be processed by a receiving processor. This permits instructions and multiple messages to be handled by a processor based on receipt of a single message interrupt. This technique also allows a processor under a temporary heavy load to defer the handling of other processing requests by delaying the transmission of a token interrupt.