Method for accessing cache and pseudo cache agent

摘要

Embodiments of the present invention disclose a method for accessing a cache and a pseudo cache agent (PCA). The method of the present invention is applied to a multiprocessor system, where the system includes at least one NC, at least one PCA conforming to a processor micro-architecture level interconnect protocol is embedded in the NC, the PCA is connected to at least one PCA storage device, and the PCA storage device stores data shared among memories in the multiprocessor system. The method of the present invention includes: if the NC receives a data request, obtaining, by the PCA, target data required in the data request from the PCA storage device connected to the PCA; and sending the target data to a sender of the data request. Embodiments of the present invention are mainly applied to a process of accessing cache data in the multiprocessor system.

主权项

1. A multiprocessor system, comprising:
at least two processors connected by a node controller (NC), wherein the NC is used to transfer data requests between the processors, and wherein each of the processors has a dedicated cache for storing data, a pseudo cache agent (PCA) which is located in the NC conforming to a processor micro-architecture level interconnect protocol, wherein the PCA is configured to: pre-fetch target data from a dedicated cache controlled by one of the processors, when network is idle, and write the target data into a PCA storage device, wherein the PCA storage device is coupled to the PCA and is dedicated to the PCA, and wherein the PCA storage device is accessible only to the PCA, wherein a first processor is configured to: receive a data request which includes an address of the target data, determine whether the target data is stored in a dedicated cache corresponding to the first processor based on the address, and send the data request to the NC to instruct the NC to forward the data request to a second processor if the target data is not stored in the dedicated cache corresponding to the first processor, and wherein the PCA is further configured to: receive the data request from the first processor, determine that the target data is stored in the PCA storage device based on the address, obtain the target data from the PCA storage device, and send the target data to the first processor.