Read-write method for forward information base and network processor

Abstract

The embodiment of the invention provides a read-write method for a forward information base and a network processor, and relates to the technical field of communication. The problem that in the prior art, due to the fact that commands sent to the Cache can not return responses in sequence and non-leaf nodes and corresponding leaf nodes can not be synchronously modified, the routing lookup goes wrong. According to the specific scheme, the network processor receives an AD writing request used for modifying the AD and modifying a Key writing request of the Key corresponding to the AD; before the AD writing request and the Key writing request pass through a buffer, when any code region where reading and writing are to be conducted includes the Key, the row address in the AD writing request is converted into the row address which is the same as the row address of the node block where the Key is located; when any code region where reading and writing are to be conducted does not contain the Key, the row address in the AD writing request and the row address in the Key writing request are converted into the same row address in any node; thus, any node area is read and written according to the same row address. The method is used for writing and reading the forward information base.

Description

technical field [0001] The invention relates to the technical field of communication, in particular to a method for reading and writing a forwarding information table and a network processor. Background technique [0002] The Forward Information Base (FIB) is an information table that provides network devices (such as bridges, switches or routers, etc.) in data communication to forward data packets to the destination address, and is stored in the network processor (Network Processor) of the network hardware. , NP), its structural schematic diagram can be found in figure 1 , the FIB table adopts a binary search tree (Binary Search Tree, Btree) structure, and each non-empty node corresponds to a FIB entry. In the Btree structure, the nodes of other layers except the last layer (Branch) nodes can be called non-leaf nodes, which are used to store keywords (Key) such as Internet Protocol (Internet Protocol, IP) addresses, and the nodes of the last layer can be called As a leaf ...

AI Technical Summary

Problems solved by technology

[0007] Embodiments of the present invention provide a method for reading and writing forwarding information tables and a network processor, which can solve the problem that the commands sent to the Cache in the prior art cannot return responses sequentially, so that non-leaf nodes and corresponding leaf nodes cannot be modified synchronously, resulting in The problem of routing lookup error

Images