Intel e1000 zero copy method

A transmission method and technology of internal data, applied in the direction of instruments, electrical digital data processing, electrical components, etc., can solve problems such as inability to meet processing requirements
CN101178694AInactive Publication Date: 2008-05-14莱克斯科技(北京)有限公司

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
莱克斯科技(北京)有限公司
Publication Date
2008-05-14
Estimated Expiration
Not applicable · inactive patent

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
Patent Text Reader

Abstract

The invention relates to a transmitting method of the data communication in a computer and is a method of direct communication through the user applying procedure and network card hardware driving procedure. The kernel and the applying layer procedure commonly share one ring-shaped memory buffering district, the memory buffering district is divided into a number of slots (can also be regarded as array which can buffer store a number of messages), and each slot buffer stores one message. The network card driving adopts a DMA method to write message into a buffering district; and the applying layer space reads data from the buffering district. When a socket connector is used for controlling the connector using zero copying mechanism, e1000 receiving message driving starts to write date into the ring-shaped buffering district, the up layer applying can directly visit the data in the ring-shaped buffering district after the socket mechanism binding the connector and getting the mapped memory.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to a transmission method for internal data communication of a computer, which is a method for direct data communication through a user application program and a network card hardware driver program. Background technique

[0002] In practical applications, the network management program often needs to receive the original packets of the network card for analysis. However, the traditional packet capture mechanism is often unable to meet the requirements of large-traffic network flows, thus becoming the bottleneck of system processing. On the Linux system, packets have been copied multiple times from the NIC to the application layer, and involve switching between kernel mode and user mode, which reduces the ability of applications to capture packets. Although the kernel adopts NAPI, MMAP and other mechanisms to greatly improve the receiving performance from the network card to the kernel, it still cannot meet the processing requirem...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More