A Multi-High-Speed ​​Interface Protocol Message Conversion System

By using a message conversion system based on programmable logic devices for multiple high-speed interface protocols, the problems of communication difficulties and excessive conversion delays caused by multiple high-speed protocol interfaces in communication systems are solved. This system achieves flexible and efficient message conversion and memory space management, improving system scalability and data access efficiency.

CN117596309BActive Publication Date: 2026-11-17THE 54TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY GROUP CORPORATION
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202311586422.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-11-24
Publication Date
2026-11-17
Estimated Expiration
2043-11-24

AI Technical Summary

Technical Problem

In existing communication systems, various high-speed protocol interfaces cause difficulties in device communication and excessive conversion delays. This is especially true in complex communication scenarios, where fixed CPU architectures struggle to handle high-speed, large-capacity data, and variable interrupt response times lead to excessive delays in message processing.

Method used

A message conversion system based on programmable logic devices is adopted, which includes PCIE, SRIO and XAUI processing modules, combined with a routing address module and a dynamic space control module. By parsing the transaction information header and dynamically allocating memory space, message conversion between different interface protocols is realized.

Benefits of technology

It supports high-speed data exchange in various scenarios, improves message conversion efficiency, reduces header overhead, and increases memory space utilization. It supports one-to-one, one-to-many, many-to-one, and many-to-many transmission, enhancing system scalability and data access efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117596309B_ABST
    Figure CN117596309B_ABST
Patent Text Reader

Abstract

This invention provides a multi-high-speed interface protocol message conversion system, belonging to the field of computer communication. It includes a PCIE processing module, an SRIO processing module, an XAUI processing module, a routing address module, a dynamic space control module, and a memory. The PCIE, SRIO, and XAUI processing modules receive messages to obtain transaction information headers and payloads, transmit the payloads to volatile memory, and send a transmission completion signal to the dynamic space control module upon completion of transmission. The routing address module sends information required for message generation based on the transaction information header. The PCIE, SRIO, and XAUI processing modules generate and transmit messages based on the required message generation information and the payload. The dynamic space control module receives information from the transaction information header and the transmission completion signal to update the start and end addresses of the PCIE, SRIO, and XAUI processing modules. This invention enables efficient conversion and transmission between PCIE, SRIO, and XAUI messages.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of computer communication technology, and in particular to a message conversion system for multiple high-speed interface protocols, which can realize message conversion between PCIE, SRIO and XAUI high-speed interface protocols in communication systems. Background Technology

[0002] With the rapid development of electronic information technology, the number of electronic information devices in communication systems has increased dramatically, and the amount of data processed has grown rapidly. Systems now require high transmission speeds, flexible processing methods, and stable transmission. Chip-to-chip and board-to-board communication typically uses high-speed interface protocols such as PCIe or SRIO, which offer high speeds and reliability. Long-distance high-speed communication often uses 10 Gigabit Ethernet. Compared to the XGMII 10 Gigabit Ethernet interface, the XAUI 10 Gigabit Ethernet interface significantly reduces the number of signal lines, and differential signals have stronger anti-interference capabilities. The lack of standardized high-speed interface protocols in communication systems leads to communication difficulties.

[0003] Traditional message conversion methods typically utilize the CPU, employing a one-to-one protocol conversion approach. However, the fixed architecture of the CPU makes it difficult to handle high-speed, large-capacity data, and the CPU usually relies on external interrupts to process or transmit messages, but the interrupt response time is variable, resulting in long delays for message processing. For example, patent application CN106027424A discloses an Ethernet switching device based on RapidIO switching technology, which uses the CPU to perform the logical management of the message conversion process. This one-to-one conversion method severely impacts communication efficiency in complex communication scenarios, resulting in long message conversion delays and a message conversion rate limited by the lowest-speed conversion device in the message conversion link. Summary of the Invention

[0004] In view of this, the present invention proposes a message conversion system for multiple high-speed interface protocols, which can realize message conversion between PCIE, SRIO and XAUI ends, so as to solve the technical problems of difficult device communication and excessive conversion delay caused by multiple high-speed protocol interfaces in the existing communication system, and realize high-speed data exchange in different scenarios.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A multi-high-speed interface protocol message conversion system, implemented based on a programmable logic device, includes a PCIE processing module, an SRIO processing module, an XAUI processing module, a routing address module, a dynamic space control module, and a memory; wherein:

[0007] The PCIe processing module, connected to the PCIe high-speed interface, is used to obtain the transaction information header from the payload of the first received PCIe message, and transmit the actual payload of the first PCIe message and the payloads of subsequent PCIe messages to the memory; or, read data from the memory, generate and send PCIe messages, and generate a transmission completion signal when the PCIe message transmission is completed.

[0008] The SRIO processing module, connected to the SRIO high-speed interface, is used to obtain the transaction information header from the payload of the first received SRIO message, and transmit the actual payload of the first SRIO message and the payloads of subsequent SRIO messages to the memory; or, read data from the memory, generate and send SRIO messages, and generate a transmission completion signal when the SRIO message transmission is completed.

[0009] The XAUI processing module, connected to the XAUI high-speed interface, is used to obtain the transaction information header from the payload of the first received XAUI message, and transmit the actual payload of the first XAUI message and the payloads of subsequent XAUI messages to the memory; or, read data from the memory, generate and send XAUI messages, and generate a transmission completion signal when the XAUI message transmission is completed.

[0010] The routing address module is used to receive and parse the source interface ID and destination interface ID in the transaction data header, and send the address information required to generate the message.

[0011] The dynamic space control module is used to receive the source interface ID and destination interface ID or the transmission completion signal in the transaction data header, and return and update the start address and end address of the allocated memory space.

[0012] The memory is used to cache valid data before the high-speed interface protocol is converted.

[0013] Furthermore, the specific method for converting PCIe messages to SRIO messages is as follows:

[0014] (101) The PCIE processing module obtains the transaction information header and the actual payload from the payload of the first PCIE message. The transaction information header includes the source interface ID, the destination interface ID and the total transmission length.

[0015] (102) The routing address module sends the SRIO device ID to the SRIO processing module based on the source interface ID and the destination interface ID;

[0016] (103) The PCIE processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent PCIE messages to the memory, and counts the length of the payload; the SRIO processing module reads the payload stored in the memory.

[0017] (104) The SRIO processing module splits the payload into an SRIO header and an SRIO payload, and generates an SRIO message based on the SRIO device ID, the SRIO header and the SRIO payload.

[0018] (105) When the length of the effective payload counted by the PCIE processing module is equal to the total transmission length, the subsequent PCIE message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

[0019] Furthermore, the specific method for converting PCIe packets to XAUI packets is as follows:

[0020] (201) The PCIE processing module obtains the transaction information header and the actual payload from the payload of the first PCIE message. The transaction information header includes the source interface ID, the destination interface ID and the total transmission length.

[0021] (202) The routing address module sends the MAC address segment, IP address segment and UDP port segment to the XAUI processing module according to the source interface ID and the destination interface ID;

[0022] (203) The PCIE processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent PCIE messages to the memory, and counts the length of the payload; the XAUI processing module reads the payload stored in the memory.

[0023] (204) The XAUI processing module splits the payload and generates XAUI packets based on the MAC address range, IP address range, UDP port range and the split payload.

[0024] (205) When the length of the effective payload counted by the PCIE processing module is equal to the total transmission length, the subsequent PCIE message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

[0025] Furthermore, the specific method for converting SRIO messages to PCIe messages is as follows:

[0026] (301) The SRIO processing module obtains the transaction information header and the actual payload from the payload of the first SRIO message; the transaction information header includes the source interface ID, the destination interface ID and the total transmission length.

[0027] (302) The routing address module sends the header type, supplementary information, message length and address to the PCIe processing module according to the source interface ID and the destination interface ID;

[0028] (303) The SRIO processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent SRIO messages to the memory, and counts the length of the payload; the PCIE processing module reads the payload stored in the memory.

[0029] (304) The PCIE processing module splits the payload and generates a PCIE message based on the header type, supplementary information, message length and address, and the split payload;

[0030] (305) When the length of the effective payload counted by the SRIO processing module is equal to the total transmission length, the subsequent SRIO message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

[0031] Furthermore, the specific method for converting SRIO messages to XAUI messages is as follows:

[0032] (401) The SRIO processing module obtains the transaction information header and the actual payload from the payload of the first SRIO message; the transaction information header includes the source interface ID, the destination interface ID and the total transmission length.

[0033] (402) The routing address module sends the MAC address segment, IP address segment and UDP port segment to the XAUI processing module according to the source interface ID and the destination interface ID;

[0034] (403) The SRIO processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent SRIO messages to the memory, and counts the length of the payload; the XAUI processing module reads the payload stored in the memory.

[0035] (404) The XAUI processing module splits the payload and generates XAUI packets based on the MAC address range, IP address range, UDP port range and the split payload.

[0036] (405) When the length of the effective payload counted by the SRIO processing module is equal to the total transmission length, the subsequent SRIO message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

[0037] Furthermore, the specific method for converting XAUI messages to PCIe messages is as follows:

[0038] (501) The XAUI processing module obtains the transaction information header and the actual payload from the payload of the first XAUI message; the transaction information header includes the source interface ID, the destination interface ID and the total transmission length;

[0039] (502) The routing address module sends the header type, supplementary information, message length and address to the PCIe processing module according to the source interface ID and the destination interface ID;

[0040] (503) The XAUI processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent XAUI messages to the memory, and counts the length of the payload; the PCIE processing module reads the payload stored in the memory.

[0041] (504) The PCIE processing module splits the payload and generates a PCIE message based on the header type, supplementary information, message length and address, and the split payload;

[0042] (505) When the length of the effective payload counted by the XAUI processing module is equal to the total transmission length, the subsequent XAUI message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

[0043] Furthermore, the specific method for converting XAUI messages to SRIO messages is as follows:

[0044] (601) The XAUI processing module obtains the transaction information header and the actual payload from the payload of the first XAUI message; the transaction information header includes the source interface ID, the destination interface ID and the total transmission length;

[0045] (602) The routing address module sends the SRIO device ID to the SRIO processing module based on the source interface ID and the destination interface ID;

[0046] (603) The XAUI processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent XAUI messages to the memory, and counts the length of the payload; the SRIO processing module reads the payload stored in the memory.

[0047] (604) The SRIO processing module splits the payload into an SRIO header and an SRIO payload, and generates an SRIO message based on the SRIO device ID, the SRIO header and the SRIO payload;

[0048] (605) When the length of the effective payload counted by the XAUI processing module is equal to the total transmission length, the subsequent XAUI message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

[0049] Furthermore, the dynamic space control module operates as follows:

[0050] When the dynamic space control module receives the source interface ID, destination interface ID, and address of the currently transmitted memory from the PCIE processing module, SRIO processing module, or XAUI processing module, it updates the start and end addresses of the PCIE processing module, SRIO processing module, or XAUI processing module to ensure that the currently transmitted space address is within the allocated address space.

[0051] When the dynamic space control module receives a transmission completion signal from the PCIE processing module, SRIO processing module, or XAUI processing module, it releases the occupied memory space and updates the start and end addresses of the PCIE processing module, SRIO processing module, or XAUI processing module.

[0052] The beneficial effects of this invention are as follows:

[0053] 1. This invention supports one-to-one, one-to-many, many-to-one, and many-to-many transmission of messages from PCIE, SRIO, and XAUI ends. It can convert messages in various scenarios. The SRIO end connects to the SRIO network and supports communication with multiple SRIO devices. The XAUI end connects to the network switch and supports communication with multiple network devices.

[0054] 2. This invention supports mixed transmission of multiple SRIO messages such as SWRITE, NWRITE_R, NWRITE, NREAD, DOORBELL, MESSAGE, and MAINTENANCE. By adding a header prefix to the SRIO message to identify multiple SRIO messages, the conversion performance of multiple SRIO messages can be further improved while ensuring flexible configuration.

[0055] 3. This invention obtains the transaction information header by decoding the first received message and uses the transaction information header to obtain the message configuration information. This eliminates the need for subsequent messages to carry transaction information headers or message configuration information, thereby reducing message header overhead and improving message conversion efficiency.

[0056] 4. This invention adopts a unified transaction information header format, which reduces the complexity of the decoding circuit for the transaction information header and improves the message conversion efficiency.

[0057] 5. This invention can dynamically partition volatile memory space. When a high-speed interface starts transmitting, it will allocate its corresponding volatile memory space to buffer the message. When the message transmission is completed, it will release the corresponding volatile memory space or allocate it to other high-speed interfaces that are transmitting, which greatly improves the utilization rate of memory space. Attached Figure Description

[0058] Figure 1 System architecture block diagram provided in Embodiment 1 of the present invention

[0059] Figure 2 This is a structural diagram of the first PCIE message provided in Embodiment 2 of the present invention.

[0060] Figure 3 This is a structural diagram of the PCIE message to SRIO message and XAUI message conversion method provided in Embodiment 2 of the present invention.

[0061] Figure 4 This is a structural diagram of the first SRIO message provided in Embodiment 3 of the present invention.

[0062] Figure 5 This is a structural diagram of the method for converting SRIO messages to PCIE messages and XAUI messages provided in Embodiment 3 of the present invention.

[0063] Figure 6 This is a structural diagram of the first XAUI message provided in Embodiment 4 of the present invention.

[0064] Figure 7 This is a structural diagram of the method for converting XAUI messages to PCIE messages and SRIO messages provided in Embodiment 4 of the present invention. Detailed Implementation

[0065] The embodiments described below illustrate the information necessary to enable those skilled in the art to practice the embodiments and to explain the best mode of practicing the embodiments. It should be noted that all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0066] Example 1

[0067] This embodiment provides a high-speed interface protocol data conversion system for PCIe, SRIO, and XAUI. The system is implemented based on a programmable logic device, such as... Figure 1 As shown, it includes a PCIe processing module, an SRIO processing module, an XAUI processing module, a routing address module, a dynamic space control module, and volatile memory, wherein:

[0068] The PCIe processing module connects to the PCIe physical high-speed interface. It extracts the transaction information header from the payload of the first received PCIe packet, transmits the actual payload of the first PCIe packet and the payloads of subsequent PCIe packets to the volatile memory, reads data from the volatile memory, and generates and sends PCIe packets. A transmission completion signal is generated when the PCIe packet transmission is complete.

[0069] The SRIO processing module is connected to the SRIO physical high-speed interface. It obtains the transaction information header from the payload of the first received SRIO message, transmits the actual payload of the first SRIO message and the payloads of subsequent SRIO messages to the volatile memory, reads data from the volatile memory, generates and sends SRIO messages, and generates a transmission completion signal when the SRIO message transmission is completed.

[0070] The XAUI processing module is connected to the XAUI physical high-speed interface. The XAUI processing module obtains the transaction information header from the payload of the first received XAUI message, transmits the actual payload of the first XAUI message and the payloads of subsequent XAUI messages to the volatile memory, reads data from the volatile memory, generates and sends XAUI messages, and generates a transmission completion signal when the XAUI message transmission is completed.

[0071] The routing address module receives and parses the source interface ID and destination interface ID in the transaction data header, and sends the address information required to generate the message.

[0072] The dynamic space control module receives the source interface ID and destination interface ID or the transmission completion signal from the transaction data header, and returns and updates the start and end addresses of the allocated volatile memory space.

[0073] Volatile memory is used to cache valid data before high-speed interface protocol conversion.

[0074] This system exists as an EP device on the PCIe bus and can access the SRIO network via the SRIO bus. It supports communication with multiple network devices through a network switch. This greatly improves system scalability, effectively increases data access efficiency, and saves on network communication costs.

[0075] Example 2

[0076] This embodiment relates to a method for converting PCIE packets to SRIO packets or XAUI packets.

[0077] The structure of the first PCIe message in this embodiment is shown below. Figure 2 The message consists of a PCIe header and a payload. The PCIe header includes header type, additional information, message length, and address. The payload contains a transaction information header and the actual payload. The transaction information header includes the source interface ID, destination interface ID, and total transmission length. The source interface ID identifies the PCIe device sending the message, and the destination interface ID identifies the specific device receiving the message. The total transmission length is the total length of the PCIe message payload during transmission.

[0078] The conversion structure from PCIE packets to SRIO packets or XAUI packets involved in this embodiment is as follows: Figure 3 As shown, it includes:

[0079] The PCIE processing module obtains the transaction information header and the actual payload from the payload of the first received PCIE message, temporarily stores the total transmission length in the transaction information header, calculates the length of the actual payload, and transmits the actual payload to volatile memory.

[0080] The routing address module sends and temporarily stores the SRIO device ID to the SRIO processing module based on the source interface ID and destination interface ID in the transaction information header, or sends and temporarily stores the MAC address range, IP address range, and UDP port range to the XAUI processing module.

[0081] The PCIe processing module transmits the payload of subsequent PCIe packets received to the volatile memory and calculates the payload length of subsequent PCIe packets. The SRIO processing module or XAUI processing module reads the payload stored in the volatile memory.

[0082] The SRIO processing module splits the payload into an SRIO header and an SRIO payload, and generates an SRIO message based on the SRIO device ID, the SRIO header, and the SRIO payload.

[0083] The XAUI processing module splits the payload and generates an XAUI message based on the MAC address segment, IP address segment, and UDP port segment and the split payload.

[0084] When the effective payload length counted by the PCIE processing module equals the temporary total transmission length, the subsequent PCIE message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

[0085] Example 3

[0086] This embodiment relates to a method for converting SRIO messages to PCIE messages or XAUI messages.

[0087] The structure of the first SRIO message in this embodiment is shown below. Figure 4The SRIO message includes an SRIO header, an SRIO device ID, and a payload. The payload consists of a transaction information header and the actual payload. The transaction information header includes the source interface ID, the destination interface ID, and the total transmission length. The source interface ID identifies the SRIO device sending the SRIO message, and the destination interface ID identifies the specific device receiving the SRIO message. The total transmission length is the total length of the SRIO message payload during transmission.

[0088] The structure of the SRIO message to PCIE message or XAUI message in this embodiment is as follows: Figure 5 As shown, it includes:

[0089] The SRIO processing module obtains the transaction information header and the actual payload from the payload of the first received SRIO message, temporarily stores the total transmission length in the transaction information header, calculates the length of the actual payload, and transmits the actual payload to the volatile memory.

[0090] The routing address module sends and temporarily stores the header type, supplementary information, message length, and address to the PCIE processing module based on the source interface ID and destination interface ID in the transaction information header, or sends and temporarily stores the MAC address range, IP address range, and UDP port range to the XAUI processing module.

[0091] The SRIO processing module transmits the payload of subsequent SRIO messages received to the volatile memory and calculates the payload length of subsequent SRIO messages. The PCIE processing module or XAUI processing module reads the payload stored in the volatile memory.

[0092] The PCIE processing module splits the payload and generates a PCIE packet based on the header type, supplementary information, packet length, and address, along with the split payload.

[0093] The XAUI processing module splits the payload and generates an XAUI message based on the MAC address segment, IP address segment, and UDP port segment and the split payload.

[0094] When the effective payload length counted by the SRIO processing module is equal to the temporary total transmission length, the subsequent SRIO message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

[0095] Example 4

[0096] This embodiment relates to a method for dynamically partitioning volatile memory space, including:

[0097] The dynamic space control module receives the source interface ID, destination interface ID, and the address of the currently transmitted volatile memory sent by the PCIE processing module, SRIO processing module, and XAUI processing module, and updates the start and end addresses of the PCIE processing module, SRIO processing module, and XAUI processing module to ensure that the space address currently being transmitted is within the allocated address space.

[0098] When the dynamic space control module receives the transmission completion signal sent by the PCIE processing module, SRIO processing module and XAUI processing module, it releases the occupied volatile memory space and updates the start address and end address of the PCIE processing module, SRIO processing module and XAUI processing module.

[0099] Example 5

[0100] This embodiment relates to a method for converting XAUI messages to PCIE messages or SRIO messages.

[0101] The structure of the first XAUI message in this embodiment is shown below. Figure 6 The XAUI message consists of an XAUI header and a payload. The XAUI header includes a MAC address field, an IP address field, and a UDP port field. The payload contains a transaction information header and the actual payload. The transaction information header includes the source interface ID, the destination interface ID, and the total transmission length. The source interface ID identifies the XAUI device sending the message, and the destination interface ID identifies the specific device receiving the message. The total transmission length is the total length of the XAUI message payload during transmission.

[0102] The method for converting XAUI messages to PCIe messages or SRIO messages in this embodiment is as follows: Figure 7 As shown, it includes:

[0103] The XAUI processing module obtains the transaction information header and the actual payload from the payload of the first received XAUI message, temporarily stores the total transmission length in the transaction information header, calculates the length of the actual payload, and transmits the actual payload to the volatile memory.

[0104] The routing address module sends and temporarily stores the header type, auxiliary information, message length, and address to the PCIE processing module based on the source interface ID and destination interface ID in the transaction information header, or sends and temporarily stores the SRIO device ID to the SRIO processing module.

[0105] The XAUI processing module transmits the payload of subsequent XAUI messages received to the volatile memory and calculates the payload length of subsequent XAUI messages. The PCIe processing module or SRIO processing module reads the payload stored in the volatile memory.

[0106] The PCIE processing module splits the payload and generates a PCIE packet based on the header type, supplementary information, packet length, and address, along with the split payload.

[0107] The SRIO processing module splits the payload into an SRIO header and an SRIO payload, and generates an SRIO message based on the SRIO device ID, the SRIO header, and the SRIO payload.

[0108] When the effective payload length calculated by the XAUI processing module is equal to the temporary total transmission length, the subsequent XAUI message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

Claims

1. A multi-high-speed interface protocol message conversion system, characterized in that, Implemented using programmable logic devices, it includes a PCIe processing module, an SRIO processing module, an XAUI processing module, a routing address module, a dynamic space control module, and memory; among which: The PCIe processing module, connected to the PCIe high-speed interface, is used to obtain the transaction information header from the payload of the first received PCIe message, and transmit the actual payload of the first PCIe message and the payloads of subsequent PCIe messages to the memory; or, read data from the memory, generate and send PCIe messages, and generate a transmission completion signal when the PCIe message transmission is completed. The SRIO processing module, connected to the SRIO high-speed interface, is used to obtain the transaction information header from the payload of the first received SRIO message, and transmit the actual payload of the first SRIO message and the payloads of subsequent SRIO messages to the memory; or, read data from the memory, generate and send SRIO messages, and generate a transmission completion signal when the SRIO message transmission is completed. The XAUI processing module, connected to the XAUI high-speed interface, is used to obtain the transaction information header from the payload of the first received XAUI message, and transmit the actual payload of the first XAUI message and the payloads of subsequent XAUI messages to the memory; or, read data from the memory, generate and send XAUI messages, and generate a transmission completion signal when the XAUI message transmission is completed. The routing address module is used to receive and parse the source interface ID and destination interface ID in the transaction data header, and send the address information required to generate the message. The dynamic space control module is used to receive the source interface ID and destination interface ID or the transmission completion signal in the transaction data header, and return and update the start address and end address of the allocated memory space. The memory is used to cache valid data before the high-speed interface protocol is converted.

2. The system of claim 1, wherein, The specific method for converting PCIe messages to SRIO messages is as follows: (101) The PCIE processing module obtains the transaction information header and the actual payload from the payload of the first PCIE message. The transaction information header includes the source interface ID, the destination interface ID and the total transmission length. (102) The routing address module sends the SRIO device ID to the SRIO processing module based on the source interface ID and the destination interface ID; (103) The PCIE processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent PCIE messages to the memory, and counts the length of the payload; the SRIO processing module reads the payload stored in the memory. (104) The SRIO processing module splits the payload into an SRIO header and an SRIO payload, and generates an SRIO message based on the SRIO device ID, the SRIO header and the SRIO payload. (105) When the length of the effective payload counted by the PCIE processing module is equal to the total transmission length, the subsequent PCIE message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

3. The system of claim 1, wherein, The specific method for converting PCIe packets to XAUI packets is as follows: (201) The PCIE processing module obtains the transaction information header and the actual payload from the payload of the first PCIE message. The transaction information header includes the source interface ID, the destination interface ID and the total transmission length. (202) The routing address module sends the MAC address segment, IP address segment and UDP port segment to the XAUI processing module according to the source interface ID and the destination interface ID; (203) The PCIE processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent PCIE messages to the memory, and counts the length of the payload. The XAUI processing module reads the effective payload stored in the memory. (204) The XAUI processing module splits the payload and generates XAUI packets based on the MAC address range, IP address range, UDP port range and the split payload. (205) When the length of the effective payload counted by the PCIE processing module is equal to the total transmission length, the subsequent PCIE message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

4. The system of claim 1, wherein, The specific method for converting SRIO messages to PCIe messages is as follows: (301) The SRIO processing module obtains the transaction information header and the actual payload from the payload of the first SRIO message; the transaction information header includes the source interface ID, the destination interface ID and the total transmission length. (302) The routing address module sends the header type, supplementary information, message length and address to the PCIe processing module according to the source interface ID and the destination interface ID; (303) The SRIO processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent SRIO messages to the memory, and counts the length of the payload. The PCIe processing module reads the effective payload stored in the memory. (304) The PCIE processing module splits the payload and generates a PCIE message based on the header type, supplementary information, message length and address, and the split payload; (305) When the length of the effective payload counted by the SRIO processing module is equal to the total transmission length, the subsequent SRIO message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

5. The system of claim 1, wherein, The specific method for converting SRIO messages to XAUI messages is as follows: (401) The SRIO processing module obtains the transaction information header and the actual payload from the payload of the first SRIO message; the transaction information header includes the source interface ID, the destination interface ID and the total transmission length. (402) The routing address module sends the MAC address segment, IP address segment and UDP port segment to the XAUI processing module according to the source interface ID and the destination interface ID; (403) The SRIO processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent SRIO messages to the memory, and counts the length of the payload; the XAUI processing module reads the payload stored in the memory. (404) The XAUI processing module splits the payload and generates XAUI packets based on the MAC address range, IP address range, UDP port range and the split payload. (405) When the length of the effective payload counted by the SRIO processing module is equal to the total transmission length, the subsequent SRIO message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

6. The system of claim 1, wherein, The specific method for converting XAUI messages to PCIe messages is as follows: (501) The XAUI processing module obtains the transaction information header and the actual payload from the payload of the first XAUI message; the transaction information header includes the source interface ID, the destination interface ID and the total transmission length; (502) The routing address module sends the header type, supplementary information, message length and address to the PCIe processing module according to the source interface ID and the destination interface ID; (503) The XAUI processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent XAUI messages to the memory, and counts the length of the payload. The PCIe processing module reads the effective payload stored in the memory. (504) The PCIE processing module splits the payload and generates a PCIE message based on the header type, supplementary information, message length and address, and the split payload; (505) When the length of the effective payload counted by the XAUI processing module is equal to the total transmission length, the subsequent XAUI message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

7. The system of claim 1, wherein, The specific method for converting XAUI messages to SRIO messages is as follows: (601) The XAUI processing module obtains the transaction information header and the actual payload from the payload of the first XAUI message; the transaction information header includes the source interface ID, the destination interface ID and the total transmission length; (602) The routing address module sends the SRIO device ID to the SRIO processing module based on the source interface ID and the destination interface ID; (603) The XAUI processing module temporarily stores the total transmission length in the transaction information header and transmits the payload of subsequent XAUI messages to the memory, and counts the length of the payload; the SRIO processing module reads the payload stored in the memory. (604) The SRIO processing module splits the payload into an SRIO header and an SRIO payload, and generates an SRIO message based on the SRIO device ID, the SRIO header and the SRIO payload; (605) When the length of the effective payload counted by the XAUI processing module is equal to the total transmission length, the subsequent XAUI message transmission is completed, and a transmission completion signal is sent to the dynamic space control module.

8. The system of claim 1, wherein, The dynamic space control module works as follows: When the dynamic space control module receives the source interface ID, destination interface ID, and address of the currently transmitted memory from the PCIE processing module, SRIO processing module, or XAUI processing module, it updates the start and end addresses of the PCIE processing module, SRIO processing module, or XAUI processing module to ensure that the currently transmitted space address is within the allocated address space. When the dynamic space control module receives the transmission completion signal sent by the PCIE processing module, the SRIO processing module or the XAUI processing module, the occupied memory space is released, and the start address and the end address of the PCIE processing module, the SRIO processing module or the XAUI processing module are updated.

Citation Information

Patent Citations

  • Ethernet exchange device based on RapidIO exchange technology

    CN106027424A

  • Method and apparatus for converting RapidIO message and Ethernet message

    CN105791214A

  • Packet conversion method and device

    CN107079017A