CN102308538A - Message processing method and device - Google Patents
Message processing method and device Download PDFInfo
- Publication number
- CN102308538A CN102308538A CN2011800011087A CN201180001108A CN102308538A CN 102308538 A CN102308538 A CN 102308538A CN 2011800011087 A CN2011800011087 A CN 2011800011087A CN 201180001108 A CN201180001108 A CN 201180001108A CN 102308538 A CN102308538 A CN 102308538A
- Authority
- CN
- China
- Prior art keywords
- message
- chip
- output port
- port
- virtual
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/74—Address processing for routing
- H04L45/742—Route cache; Operation thereof
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Multi Processors (AREA)
Abstract
The invention provides a message processing method and a device. The business process of the message is completed in the chip of the received message, and the chip is sent to the output port to complete the output. In this case, it is not required to maintain the Cache consistency between network chips. The system design is simple, and the processed message can be transmitted between network chips for one time at most, and the width demand and system consumption of network bus can be reduced.
Description
Technical field
The embodiment of the invention relates to communication technical field, especially a kind of message processing method and device.
Background technology
One of characteristics of network traffic are exactly that message is handled, and transmit after the routing.Export from the definite output port of routing after from the message of input port entering, as shown in Figure 1 through treatment steps such as decapsulation, routing, service quality (Quality Of Service is called for short QOS) control, encapsulation.
When carrying out power system capacity when expansion, can increase the input/output port order of system, the mode through two or more chip interconnect improves the system handles performance, as shown in Figure 2.See from system perspective, can export from the output port of forwarding decision after being processed from the message that arbitrary input port gets into, and keep the transmission and the characteristic of relevant parameters such as QOS at total system.In such scheme, need to keep cache memory (Cache) consistency of each chip.The Cache consistency is meant that a chip can be visited the memory space that preestablishes of other chips, and data consistency is guaranteed, and is equivalent to, and preserves a plurality of copies of same shared data among the Cache of each chip, and needs to keep these a plurality of copies consistent.The benefit of this scheme is; See that from the software angle data of independently two or more chip visits are that the overall situation is consistent; Need not guarantee the unanimity of data through mechanism such as software locks; Can adopt flexibly and implement completion (Run To Complete, be called for short RTC) or the mode of streamline (Pipeline) is disposed business.
But, for guaranteeing each chip chamber Cache consistency, need the operation affairs of labor to intercept in the above-mentioned existing scheme, this will influence the effective bandwidth of interconnection.
Summary of the invention
The embodiment of the invention provides a kind of message processing method and device, causes influencing the problem of the effective bandwidth of interconnection in order to solve existing chip interconnect system for guaranteeing each chip chamber Cache consistency.
On the one hand, the embodiment of the invention provides a kind of message processing method, comprising:
The first input end mouth receives first message, and said first message is transmitted to first Service Processing Unit, and said first Service Processing Unit and said first input end mouth are positioned at first chip;
Said first Service Processing Unit is handled said first message, generates second message, and the output port of confirming said second message is first output port;
If said first output port is positioned at second chip, then said second message is sent to said second chip;
Said second chip sends said second message through said first output port.
On the other hand, the embodiment of the invention provides a kind of message process device, comprising:
First chip and second chip;
Said first chip comprises:
The first input end mouth is used to receive first message, and said first message is transmitted to first Service Processing Unit;
First Service Processing Unit is used for said first message is handled, and generates second message, and the output port of confirming said second message is first output port; If said first output port is positioned at said second chip, then said second message is sent to said second chip;
Said second chip comprises said first output port, is used for sending said second message through said first output port.
More than a technical scheme in a plurality of technical schemes have following advantage or beneficial effect:
Chip was to accomplish the technological means of output under the embodiment of the invention was employed in and just sends to output port behind the Business Processing main process of completion to said message in the chip that receives message; The Cache consistency that need not keep interconnected chip chamber; System design realizes simple; And the message that is processed is meeting transmission primaries between interconnected chip at most only, has reduced bandwidth demand and system power dissipation to interconnection.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art; To do one to the accompanying drawing of required use in embodiment or the description of the Prior Art below introduces simply; Obviously; Accompanying drawing in describing below is some embodiments of the present invention; For those of ordinary skills; Under the prerequisite of not paying creative work property, can also obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is a kind of message transaction module sketch map of the prior art.
Fig. 2 is a kind of message transaction module sketch map when power system capacity is expanded in the prior art.
Fig. 3 is a kind of message transaction module sketch map of the embodiment of the invention.
The schematic flow sheet of a kind of message processing method embodiment one that Fig. 4 provides for the embodiment of the invention.
The schematic flow sheet of a kind of message processing method embodiment two that Fig. 5 provides for the embodiment of the invention.
The schematic flow sheet of a kind of message processing method embodiment three that Fig. 6 provides for the embodiment of the invention.
Fig. 7 is a kind of application sketch map embodiment illustrated in fig. 6.
Fig. 8 is that embodiment illustrated in fig. 6 another used sketch map.
The structural representation of a kind of message process device embodiment one that Fig. 9 provides for the embodiment of the invention.
The structural representation of a kind of message process device embodiment two that Figure 10 provides for the embodiment of the invention.
Embodiment
For the purpose, technical scheme and the advantage that make the embodiment of the invention clearer; To combine the accompanying drawing in the embodiment of the invention below; Technical scheme in the embodiment of the invention is carried out clear, intactly description; Obviously; Described embodiment is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills are not making the every other embodiment that is obtained under the creative work prerequisite, all belong to the scope of the present invention's protection.
For the embodiment of the invention being known detailed introduction, introduce some correlation techniques here earlier.In multicore sheet interconnected systems (being multiple nucleus system), professional processing mode generally can be divided into two kinds of RTC pattern and Pipeline patterns according to operation flow.Under the RTC pattern, single Service Processing Unit can be accomplished all processing of single business; Under the Pipeline pattern, the different phase that each processing unit is managed business respectively, each stage is accomplished Business Processing together.
Master-slave mode is adopted in the performance expansion between the chip interconnect system chips that has.Particularly, connect through high-speed serializer/deserializer (Serializer/Deserializer is called for short Serdes) or high speed parallel interface between the chip, be provided with direct memory access (Direct Memory Access is called for short DMA) engine at interface.One chip disposes the DMA engine then to the abundant memory space of an other chip application, and data are moved the memory space of an other chip from a chip, and the Service Processing Unit that starts final data place chip then carries out Business Processing.But there is following shortcoming in such scheme:
1) adopt master-slave mode work, need repeatedly communication between two or more chips, efficient is lower.
2) Network is transmitted based on stream or based on bag, and principal and subordinate's mode of operation can not be efficiently be incorporated into the Business Stream of a chip in the Business Stream of another one chip, both has been unfavorable for the service deployment of RTC mode, also is unfavorable for the service deployment of Pipeline mode.
3),, be unfavorable for that the QOS of whole system realizes owing to the overall real time information deviation that time-delay causes through in the system of Extended Capabilities Port expanding system capacity.
The mode that two or more chips flow according to Business Processing in the chip interconnect system that also has is together in series and carries out the Pipeline processing; Chip A accomplishes after the section processes; The intermediate object program of Business Processing is sent to chip B; Chip B is after accomplishing the partial service processing; The result is returned to chip A; Chip A accomplishes after next step processing, as result's output of system handles.This extended mode is applicable to that when single chip can't be accomplished the complicated business processing, carrying out cascade handles, and is suitable for the expansion of professional complex situations.But, under the situation that increases port number expanding system capacity, needing newly-increased chip is added in the Business Processing stream of existing chip, this will cause revising the main process of Business Processing, and will be bigger to service impact.
The mode that the embodiment of the invention adopts RTC and Pipeline to combine is carried out capacity extension.Single chip is accomplished this chip with the RTC mode and is input to professional output Business Processing before from port; When final output need be from the output of the port of another one chip; Then Pipeline exports to the another one chip, otherwise from the output of this chip, the message transaction module as shown in Figure 3.It is this according to business processing flow that the embodiment of the invention adopts, and the scheme of carrying out interconnected expansion by different level with the mode of complete equity has following beneficial effect:
1) service deployment of single-chip is transplanted conveniently to two or more chips, need not revise the main process of Business Processing.When single-chip traffic diversion multicore sheet is expanded, the just increase of Business Processing output port that single-chip is seen.
2) two or more chips are reciprocity fully, initiatively send data mutually, and the system handles time-delay is little, and efficient is high.
3) do not need to keep the Cache consistency between two or more chips, design realizes simple.
4) data message that is processed is meeting transmission primaries between interconnected chip at most only, reduces interconnected bandwidth demand of chip chamber and system power dissipation.
Particularly, the embodiment of the invention can realize through method shown in Figure 4.
The schematic flow sheet of a kind of message processing method embodiment one that Fig. 4 provides for the embodiment of the invention.As shown in Figure 4, this method comprises:
First message here can be any message that external system sends to the first input end mouth, and present embodiment does not limit this.
Step 402, said first Service Processing Unit are handled said first message, generate second message, and the output port of confirming said second message is first output port;
Here, first Service Processing Unit can adopt method of the prior art that said first message is handled, and the output port of confirming second message, and present embodiment does not limit this.In the application, all before first Service Processing Unit can be exported said first message with the RTC mode before the Business Processing are handled, and need second message of output until generation.Business Processing is accomplished data processing according to business demand before being included in output port before the output here, such as inserting timestamp, adds the message encapsulation format of corresponding output port etc.; Business Processing can not cause the change of the output port of message before the output.
Second chip here and first chip interconnect can be transmitted said second message through interconnecting interface, and interconnecting interface can be high speed Serdes or high speed parallel interface.If said first output port is positioned at first chip; Promptly be arranged in same chip with first input end mouth, first Service Processing Unit; Then can adopt the method for prior art to handle, such as directly in first chip, accomplishing output, present embodiment does not limit this.
In the application, said second chip can also carry out Business Processing before the aforesaid output to said second message that sends over from first chip, afterwards again through said first output port output.
Chip was to accomplish the technological means of output under the embodiment of the invention was employed in and just sends to output port behind the Business Processing main process of completion to said message in the chip that receives message; The Cache consistency that need not keep interconnected chip chamber; System design realizes simple; And the message that is processed is meeting transmission primaries between interconnected chip at most only, has reduced bandwidth demand and system power dissipation to interconnection.
The schematic flow sheet of a kind of message processing method embodiment two that Fig. 5 provides for the embodiment of the invention.As shown in Figure 5, this method comprises:
First output port here is the physical port of second chip, and the information of first output port comprises port-mark, information such as ports physical characteristic.
In the application; Can when system initialization, pass through the interconnecting interface physical port information of each interconnected chip synchronously; And in each interconnected chip the physical port corresponding virtual port of virtual generation and other interconnected chips, and with the information configuration of physical port in the corresponding virtual port.Certainly, also can be in advance in each interconnected chip the physical port corresponding virtual port of good and other the interconnected chips of manual configuration, present embodiment does not limit this.
The first input end mouth here is the physical port of first chip.
In the application, Service Processing Unit not only can be confirmed the output port of message, can also further correspond to the subqueue of output port, and corresponding to which subqueue here can confirm that present embodiment does not limit this according to the priority of business etc.
In the application, the virtual port of each interconnected chip can be converged to interconnecting interface, virtual port can be sent to other interconnected chips with message through interconnecting interface.Particularly, first virtual port here can send to second chip with second message through the interconnecting interface of first chip and second chip.
Particularly, the information of said first output port that in above-mentioned steps 501, gets access to comprises the port-mark of said first output port, and step 506 specifically can comprise:
Said second message is encapsulated, in said encapsulation, add the port-mark of said first output port;
Second message after the said encapsulation is sent to said second chip.
Accordingly, step 507 specifically can comprise:
Said second chip is resolved second message after the said encapsulation, obtains the port-mark of said first output port;
Send said second message through said first output port.
Need to prove that the port-mark of physical port is an overall identification, can physical port of unique identification.More excellent ground under the scene of a plurality of chip interconnect, also can confirm which interconnected chip the corresponding physical port is positioned at according to port-mark.For instance, each system one of being made up of two chip interconnect of 8 physical ports has 16 physical ports, the port-mark of the physical port of a chip can be made as 1～8 respectively, and the port-mark of the physical port of another chip is made as 9～16 respectively.
In order to cause flow pressure for other interconnected chips, each interconnected chip can also carry out traffic shaping earlier before the message that Service Processing Unit is handled is given to other interconnected chips.Under this scene, the information of said first output port comprises the band width in physical of said first output port, and step 506 specifically can comprise:
According to said band width in physical, the data flow of carrying said second message is carried out traffic shaping;
The data flow of said second message after through said shaping sent to said second chip.
More excellent ground, the flow bandwidth after the shaping is consistent with the band width in physical of first output port.
In the application, each interconnected chip is realized formation usually on the hardware logic of this chip or institute's operating software, more excellent ground, the individual queue numbering corresponding the port-mark of physical port.For instance, need be that the message that 1 physical port sends is put into formation 1 from port-mark.Under this scene, can also comprise after the step 504:
Said first Service Processing Unit joins said second message in the formation of said first virtual port;
The formation of said first virtual port is sent to said second chip;
Step 507 specifically can comprise:
According to the task of sending said second message in the formation of said first virtual port, said second chip sends said second message through said first output port.
Here; For first chip; Self with the mode of software queue or hardware queue virtual the interconnected right physical port of second chip; Make E-Packeting of self in processing procedure, see all of the port of the interconnected expansion of all entire chip back system; Message outlet after finish dealing with is on certain virtual port the time, the formation that this message transmission task can be put into this virtual port.For instance; The port-mark of first output port is 10; Then the port-mark of first virtual port also is 10 in first chip; First chip can be put into formation 10 with second message; Because formation 10 correspondences is virtual port; First chip is given second chip synchronously with formation 10, and second chip is accomplished the task of sending second message through first output port according to formation 10.
In the application; Second chip is when sending said second message through first output port; Maybe be also need send the message after the second chip own service processing unit processes through first output port; Under this scene, Business Processing and Port Scheduling before can exporting in the lump these messages.Particularly, can also comprise before the step 507:
Second input port receives the 3rd message, and said the 3rd message is transmitted to second Service Processing Unit, and said second Service Processing Unit and said second input port are located on said second chip;
Said second Service Processing Unit is handled said the 3rd message, generates the 4th message, and the output port of confirming said the 4th message is said first output port;
Step 507 specifically can comprise:
Said second chip is dispatched said second message and the 4th message;
Said first output port sends said second message and the 4th message according to said scheduling.
In the application; Second Service Processing Unit can be put into the 4th message the formation of first output port; Business Processing and scheduling before the second chip unification is exported the message in the formation of first output port then finally send through first output port.
The embodiment of the invention is employed in the physical port of virtual other interconnected chips in each interconnected chip; Make local terminal in the business processing flow of message, can see all of the port of whole system; After the message that receives is accomplished the Business Processing main process; Output port according to this message is the technological means that virtual port sends to this message the corresponding physical port of other interconnected chips; Make when single-chip traffic diversion multicore sheet is expanded; The just increase of Business Processing output port that this chip is seen; Service deployment is transplanted convenient; Need not revise the Business Processing main process; The Cache uniformity that need not keep interconnected chip chamber; System's design realizes simple; The message that is processed is meeting transmission primaries between interconnected chip at most only, has reduced bandwidth demand and system power dissipation to interconnection.
The schematic flow sheet of a kind of message processing method embodiment three that Fig. 6 provides for the embodiment of the invention.To be example based on chip A and chip B system for interconnecting, as shown in Figure 6, this method comprises:
The formation of step 611 and other virtual ports of chip A is dispatched together, converges to interconnecting interface;
Fig. 7 is a kind of application sketch map of present embodiment.As shown in Figure 7, the system of chip A and chip B composition is divided into following several level:
Level 1, input physical port.The port input data of each chip are given this chip Service Processing Unit and are begun to handle in this chip porch.
Level 2, Business Processing.Service Processing Unit is handled the business datum of this chip input with the RTC mode; Accomplish the main body Business Processing; And system's output port of decision communication message, outlet possibly be the physical port of this chip or the physical port of opposite end, or the subqueue on the corresponding port.
Level 3, global port and formation.Comprise formation and port subqueue, the formation of local physical port and the port subqueue of the physical port of the interconnected chip in opposite end at the virtual port of this chip.
Level 4, go out Business Processing and Port Scheduling before the output before the physical port.Before output port, accomplish data processing,, add the message encapsulation format of corresponding physical port etc. such as inserting timestamp according to business demand.Send after the formation of local physical port and the queue scheduling of virtual port.These handle the change that can not cause the message output port.
Level 5, output physical port.From system, send the message that processing finishes.
Wherein, the virtual port back of local terminal chip is provided with shaping unit, and bandwidth traffic is carried out shaping.Flow bandwidth after the shaping is consistent with the actual physical bandwidth of actual physics output port.When data are transmitted at high-speed interface, data to be transmitted is encapsulated, comprise the information field section of the corresponding actual physics output port of current transmission data in the encapsulation.
In communication message was handled, the certain applications meeting is differentiated service priority, subport, child user etc. on output port, was used for realizing more fine-grained QOS function.Through sync cap shown in Figure 7; Chip A and chip B periodically transmit the chip internal state information, comprise the waiting task quantity of formation, the subqueue on the physical port, the subqueue on the virtual port of formation, the virtual port of physical port, send state and statistical informations such as data bulk.Waiting task quantity addition in the subqueue that the opposite end chip is sended in waiting task quantity and the local corresponding subqueue; The approximate waiting task quantity that obtains corresponding this subqueue of total system, these data can be used to carry out the professional back-pressure, packet loss of congested control etc.Sent the data bulk of the transmission addition in data bulk and the local corresponding subqueue in the subqueue that the opposite end chip is sended over, the approximate data bulk of transmission that obtains corresponding this subqueue of total system.
High-speed interface shown in Figure 7 is the shared interface of the virtual port of each interconnected chip; Virtual port and corresponding physical port communication can use based on the Flow Control mode of credit (Credit-based) and carry out mutual transfer of data; Take high-speed interface when avoiding a pair of virtual port to communicate by letter, make other ports can not get guaranteeing communication with physical port.
Fig. 8 is another application sketch map of present embodiment.Different with application shown in Figure 7 is, should with in sync cap and high-speed interface are combined, accomplish simultaneously at an interconnecting interface.In interconnecting interface, divide the transmission channel of different priorities, the data that sync cap is transmitted are through the high-priority channel transmission.
Chip was to accomplish the technological means of output under the embodiment of the invention was employed in and just sends to output port behind the Business Processing main process of completion to said message in the chip that receives message; The Cache consistency that need not keep interconnected chip chamber; System design realizes simple; And the message that is processed is meeting transmission primaries between interconnected chip at most only, has reduced bandwidth demand and system power dissipation to interconnection.
The structural representation of a kind of message process device embodiment one that Fig. 9 provides for the embodiment of the invention.As shown in Figure 9, this device comprises:
First input end mouth 911 is used to receive first message, and said first message is transmitted to first Service Processing Unit 912;
First Service Processing Unit 912 is used for said first message is handled, and generates second message, and the output port of confirming said second message is first output port 921; If first output port 921 is positioned at second chip 92, then said second message is sent to second chip 92;
A kind of message processing method embodiment one that the concrete realization of present embodiment provides with reference to the embodiment of the invention.
Chip was to accomplish the technological means of output under the embodiment of the invention was employed in and just sends to output port behind the Business Processing main process of completion to said message in the chip that receives message; The Cache consistency that need not keep interconnected chip chamber; System design realizes simple; And the message that is processed is meeting transmission primaries between interconnected chip at most only, has reduced bandwidth demand and system power dissipation to interconnection.
The structural representation of a kind of message process device embodiment two that Figure 10 provides for the embodiment of the invention.As shown in figure 10, this device comprises:
First input end mouth 111 is used to receive first message, and said first message is transmitted to first Service Processing Unit 112;
First Service Processing Unit 112 is used for obtaining from second chip 12 information of first output port; According to the information of said first output port, first virtual port 113 that virtual generation is corresponding with first output port 121; Said first message is handled, generated second message, the output port of confirming said second message is first virtual port 113, and said second message is sent to first virtual port 113;
First virtual port 113 is used for said second message is sent to second chip 12;
In an optional embodiment of the present invention, the information of said first output port comprises the port-mark of first output port 121, and first virtual port 113 specifically is used for,
Said second message is encapsulated, in said encapsulation, add the port-mark of first output port 121;
Second message after the said encapsulation is sent to second chip 12.
Accordingly, second chip 12 specifically is used for,
Resolve second message after the said encapsulation, obtain the port-mark of said first output port 121, send said second message through first output port 121.
In another optional embodiment of the present invention, the information of said first output port comprises the band width in physical of said first output port, and first virtual port 113 specifically is used for,
According to said band width in physical, said second message is carried out traffic shaping;
Second message behind the traffic shaping is sent to second chip 12.
In another optional embodiment of the present invention, first Service Processing Unit 112 also is used for,
Said second message is joined in the formation of first virtual port 113;
The formation of said first virtual port 113 is sent to second chip 12;
In another optional embodiment of the present invention, second chip 12 also comprises:
Second Service Processing Unit 123 is used for said the 3rd message is handled, and generates the 4th message, and the output port of confirming said the 4th message is first output port 121;
A kind of message processing method embodiment one, two or three that the concrete realization of present embodiment provides with reference to the embodiment of the invention.The embodiment of the invention is employed in the physical port of virtual other interconnected chips in each interconnected chip; Make local terminal in the business processing flow of message, can see all of the port of whole system; After the message that receives is accomplished the Business Processing main process; Output port according to this message is the technological means that virtual port sends to this message the corresponding physical port of other interconnected chips; Make when single-chip traffic diversion multicore sheet is expanded; The just increase of Business Processing output port that this chip is seen; Service deployment is transplanted convenient; Need not revise the Business Processing main process; The Cache uniformity that need not keep interconnected chip chamber; System's design realizes simple; The message that is processed is meeting transmission primaries between interconnected chip at most only, has reduced bandwidth demand and system power dissipation to interconnection.
One of ordinary skill in the art will appreciate that: all or part of step that realizes said method embodiment can be accomplished through the relevant hardware of program command; Aforesaid program can be stored in the computer read/write memory medium; This program the step that comprises said method embodiment when carrying out; And aforesaid storage medium comprises: various media that can be program code stored such as ROM, RAM, magnetic disc or CD.
What should illustrate at last is: above embodiment is only in order to illustrating technical scheme of the present invention, but not to its restriction; Although with reference to previous embodiment the present invention has been carried out detailed explanation, those of ordinary skill in the art is to be understood that: it still can be made amendment to the technical scheme that aforementioned each embodiment put down in writing, and perhaps part technical characterictic wherein is equal to replacement; And these are revised or replacement, do not make the spirit and scope of the essence disengaging various embodiments of the present invention technical scheme of relevant art scheme.
Claims (12)
1. a message processing method is characterized in that, comprising:
The first input end mouth receives first message, and said first message is transmitted to first Service Processing Unit, and said first Service Processing Unit and said first input end mouth are positioned at first chip;
Said first Service Processing Unit is handled said first message, generates second message, and the output port of confirming said second message is first output port;
If said first output port is positioned at second chip, then said second message is sent to said second chip;
Said second chip sends said second message through said first output port.
2. method according to claim 1 is characterized in that, said first input end mouth also comprises before receiving first message:
Obtain the information of said first output port from said second chip;
According to the information of said first output port, first virtual port that virtual generation is corresponding with said first output port;
The output port of said definite said second message is that first output port specifically comprises:
The output port of confirming said second message is said first virtual port;
Saidly said second message sent to said second chip specifically comprise:
Said second message is sent to said first virtual port;
Said first virtual port sends to said second chip with said second message.
3. method according to claim 2 is characterized in that the information of said first output port comprises the port-mark of said first output port, and said first virtual port sends to said second chip with said second message and specifically comprises:
Said second message is encapsulated, in said encapsulation, add the port-mark of said first output port;
Second message after the said encapsulation is sent to said second chip;
Said second chip sends said second message through said first output port and specifically comprises:
Said second chip is resolved second message after the said encapsulation, obtains the port-mark of said first output port;
Send said second message through said first output port.
4. method according to claim 2, the information of said first output port comprises the band width in physical of said first output port, and said first virtual port sends to said second chip with said second message and specifically comprises:
According to said band width in physical, the data flow of carrying said second message is carried out traffic shaping;
The data flow of said second message after through said shaping sent to said second chip.
5. method according to claim 2 is characterized in that, the output port of said definite said second message is that first output port also comprises afterwards:
Said first Service Processing Unit joins said second message in the formation of said first virtual port;
The formation of said first virtual port is sent to said second chip;
Said second chip sends said second message through said output port and specifically comprises:
According to the formation of said first virtual port, said second chip sends said second message through said first output port.
6. method according to claim 1 is characterized in that, said second chip also comprises before sending said second message through said first output port:
Second input port receives the 3rd message, and said the 3rd message is transmitted to second Service Processing Unit, and said second Service Processing Unit and said second input port are positioned at said second chip;
Said second Service Processing Unit is handled said the 3rd message, generates the 4th message, and the output port of confirming said the 4th message is said first output port;
Said second chip sends said second message through said first output port and specifically comprises:
Said second chip is dispatched said second message and the 4th message;
Said first output port sends said second message and the 4th message according to said scheduling.
7. a message process device is characterized in that, comprising: first chip and second chip;
Said first chip comprises:
The first input end mouth is used to receive first message, and said first message is transmitted to first Service Processing Unit;
First Service Processing Unit is used for said first message is handled, and generates second message, and the output port of confirming said second message is first output port; If said first output port is positioned at said second chip, then said second message is sent to said second chip;
Said second chip comprises said first output port, is used for sending said second message through said first output port.
8. device according to claim 7 is characterized in that, said first Service Processing Unit specifically is used for:
Obtain the information of said first output port from said second chip;
According to the information of said first output port, first virtual port that virtual generation is corresponding with said first output port;
The output port of confirming said second message is said first virtual port, and said second message is sent to said first virtual port;
Said first chip also comprises:
First virtual port is used for said second message is sent to said second chip.
9. device according to claim 8 is characterized in that the information of said first output port comprises the port-mark of said first output port, and said first virtual port specifically is used for,
Said second message is encapsulated, in said encapsulation, add the port-mark of said first output port;
Second message after the said encapsulation is sent to said second chip;
Said second chip specifically is used for,
Resolve second message after the said encapsulation, obtain the port-mark of said first output port, send said second message through said first output port.
10. device according to claim 8 is characterized in that the information of said first output port comprises the band width in physical of said first output port, and said first virtual port specifically is used for,
According to said band width in physical, the data flow of carrying said second message is carried out traffic shaping;
The data flow of said second message after through said shaping sent to said second chip.
11. device according to claim 8 is characterized in that, said first Service Processing Unit also is used for,
Said second message is joined in the formation of said first virtual port;
The formation of said first virtual port is sent to said second chip;
Said second chip specifically is used for, and according to the formation of said first virtual port, sends said second message through said first output port.
12. device according to claim 7 is characterized in that, said second chip also comprises:
Second input port is used to receive the 3rd message, and said the 3rd message is transmitted to second Service Processing Unit;
Second Service Processing Unit is used for said the 3rd message is handled, and generates the 4th message, and the output port of confirming said the 4th message is said first output port;
The Port Scheduling unit is used for said second message and the 4th message are dispatched;
Said first output port specifically is used for, and sends said second message and the 4th message according to said scheduling.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2011/077361 WO2012106905A1 (en) | 2011-07-20 | 2011-07-20 | Message processing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102308538A true CN102308538A (en) | 2012-01-04 |
| CN102308538B CN102308538B (en) | 2014-11-05 |
Family
ID=45381271
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201180001108.7A Active CN102308538B (en) | 2011-07-20 | 2011-07-20 | Message processing method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN102308538B (en) |
| WO (1) | WO2012106905A1 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014187428A1 (en) * | 2013-12-11 | 2014-11-27 | 中兴通讯股份有限公司 | Method for realizing message forwarding and switch |
| CN104486261A (en) * | 2014-12-19 | 2015-04-01 | 盛科网络（苏州）有限公司 | Interlaken interface-based two-chip interconnection method |
| CN113504985A (en) * | 2016-07-29 | 2021-10-15 | 华为技术有限公司 | Task processing method and network equipment |
| CN115114200A (en) * | 2022-06-29 | 2022-09-27 | 海光信息技术股份有限公司 | Multi-chip system and starting method based on same |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7069372B1 (en) * | 2001-07-30 | 2006-06-27 | Cisco Technology, Inc. | Processor having systolic array pipeline for processing data packets |
| CN1866845A (en) * | 2006-05-08 | 2006-11-22 | 国家数字交换系统工程技术研究中心 | Virtual identifying method for multicast forwarding table output port |
| CN101089829A (en) * | 2007-08-01 | 2007-12-19 | 杭州华三通信技术有限公司 | Shared buffer store system and implementing method |
| CN201051079Y (en) * | 2007-06-05 | 2008-04-23 | 浙江双元科技开发有限公司 | High-speed processing module for paper disease image data |
| CN101296186A (en) * | 2008-06-13 | 2008-10-29 | 杭州华三通信技术有限公司 | Routing switching device, method and switching cable fastener plate |
| CN201374784Y (en) * | 2009-03-13 | 2009-12-30 | 深圳市泛海三江电子有限公司 | Eight-channel embedded type digital video recorder |
| CN101719877A (en) * | 2010-01-15 | 2010-06-02 | 福建星网锐捷网络有限公司 | Message forwarding device, network equipment and method |
| CN101808428A (en) * | 2010-04-21 | 2010-08-18 | 华为终端有限公司 | Communication method and device of double-card dual-standby cell phone |
-
2011
- 2011-07-20 WO PCT/CN2011/077361 patent/WO2012106905A1/en active Application Filing
- 2011-07-20 CN CN201180001108.7A patent/CN102308538B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7069372B1 (en) * | 2001-07-30 | 2006-06-27 | Cisco Technology, Inc. | Processor having systolic array pipeline for processing data packets |
| CN1866845A (en) * | 2006-05-08 | 2006-11-22 | 国家数字交换系统工程技术研究中心 | Virtual identifying method for multicast forwarding table output port |
| CN201051079Y (en) * | 2007-06-05 | 2008-04-23 | 浙江双元科技开发有限公司 | High-speed processing module for paper disease image data |
| CN101089829A (en) * | 2007-08-01 | 2007-12-19 | 杭州华三通信技术有限公司 | Shared buffer store system and implementing method |
| CN101296186A (en) * | 2008-06-13 | 2008-10-29 | 杭州华三通信技术有限公司 | Routing switching device, method and switching cable fastener plate |
| CN201374784Y (en) * | 2009-03-13 | 2009-12-30 | 深圳市泛海三江电子有限公司 | Eight-channel embedded type digital video recorder |
| CN101719877A (en) * | 2010-01-15 | 2010-06-02 | 福建星网锐捷网络有限公司 | Message forwarding device, network equipment and method |
| CN101808428A (en) * | 2010-04-21 | 2010-08-18 | 华为终端有限公司 | Communication method and device of double-card dual-standby cell phone |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014187428A1 (en) * | 2013-12-11 | 2014-11-27 | 中兴通讯股份有限公司 | Method for realizing message forwarding and switch |
| CN104717138A (en) * | 2013-12-11 | 2015-06-17 | 中兴通讯股份有限公司 | Method of realizing message forwarding and switch |
| CN104717138B (en) * | 2013-12-11 | 2019-07-12 | 中兴通讯股份有限公司 | A kind of method and interchanger for realizing message forwarding |
| CN104486261A (en) * | 2014-12-19 | 2015-04-01 | 盛科网络（苏州）有限公司 | Interlaken interface-based two-chip interconnection method |
| CN113504985A (en) * | 2016-07-29 | 2021-10-15 | 华为技术有限公司 | Task processing method and network equipment |
| CN113504985B (en) * | 2016-07-29 | 2022-10-11 | 华为技术有限公司 | Task processing method and network equipment |
| CN115114200A (en) * | 2022-06-29 | 2022-09-27 | 海光信息技术股份有限公司 | Multi-chip system and starting method based on same |
| CN115114200B (en) * | 2022-06-29 | 2023-11-17 | 海光信息技术股份有限公司 | Multi-chip system and starting method based on same |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2012106905A1 (en) | 2012-08-16 |
| CN102308538B (en) | 2014-11-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10027433B2 (en) | Multiple clock domains in NoC | |
| US9860197B2 (en) | Automatic buffer sizing for optimal network-on-chip design | |
| US9426099B2 (en) | Router, method for controlling router, and program | |
| CN103348640B (en) | Relay | |
| CN100407701C (en) | Network processor | |
| US9356881B2 (en) | Traffic scheduling device | |
| CN101199170B (en) | Packet processing switch and methods of operation thereof | |
| CN100527697C (en) | Means and a method for switching data packets or frames | |
| JP4763405B2 (en) | Network-on-chip semi-automatic communication architecture for data flow applications | |
| CN110995598B (en) | Variable-length message data processing method and scheduling device | |
| CN101001209B (en) | System for switching variable-length data packets of heterogeneous network and method thereof and method for forming address list using signal loop interface | |
| CN101729407B (en) | Low delay jitter exchanging method and equipment based on unicast and multicast differentiated treatment | |
| US8949578B2 (en) | Sharing of internal pipeline resources of a network processor with external devices | |
| WO2022094771A1 (en) | Network chip and network device | |
| CN102308538B (en) | Message processing method and device | |
| CN105095150A (en) | Network interface supporting network-on-chip | |
| CN103530245A (en) | SRIO interconnection exchanging device based on field programmable gate array (FPGA) | |
| CN104243348A (en) | Data processing method and device | |
| WO2023029487A1 (en) | Method and apparatus for determining topological structure of network-on-chip, and chip | |
| CN103902505A (en) | Configurable FFT processor circuit structure based on switching network | |
| CN104009942A (en) | Kilomega AFDX switch and switching method thereof | |
| CN114679415A (en) | Non-blocking banyan network meeting AXI5-Lite protocol standard | |
| CN103347077A (en) | Method and device for data transmission of distributed file system | |
| US7398501B2 (en) | System and method for optimizing an integrated circuit design | |
| US8645557B2 (en) | System of interconnections for external functional blocks on a chip provided with a single configurable communication protocol |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |