CN116055381A - Routing table updating method and device based on many-core system and electronic equipment - Google Patents

Abstract

The present disclosure provides a routing table updating method, a system, an electronic device and a readable medium based on a many-core system, wherein the method comprises: under the condition that the update condition of the routing table is met, determining routing transmission information between the first processing core and the second processing core according to the current physical position of the first processing core and the current physical position of the second processing core; updating a routing table corresponding to the first processing core according to the routing transmission information; wherein the second processing core is a processing core that receives the first processing core packet, and the routing table update condition includes at least one of: the routing table corresponding to the first processing core is empty, and the physical position of the first processing core and/or the physical position of the second processing core are changed. The method can flexibly set the route transmission information, thereby being convenient for dynamically adjusting the route and being flexibly applicable to complex application scenes in which the physical position of the processing core can be dynamically changed.

Description

Routing table update method, device and electronic equipment based on many-core system

technical field

Embodiments of the present disclosure relate to the field of computer technology, and in particular to a method, device and electronic device for updating a routing table based on a many-core system.

Background technique

Many-core architecture is a parallel processing architecture widely used to execute neural network models. In the many-core architecture, each processing core can perform a certain computing function, a certain number of processing cores are connected through a certain topology to form a chip, a certain number of chips are connected through a certain topology to form a chip array board, and so on , can be extended to a larger system. It can be seen that in a many-core system, a large number of processing cores are involved. Therefore, how to realize efficient communication between each processing core under the premise of a large number of processing cores has become an urgent technical problem in the field of routing computing.

Contents of the invention

In view of the above problems, the present disclosure is proposed to provide a method, device and electronic device for updating a routing table based on a many-core system that overcomes the above problems or at least partially solves the above problems.

According to an aspect of an embodiment of the present disclosure, a method for updating a routing table based on a many-core system is provided, including:

If the routing table update condition is met, determine the route between the first processing core and the second processing core according to the current physical location of the first processing core and the current physical location of the second processing core routing information;

updating the routing table corresponding to the first processing core according to the routing transmission information;

Wherein, the second processing core is a processing core that receives the first processing core data packet, and the routing table update condition includes at least one of the following: the routing table corresponding to the first processing core is empty, the first processing core and/or the physical location of the second processing core changes.

In a second aspect, an embodiment of the present disclosure provides an apparatus for updating a routing table based on a many-core system, including:

The determining module is adapted to determine the first processing core and the second processing core according to the current physical location of the first processing core and the current physical location of the second processing core when the routing table update condition is satisfied. Handle routing information between cores;

An update module, adapted to update the routing table corresponding to the first processing core according to the routing transmission information;

Wherein, the second processing core is a processing core that receives the first processing core data packet, and the routing table update condition includes at least one of the following: the routing table corresponding to the first processing core is empty, the first processing core and/or the physical location of the second processing core changes.

In a third aspect, an embodiment of the present disclosure provides an electronic device, which includes:

one or more processors;

A storage device, on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement at least one of the following methods:

According to the routing table updating method described in the first aspect of the embodiments of the present disclosure.

In a fourth aspect, an embodiment of the present disclosure provides a computer-readable medium on which a computer program is stored, and when the program is executed by a processor, at least one of the following methods is implemented:

According to the routing table updating method described in the first aspect of the embodiments of the present disclosure.

In the routing table update method, device, and electronic device based on the many-core system provided by the embodiments of the present disclosure, when the routing table update condition is met, the current physical location of the first processing core and the current location of the second processing core can be The physical location determines routing transmission information between the first processing core and the second processing core; and updates a routing table corresponding to the first processing core according to the routing transmission information. Wherein, the routing table update condition includes at least one of the following: the routing table corresponding to the first processing core is empty, and the physical location of the first processing core and/or the second processing core changes. It can be seen that by setting the routing table update conditions, the routing transmission information between the two processing cores can be obtained in time when the routing table needs to be updated, and the routing table can be updated based on the routing transmission information, which can ensure the timeliness and accuracy of the routing table update In this way, efficient communication between each processing core can be realized under the premise of a large number of processing cores. In addition, by updating the routing table when the physical location of the first processing core and/or the second processing core changes, it can be flexibly adapted to the application scenario where the location of the processing core changes.

Description of drawings

The accompanying drawings are used to provide a further understanding of the embodiments of the present disclosure, and constitute a part of the specification, and are used together with the embodiments of the present disclosure to explain the present disclosure, and do not constitute limitations to the present disclosure. The above and other features and advantages will become more apparent to those skilled in the art by describing detailed example embodiments with reference to the accompanying drawings, in which:

FIG. 1 is a flowchart of a method for updating a routing table based on a many-core system according to an embodiment of the present disclosure;

FIG. 2 is a flowchart of a method for updating a routing table based on a many-core system according to another embodiment of the present disclosure;

FIG. 3 is a structural diagram of an apparatus for updating a routing table based on a many-core system according to yet another embodiment of the present disclosure;

FIG. 4 is a composition block diagram of an electronic device provided by an embodiment of the present disclosure;

Fig. 5 is a composition block diagram of a computer-readable medium provided by an embodiment of the present disclosure.

Detailed ways

In order for those skilled in the art to better understand the technical solution of the present disclosure, the routing method, system, electronic device and computer-readable medium provided by the present disclosure will be described in detail below with reference to the accompanying drawings.

Example embodiments will be described more fully hereinafter with reference to the accompanying drawings, but may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

In the case of no conflict, various embodiments of the present disclosure and various features in the embodiments can be combined with each other.

As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

The terminology used herein is for describing particular embodiments only and is not intended to limit the present disclosure. As used herein, the singular forms "a" and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise. It will also be understood that when the terms "comprising" and/or "consisting of" are used in this specification, the stated features, integers, steps, operations, elements and/or components are specified to be present but not excluded to be present or Add one or more other features, integers, steps, operations, elements, components and/or groups thereof.

Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant art and the present disclosure, and will not be interpreted as having idealized or excessive formal meanings, Unless expressly so limited herein.

In related technologies, routing precompilation is usually used to determine routing modes between processing cores. However, the inventor found in the process of implementing the present invention that the above-mentioned pre-compilation method has at least the following defects: the routing flexibility set by the pre-compilation method is poor, once the compilation is completed, it cannot be dynamically adjusted, and cannot be applied to the physical location of the processing core. Complex application scenarios that can change dynamically.

An embodiment of the present disclosure provides a method for updating a routing table based on a many-core system. In the method provided in this embodiment, one chip in the many-core system includes multiple processing cores capable of communicating with each other, and the physical location of each processing core can be dynamically adjusted according to the running state of the application. Fig. 1 shows a schematic flowchart of a method for updating a routing table based on a many-core system provided by an embodiment of the present disclosure. As shown in Fig. 1 , the method includes:

Step S110: If the routing table update condition is met, determine routing transmission information between the first processing core and the second processing core according to the current physical location of the first processing core and the current physical location of the second processing core.

Wherein, the routing table update condition includes at least one of the following: the routing table corresponding to the first processing core is empty, and the physical location of the first processing core and/or the second processing core changes. For example, when the routing table corresponding to the first processing core is empty, it means that the routing table needs to be updated; The stored routing information is out of date, requiring an update to the routing table.

Correspondingly, according to the current physical location of the first processing core and the current physical location of the second processing core, route transmission information between the first processing core and the second processing core is determined. Wherein, the second processing core is a processing core that receives the data packet of the first processing core. When determining the route transmission information between the first processing core and the second processing core, it can be implemented in multiple ways. For example, in the first way, the route transmission information between the first processing core and the second processing core is determined through path calculation; as another example, in the second way, the first Routing transfers information between the processing core and the second processing core.

In addition, when the routing table update condition is that the routing table corresponding to the first processing core is empty, the current physical location of the second processing core can be determined by: broadcasting a location query message to the second processing core; receiving the second processing core The location response message sent by the core, and the current physical location of the second processing core is determined according to the location response message.

Step S120: Update the routing table corresponding to the first processing core according to the routing transmission information.

Specifically, when the routing table is empty, the routing table corresponding to the first processing core may be updated directly according to the routing transmission information. In addition, when the routing table is not empty, you can also query the stored routing information between the first processing core and the second processing core in the routing table corresponding to the first processing core; An update operation; wherein, the update operation includes at least one of the following: a replace class update operation, and an add class update operation.

It can be seen that by setting the routing table update conditions, the routing transmission information between the two processing cores can be obtained in time when the routing table needs to be updated, and the routing table can be updated based on the routing transmission information, which can ensure the timeliness and accuracy of the routing table update In this way, efficient communication between each processing core can be realized under the premise of a large number of processing cores. In addition, by updating the routing table when the physical location of the first processing core and/or the second processing core changes, it can be flexibly adapted to the application scenario where the location of the processing core changes.

Fig. 2 shows a schematic flowchart of a method for updating a routing table based on a many-core system provided by another embodiment of the present disclosure. In the embodiment shown in FIG. 2 , a many-core chip is provided with multiple processing cores capable of communicating with each other. Each processing core has a logical address (also called a logical identifier) and a physical address (also called a physical identifier). The logical address is used to uniquely identify a processing core from the perspective of the business function of the processing core. The angle of the core's actual physical location to uniquely identify a processing core. In this embodiment, the physical addresses of the processing cores identified by the same logical address can change dynamically as the running state of the application changes. As shown in Figure 2, the method includes:

Step S200: Set up a routing table according to the positional relationship between each processing core.

This step is an optional step. In actual situations, this step can also be omitted, and the routing table can be dynamically created directly during the routing process. Specifically, in order to facilitate communication, a routing table is set in each processing core, and the routing table is used to store the transmission path and/or transmission location when the processing core communicates with other processing cores. Among them, the transmission path is used to describe the specific path information, specifically including the position information and transmission order information of each processing core that the data packet passes through during the transmission process; the transmission position is used to describe the relative The location offset of the current processing core (that is, the current starting address).

Specifically, the routing table is empty in the initial state, and when the current processing core needs to send a data packet to other processing cores, the route is calculated according to the physical location of the target processing core (ie, the second processing core) used to receive the data packet The information is transmitted and updated in the routing table.

In this embodiment, since each processing core is independent of each other, in order to improve communication efficiency and avoid query delay, a routing table is separately maintained inside each processing core, that is, there is a one-to-one correspondence between the routing table and the processing cores. Of course, in other embodiments of the present disclosure, the routing table related to each processing core can also be stored in a processing core (referred to as the routing management processing core) for implementing the routing management function, and other processing cores can communicate with the routing table. The communication between the management processing cores implements the query and update operations on the routing table.

Step S210: According to the historical communication state of the first processing core, adjust the physical position of the first processing core, and broadcast the update message of the physical position of the first processing core to other processing cores.

Wherein, this step can be performed at any timing, that is, the present invention does not limit the sequence between this step and other steps.

Among them, the historical communication state is used to describe information transmission path, information transmission delay, information transmission frequency, information transmission quantity, information transmission type, etc. between a processing core and other processing cores. Correspondingly, the current processing core (that is, the first processing core) dynamically detects at least one of the following data: information transmission delay between this processing core and each processing core, information transmission frequency, information transmission quantity, and information transmission type , and determine whether to adjust the physical position of the processing core according to the detection result. Specifically, when the information transmission delay between one processing core and other processing cores exceeds the preset delay threshold, the physical location of the relevant processing core can be dynamically adjusted so that the transmission time between the two processing cores communicating with each other It is also possible to dynamically adjust the physical location of the relevant processing cores when the frequency of information transmission between one processing core and other processing cores exceeds the preset frequency threshold and/or the number of information transmissions exceeds the preset number threshold, so that the mutual The communication between the two processing cores of the communication is more convenient; when the information transmission type between one processing core and other processing cores belongs to the preset type (such as information types requiring higher delay and reliability), Dynamically adjust the physical location of the relevant processing cores.

For example, the current processing core dynamically detects the information transmission delay, information transmission frequency, information transmission quantity, and information transmission type between the processing core and each processing core, and determines the corresponding communication processing core ( That is: the processing core that communicates more with the current processing core); when it is determined that the communication processing core has a corresponding adjacent idle core, update the physical location of the current processing core according to the physical location of the adjacent idle core. When updating the physical location of the processing core according to the physical location of the adjacent idle core, the physical address of the processing core can be directly updated to the physical address of the adjacent idle core, so that the logical address of the adjacent idle core is the same as that of the processing core. The logical addresses of the two nodes are the same, so as to achieve the effect of moving the current processing core to the adjacent idle core.

Specifically, when the physical location of the current processing core is updated, a physical location update message may be sent to other processing cores by broadcasting, so that other processing cores can determine the latest location of the current processing core according to the physical location update message. During specific implementation, the current processing core may directly broadcast the physical location update message to other processing cores; or, the current processing core may first send the physical location update message to the controller or the processing core used to implement the control function (which may be The routing management processing core mentioned above), and then the controller or the processing core for realizing the control function broadcasts the physical location update message to other processing cores. In a word, the way of global broadcasting can ensure that the position change of any processing core can be sensed by other processing cores in time.

Correspondingly, the current processing core is further configured to: receive the physical position update message of the second processing core sent by the second processing core, the physical position update message is that the second processing core adjusts the second processing core according to the historical communication state of the second processing core. Message broadcast after processing the physical location of a core. For example, after any processing core receives a broadcasted physical location update message, it parses the physical location update message to obtain the identification information (such as the logical address mentioned above) of the processing core whose location has changed and the updated physical location update message contained therein. Location. In order to facilitate querying the above location update information in the subsequent routing process, the current processing core can uniformly store the analysis results corresponding to the received physical location update messages in the storage unit; or, the current processing core can also traverse the routing table Store each piece of routing information related to the processing core whose location has changed, and set the routing information related to the processing core whose location has changed to an expired or invalid state, so as to update the routing table in the subsequent routing process . When the routing dependency between the first processing core and the second processing core changes, that is, when the physical location of the first processing core changes and/or the physical location of the second processing core changes, it is determined that the routing table needs to be updated. In an actual situation, the position of one processing core among the first processing core and the second processing core may change, while the position of the other processing core remains unchanged; it may also be that the positions of the two processing cores change simultaneously.

Correspondingly, in this step, it is necessary to detect whether the physical positions of the first processing core and the second processing core change, so as to obtain the latest physical position after the change. Wherein, if the physical position of the first processing core changes, the current physical position of the first processing core refers to the changed physical position; similarly, if the physical position of the second processing core changes, the current physical position of the second processing core The current physical location refers to a changed physical location.

During specific implementation, the current physical locations of the first processing core and the second processing core can be determined according to the above-mentioned physical location update message and the storage unit; or, by querying the above-mentioned routing table, according to the route It is determined whether the relevant routes stored in the table are set with an expired flag or an invalid flag, and this disclosure does not limit the specific details.

In addition, in an actual situation, the current physical location of the second processing core may also be determined by means of broadcasting and sending an addressing message by the first processing core. Specifically, the first processing core broadcasts and sends an addressing message including the identification information of the second processing core, and after receiving the addressing message, the second processing core feeds back an address response message including the current address information to the first processing core. Correspondingly, the first processing core compares the address information of the second processing core contained in the address response message with the address information of the second processing core stored locally, and if the two are inconsistent, it determines the location of the second processing core. Change.

Step S220: If the routing table update condition is met, determine routing transmission information between the first processing core and the second processing core according to the current physical location of the first processing core and the current physical location of the second processing core.

Wherein, the routing table update condition includes: the routing table corresponding to the first processing core is empty, and the physical location of the first processing core and/or the second processing core changes. In this embodiment, at least the following two ways are implemented to determine the route transmission information between the first processing core and the second processing core:

(1) Path calculation method

This method can directly determine the route transmission information through calculation, which is convenient and quick. Specifically, route transmission information between the first processing core and the second processing core is determined through path calculation. During specific implementation, the first position coordinates corresponding to the current physical position of the first processing core and the second position coordinates corresponding to the current physical position of the second processing core are obtained; according to the preset path setting strategy, the first position coordinates are determined At least one path information corresponding to the first position coordinate and the second position coordinate; calculating the transmission cost of each path information respectively, and determining the route transmission information according to the calculation result.

Wherein, the path setting strategy can be pre-configured, specifically including at least one of the following strategies:

The first type of strategy is an avoidance strategy, which is used to set the transmission nodes corresponding to the processing cores that need to be avoided during the path selection process, so as to bypass such transmission nodes during the path calculation process. Specifically, avoidance strategies include: absolute avoidance strategies and suggested avoidance strategies. Among them, the absolute avoidance strategy usually corresponds to the situation where the transmission node is physically damaged and cannot transmit. At this time, such nodes need to be absolutely avoided; the recommended avoidance strategy usually corresponds to the situation where the transmission node has low transmission efficiency due to excessive load and other reasons. , at this time, you can choose whether to avoid such nodes according to the actual situation.

The second strategy is a selection strategy, which is used to select the transmission node corresponding to the processing core with a higher priority in the path selection process, so as to select this type of transmission node as much as possible in the path calculation process. Specifically, the selection strategy may be set according to factors such as the load of each transmission node, transmission delay, and the like. During specific implementation, selection strategies include: absolute selection strategies and suggested selection strategies. Among them, the absolute selection strategy can be configured by the operator to flexibly define the default routing selection method.

The above two strategies can be used alone or in combination, which is not limited in the present disclosure. During specific implementation, the transmission cost of each path can be determined according to the routing line statistics of the previous period, and then the routing transmission information can be determined.

(2) Exploration methods

This method determines the route transmission information by exploring the data packets, which is beneficial to dynamically adjust the route according to the dynamic change of the physical core location. Specifically, the route transmission information between the first processing core and the second processing core is determined by transmitting a discovery data packet.

Specifically, according to the relative positional relationship between the current physical position of the first processing core and the current physical position of the second processing core, an exploration data packet is generated; the exploration data packet is transmitted to the second processing core by means of exploration; the exploration data packet It is used for recording transmission parameters corresponding to the exploration data packets during the transmission process; and determining routing transmission information between the first processing core and the second processing core according to the transmission parameters of the exploration data packets. Wherein, the transmission parameters include a transmission path and/or a position offset. Wherein, after each transmission node receives the exploration data packet, it transmits the data packet to the transmission node located at the next hop according to the exploration strategy, and updates the transmission path and/or position offset of the exploration data packet according to the transmission path; wherein, the transmission node is a transmission node located between the first processing core and the second processing core, and specifically may be a routing node located between the first processing core and the second processing core. In this embodiment, each processing core corresponds to a routing node, therefore, the transfer node at least includes: a first routing node corresponding to the first processing core, a second routing node corresponding to the second processing core, and a A routing node between the core and the second processing core. In addition, the exploration strategy includes: a random selection strategy or a sequential selection strategy, and the present disclosure does not limit the specific strategy, as long as different routing directions can be tried as much as possible. For example, a genetic algorithm can be used to try: send a large number of exploration data packets at the fork node, so as to find the fastest path or paths. For example, at the initial stage of transmission, you can try any fork point. Whenever you encounter an uncertain routing node, some data packets go in the first direction, and some data packets go in the second direction, so as to find out the fastest one. To N paths as transfer paths, N is a natural number. It can be seen that by sending a large number of trial packets at the fork point, the fastest or TOP N path can be found. This method belongs to slow update, sending exploration data packets until a better path is determined.

Among them, the above-mentioned transmission parameters are used to record the information of each transmission node passed by the data packet during the transmission process. Correspondingly, the information content of the transmission parameters is dynamically updated with the change of the transmission hops of the data packet, and the data can be determined through the transmission parameters. The transmission path of the package. Since the transmission path of each data packet in the exploration transmission mode is unknown, it is necessary to dynamically record the transmission path by means of transmission parameters. During specific implementation, the transmission path in the transmission parameters may be determined in various ways. For example, corresponding digital numbers are set for each transmission direction, and the transmission path is represented by a digital sequence composed of a group of digital numbers. For example, assume that transmission in the first direction (such as the left side) corresponds to the number 0, transmission in the second direction (such as the right) corresponds to the number 1, and transmission in the third direction (such as the top) corresponds to the number number 2. Assume that the transmission to the fourth direction (such as the bottom) corresponds to the number 3. Correspondingly, if the transmission is first to the left and then to the top, it is correspondingly expressed as a digital sequence 02, and so on.

In addition, when determining the routing transmission information between the first processing core and the second processing core according to the transmission parameters of the exploration data packet, it can be realized in at least one of the following two ways:

In the first way, the transmission parameters of each exploration data packet sent by the second processing core are received, at least one transmission path is selected as the target path according to the transmission parameters of each exploration data packet, and the first processing core and the second processing core are determined according to the target path. The second handles the routing and transmission of information between cores. In this way, after the second processing core receives the exploration data packet, it directly sends the transmission parameters of each exploration data packet to the current processing core (that is, the first processing core), and the first processing core transmits the parameters according to the transmission parameters of each exploration data packet. parameters, at least one transmission path is selected as a target path, and the route transmission information between the first processing core and the second processing core is determined according to the target path.

In the second method, the transmission parameters of each exploration data packet sent by the second processing core are received, at least one transmission path is selected as the target path according to the transmission parameters of each exploration data packet, and the first processing core and the second processing core are determined according to the target path. The second handles the routing and transmission of information between cores. In this way, after receiving the exploration data packets, the second processing core obtains the transmission parameters of each exploration data packet, selects at least one transmission path as the target path according to the transmission parameters of each exploration data packet, and directly sends the target path to The first processing core.

It can be seen that the difference between the above two methods lies in that the execution subjects of the screening target paths are different. In the first manner, the target path is screened by the first processing core; in the second manner, the target path is screened by the second processing core.

In addition, when determining the route transmission information between the first processing core and the second processing core by transmitting the exploration data packet, in order to facilitate the determination of the transmission efficiency of the exploration data packet, a certain number of preset path data packets can also be set, The final route transmission information is determined by comparing the transmission result of the exploration data packet with the preset path data packet. Specifically, the first number of exploration data packets are transmitted to the second processing core through exploration; the second number of preset path data packets are transmitted to the second processing core according to the preset transmission path; according to the exploration data packets and the preset The path data packet determines routing transmission information between the first processing core and the second processing core. Wherein, the preset transmission path may be determined according to a path calculation method. For example, assuming that multiple paths are determined through path calculation, at least one path is selected from the multiple paths as the preset transmission path of the preset path data packet.

In addition, in the case of setting the exploration data packet and the preset path data packet at the same time as mentioned above, at least one of the following three methods can be used to determine whether the data packet to be transmitted is an exploration data packet:

(1) Acquiring sequence numbers of data packets to be transmitted, and determining the first number of data packets as exploration data packets according to the sequence numbers of data packets to be transmitted. For example, the ratio between the exploration data packets and the preset path data packets can be set through sequence numbers. For example, the data packets of the first proportion are transmitted according to the preset optimal path, and the data packets of the second proportion are used as exploration data packets to try different paths. In a specific case, 90% of the data packets may be used as preset path data packets, and 10% of the data packets may be used as exploration data packets.

(2) Obtain the information type of the data packet to be transmitted, and determine the data packet whose information type belongs to the preset type as the exploration data packet. For example, judge the delay requirements and reliability requirements of data packets according to the information type, so that important data packets go to the preset, and judge whether to change the path according to the transmission delay of the exploration data packets. Among them, the preset type is usually an information type with important content and high timeliness requirements.

(3) Determine whether the current period information belongs to the preset exploration period, and if so, determine the data packets to be transmitted in the current period as exploration data packets. This method can flexibly configure different time periods as the exploration time period according to the network status of each business period. This method mainly divides the transmission mode of the data packet from the time dimension, so that a certain period of time is used as the exploration period. Specifically, it is possible to pre-monitor the communication traffic of each period of the day, and determine a low-peak business period with a relatively low traffic volume as a preset exploration period.

In addition, the above-mentioned transmission parameters can further record the time stamp when each transmission node forwards the data packet, so as to determine the transmission delay of the data packet, so that when screening the target path, according to the transmission hop count and/or transmission The delay calculates the transmission cost, and then filters the target path according to the transmission cost. For example, the routing table is updated when the transmission cost is lower than the preset cost, so as to add the target path with the lower transmission cost to the routing table. During specific implementation, the transmission cost of each transmission path may be calculated separately, and each transmission path may be sorted according to the order of transmission cost from low to high, and then several transmission paths with the highest ranking may be extracted as target paths. In addition, in the scenario where the exploration data packet is combined with the preset path data packet, the transmission cost of the exploration data packet can also be compared with the transmission cost of the preset path data packet. Only when the transmission cost of the exploration data packet is less than the preset When the transmission cost of the path data packet is determined, the routing table is updated based on the transmission path of the exploration data packet.

In addition, in the exploration data packet, the destination address of the data packet also needs to be included, which can be expressed in various ways. For example, the physical address of the second processing core may be directly used for identification, or the position offset of the second processing core relative to the first processing core may be used for identification, which is not limited in the present disclosure. In addition, in addition to including the destination address information, the search data packet may further include information such as start address information and an information type used to indicate the importance of the data packet. Certainly, the exploration data packet may further set an exploration flag, so that each routing node may determine its transmission mode as the exploration transmission mode according to the exploration flag.

Step S230: Update the routing table corresponding to the first processing core according to the routing transmission information.

Specifically, in the routing table corresponding to the first processing core, query the stored routing information between the first processing core and the second processing core; update the stored routing information according to the routing transmission information; wherein, the updating operation includes At least one of the following: replacing a class update operation, and adding a class update operation.

In summary, through the method in this embodiment, by dynamically determining the physical positions of the first processing core and the second processing core, the routing transmission information can be flexibly set, so that it is convenient to dynamically update the routing table and adjust the routing, and can be flexibly applied For complex application scenarios where the physical location of the processing core can change dynamically. Among them, the combination of exploration transmission and deterministic path can take into account the requirements of path reliability and flexibility.

In addition, another embodiment of the present disclosure also provides a device for updating a routing table based on a many-core system. FIG. 3 shows a schematic structural diagram of the device. As shown in FIG. 3 , the device includes:

The determination module 31 is adapted to determine the routing transmission between the first processing core and the second processing core according to the current physical location of the first processing core and the current physical location of the second processing core when the routing table update condition is satisfied information;

The update module 32 updates the routing table corresponding to the first processing core according to the routing transmission information; wherein, the second processing core is a processing core receiving the first processing core data packet, and the routing table update condition includes at least one of the following: first The routing table corresponding to the processing core is empty, and the physical location of the first processing core and/or the second processing core changes.

Optionally, when the routing table update condition is that the routing table corresponding to the first processing core is empty, the current physical location of the second processing core is determined by broadcasting a location query message to the second processing core ; Receive a location response message sent by the second processing core, and determine the current physical location of the second processing core according to the location response message.

Optionally, the determination module is specifically adapted to:

determining routing transmission information between the first processing core and the second processing core by way of path calculation; and/or

The route transmission information between the first processing core and the second processing core is determined by transmitting a discovery data packet.

Optionally, the determination module is specifically adapted to:

obtaining first location coordinates corresponding to the current physical location of the first processing core, and second location coordinates corresponding to the current physical location of the second processing core;

Determine at least one piece of path information corresponding to the first location coordinate and the second location coordinate according to a preset path setting strategy;

The transmission cost of each path information is calculated respectively, and the route transmission information is determined according to the calculation result.

Optionally, the determining module is specifically adapted to, the determining the route transmission information between the first processing core and the second processing core by transmitting the exploration data packet includes:

generating an exploration data packet according to a relative positional relationship between the current physical position of the first processing core and the current physical position of the second processing core;

Transmitting the exploration data packet to the second processing core in an exploratory manner; the exploration data packet is used to record transmission parameters corresponding to the exploration data packet during the transmission process, and the transmission parameters include a transmission path and/or position offset;

receiving the transmission parameters of each exploration data packet sent by the second processing core, selecting at least one transmission path as a target path according to the transmission parameters of each exploration data packet, and determining the relationship between the first processing core and the Routing transfer information between the second processing cores.

Optionally, the determination module is specifically adapted to:

generating an exploration data packet according to a relative positional relationship between the current physical position of the first processing core and the current physical position of the second processing core;

Transmitting the exploration data packet to the second processing core in an exploratory manner; the exploration data packet is used to record transmission parameters corresponding to the exploration data packet during the transmission process, and the transmission parameters include a transmission path and/or position offset;

Receiving at least one target path obtained by screening by the second processing core according to the transmission parameters of each exploration data packet, and determining routing transmission information between the first processing core and the second processing core according to the target path.

Optionally, after each transmission node receives the exploration data packet, it transmits the data packet to the next-hop transmission node according to the exploration strategy, and updates the transmission path and/or the transmission path of the exploration data packet according to the transmission path. Position offset; wherein, the transfer node is a transfer node located between the first processing core and the second processing core, and the exploration strategy includes: a random selection strategy, or a sequential selection strategy.

Optionally, the determination module is specifically adapted to:

transmitting the first number of exploration data packets to the second processing core in an exploration manner;

transmitting a second number of preset path data packets to the second processing core according to a preset transmission path;

Determine routing transmission information between the first processing core and the second processing core according to the exploration data packet and the preset path data packet.

Optionally, the determination module is specifically adapted to:

Acquiring the sequence numbers of the data packets to be transmitted, and determining the first number of data packets as exploration packets according to the sequence numbers of the data packets to be transmitted; or,

Acquire the information type of the data packet to be transmitted, and determine the data packet whose information type belongs to the preset type as the exploration data packet; or,

It is judged whether the current period information belongs to the preset exploration period, and if so, the data packet to be transmitted in the current period is determined as the exploration data packet.

Optionally, the update module is specifically adapted to:

In the routing table corresponding to the first processing core, query the stored routing information between the first processing core and the second processing core;

An update operation is performed on the stored route information according to the route transmission information; wherein the update operation includes at least one of the following: a replace type update operation, and an add type update operation.

Optionally, the determining module is further adapted to:

According to the historical communication state of the first processing core, the physical position of the first processing core is adjusted, and the update message of the physical position of the first processing core is broadcast to other processing cores.

Optionally, the determining module is further adapted to: receive a physical location update message of the second processing core sent by the second processing core, where the physical location update message is adjusted by the second processing core according to the historical communication state of the second processing core. The message broadcast after the physical location of the second processing core.

In addition, referring to FIG. 4 , an embodiment of the present disclosure provides an electronic device, which includes:

one or more processors 101;

The memory 102 stores one or more programs thereon, and when the one or more programs are executed by one or more processors, the one or more processors implement the routing method in any one of the foregoing embodiments.

One or more I/O interfaces 103 are connected between the processor and the memory, and are configured to realize information exchange between the processor and the memory.

Wherein, the processor 101 is a device with data processing capability, which includes but not limited to a central processing unit (CPU), etc.; the memory 102 is a device with data storage capability, which includes but not limited to a random access memory (RAM, more specifically Such as SDRAM, DDR, etc.), read-only memory (ROM), electrified erasable programmable read-only memory (EEPROM), flash memory (FLASH); I/O interface (read-write interface) 103 is connected between processor 101 and memory 102 , can realize information exchange between the processor 101 and the memory 102, which includes but not limited to a data bus (Bus) and the like.

In some embodiments, the processor 101 , the memory 102 and the I/O interface 103 are connected to each other through the bus 104 , and further connected to other components of the computing device.

In addition, referring to FIG. 5 , an embodiment of the present disclosure provides a computer-readable medium on which a computer program is stored, and when the program is executed by a processor, the routing method provided by any of the foregoing embodiments is implemented.

Those of ordinary skill in the art can understand that all or some of the steps in the methods disclosed above, the functional modules/units in the system, and the device can be implemented as software, firmware, hardware, and an appropriate combination thereof. In a hardware implementation, the division between functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be composed of several physical components. Components cooperate to execute. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit . Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). As known to those of ordinary skill in the art, the term computer storage media includes both volatile and nonvolatile media implemented in any method or technology for storage of information, such as computer readable instructions, data structures, program modules, or other data. permanent, removable and non-removable media. Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cartridges, magnetic tape, magnetic disk storage or other magnetic storage devices, or can Any other medium used to store desired information and which can be accessed by a computer. In addition, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media .

Example embodiments have been disclosed herein, and while specific terms have been employed, they are used and should be construed in a generic descriptive sense only and not for purposes of limitation. In some instances, it will be apparent to those skilled in the art that features, characteristics and/or elements described in connection with a particular embodiment may be used alone, or may be described in combination with other embodiments, unless explicitly stated otherwise. Combinations of features and/or elements. Accordingly, it will be understood by those of ordinary skill in the art that various changes in form and details may be made without departing from the scope of the present disclosure as set forth in the appended claims.

Claims (15)

1. A routing table update method based on many-core system, comprising: When the routing table update condition is met, determine routing transmission information between the first processing core and the second processing core according to the current physical location of the first processing core and the current physical location of the second processing core; updating the routing table corresponding to the first processing core according to the routing transmission information; Wherein, the second processing core is a processing core that receives the first processing core data packet, and the routing table update condition includes at least one of the following: the routing table corresponding to the first processing core is empty, the first processing core and/or the physical location of the second processing core changes. 2. The method according to claim 1, wherein, when the routing table update condition is that the routing table corresponding to the first processing core is empty, the current physical location of the second processing core is in the following manner Sure: broadcast a location query message to the second processing core; Receive a location response message sent by the second processing core, and determine the current physical location of the second processing core according to the location response message. 3. The method according to claim 1, wherein said determining routing transmission information between said first processing core and said second processing core comprises: determining routing transmission information between the first processing core and the second processing core by way of path calculation; and/or The route transmission information between the first processing core and the second processing core is determined by transmitting a discovery data packet. 4. The method according to claim 3, wherein the determining the route transmission information between the first processing core and the second processing core through path calculation comprises: obtaining first location coordinates corresponding to the current physical location of the first processing core, and second location coordinates corresponding to the current physical location of the second processing core; Determine at least one piece of path information corresponding to the first location coordinate and the second location coordinate according to a preset path setting strategy; The transmission cost of each path information is calculated respectively, and the route transmission information is determined according to the calculation result. 5. The method according to claim 3, wherein said determining the routing transmission information between the first processing core and the second processing core by transmitting the exploration data packet comprises: generating an exploration data packet according to a relative positional relationship between the current physical position of the first processing core and the current physical position of the second processing core; Transmitting the exploration data packet to the second processing core in an exploratory manner; the exploration data packet is used to record transmission parameters corresponding to the exploration data packet during the transmission process, and the transmission parameters include a transmission path and/or position offset; receiving the transmission parameters of each exploration data packet sent by the second processing core, selecting at least one transmission path as a target path according to the transmission parameters of each exploration data packet, and determining the relationship between the first processing core and the Routing transfer information between the second processing cores. 6. The method according to claim 3, wherein the determining the routing transmission information between the first processing core and the second processing core by transmitting the exploration data packet comprises: generating an exploration data packet according to a relative positional relationship between the current physical position of the first processing core and the current physical position of the second processing core; Transmitting the exploration data packet to the second processing core in an exploratory manner; the exploration data packet is used to record transmission parameters corresponding to the exploration data packet during the transmission process, and the transmission parameters include a transmission path and/or position offset; Receiving at least one target path obtained by screening by the second processing core according to the transmission parameters of each exploration data packet, and determining routing transmission information between the first processing core and the second processing core according to the target path. 7. The method according to claim 5 or 6, wherein, after receiving the exploration data packet, each transmission node transmits the data packet to the transmission node located at the next hop according to the exploration strategy, and updates the data packet according to the transmission path The transmission parameters of the exploration data packet; wherein, the transmission node is a transmission node located between the first processing core and the second processing core, and the exploration strategy includes: a random selection strategy or a sequential selection strategy. 8. The method according to claim 3, wherein said determining the routing transmission information between the first processing core and the second processing core by transmitting the exploration data packet comprises: transmitting the first number of exploration data packets to the second processing core in an exploration manner; transmitting a second number of preset path data packets to the second processing core according to a preset transmission path; Determine routing transmission information between the first processing core and the second processing core according to the exploration data packet and the preset path data packet. 9. The method according to claim 8, wherein, before determining the route transmission information between the first processing core and the second processing core by transmitting the exploration data packet, further comprising: Acquiring the sequence numbers of the data packets to be transmitted, and determining the first number of data packets as exploration packets according to the sequence numbers of the data packets to be transmitted; or, Acquire the information type of the data packet to be transmitted, and determine the data packet whose information type belongs to the preset type as the exploration data packet; or, It is judged whether the current period information belongs to the preset exploration period, and if so, the data packet to be transmitted in the current period is determined as the exploration data packet. 10. The method according to claim 1, wherein, according to the routing transmission information, updating the routing table corresponding to the first processing core comprises: In the routing table corresponding to the first processing core, query the stored routing information between the first processing core and the second processing core; An update operation is performed on the stored route information according to the route transmission information; wherein the update operation includes at least one of the following: a replace type update operation, and an add type update operation. 11. The method according to any one of claims 1-10, wherein, before the method is executed, further comprising: According to the historical communication state of the first processing core, the physical position of the first processing core is adjusted, and the update message of the physical position of the first processing core is broadcast to other processing cores. 12. The method according to claim 11, wherein the method further comprises: receiving a physical location update message of the second processing core sent by the second processing core, the physical location update message is the second processing core according to the second processing The historical communication state of the core, the message broadcast after adjusting the physical location of the second processing core. 13. A routing table updating device based on a many-core system, comprising: The determining module is adapted to determine the distance between the first processing core and the second processing core according to the current physical location of the first processing core and the current physical location of the second processing core when the routing table update condition is satisfied. route transmission information; An update module, adapted to update the routing table corresponding to the first processing core according to the routing transmission information; Wherein, the second processing core is a processing core that receives the first processing core data packet, and the routing table update condition includes at least one of the following: the routing table corresponding to the first processing core is empty, the first processing core and/or the physical location of the second processing core changes. 14. An electronic device comprising: one or more processors; A storage device, on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement any of the A method for updating a routing table based on a many-core system. 15. A computer-readable medium on which a computer program is stored, and when the program is executed by a processor, at least one of the following methods is implemented: The method for updating a routing table based on a many-core system according to any one of claims 1-12.

Images