WO2020062297A1 - Token management method and apparatus, chip, and mobile platform - Google Patents

Abstract

A token management method: after dividing a command into a plurality of command fragments and sequentially allocating a plurality of tokens, sequentially storing the tokens and generating token storage information, the storage information of each token being used for indicating: whether the token is the last token of the command, whether the token is the last token corresponding to a command ID to which the command belongs, and the storage address of the next token after said token in the command. By means of said storage information, all of the tokens under the command can be acquired on the basis of the order of allocation of the tokens, thereby ensuring that the data under the same command will not be in error. The tokens do not need to be stored in an FIFO form in the order of allocation of the tokens, saving storage resources and avoiding resource wastage.

Description

Token management method, device, chip and mobile platform Technical field

Embodiments of the present invention relate to the technical field of chips, and in particular, to a token management method, device, chip, and mobile platform.

Background technique

System-on-chip (SoC) refers to an integrated chip with embedded system as the core, integrating software and hardware, and pursuing the maximum tolerance of the product system. It can be applied to mobile phones, image processing, TV games and other fields. For software application scenarios that require large-capacity memory, the SoC usually integrates a Double Data Rate (DDR) controller to connect DDR devices to provide system memory space. Generally improving the access efficiency of the DDR controller can improve the processing performance of the entire SoC. In order to improve the access efficiency of the DDR controller, multiple access ports are configured for the DDR controller. These access ports follow certain bus standards, such as the advanced scalable interface of the Advanced Microcontroller Bus Architecture (AMBA). (Advanced Xtensible Interface, AXI), the Advanced High-performance Bus (AHB) bus interface standard. As a SLAVE, the DDR controller has multiple access masters externally. These access masters can come from, for example, an application processor, a digital signal processor, a graphics processing unit, and multimedia.

Among them, if the bus standard between the DDR controller and the access port is AXI of AMBA, the command with the same ID issued by the same access master is consistent with the order of reading data, that is, the read data corresponding to the read command issued first return. However, the DDR device is not an AXI bus standard interface, so a DDR controller is required to convert the read instruction and timing transmitted by the AXI bus into the read instruction and timing of the DDR device. The details are as follows: After the DDR controller receives the read instruction through the AXI bus, it divides each read instruction into multiple instruction fragments, and assigns a token to each instruction fragment to record the order of the read instructions. The DDR controller then re-adjusts the order of these instruction fragments according to the current state of the DDR device, and sends the re-adjusted instruction fragments to the DDR device. The DDR device sequentially returns the read data according to the order of the received instruction fragments, and reads them for each The data is added to the token of the corresponding instruction fragment, wherein the token is generally stored in a first-in-first-out (FIFO) memory, and the read data corresponding to the token generated first is read out first, and the token generated later The corresponding read data is read out later, so the read data of the same read instruction is combined in order to obtain the sorted read data, where the sorted read data is consistent with the order of the corresponding read instruction, and then the sorted The read data is sent to the access master in turn.

In the prior art, the number and depth of FIFO memories are generally determined according to the maximum number of possible IDs N to access the MASTER and the depth of the CAM instruction of the DDR device (usually represented by 2 ^M ). The maximum number of possible master IDs. In order to cover the limit case that the read instructions all come from the same ID, the depth of each FIFO memory is equal to the CAM depth of the above DDR device instruction, that is, the width of the FIFO memory is M. It can be seen that the total storage of the FIFO memory is N times the depth of the CAM, which causes the storage resources of the FIFO to be much larger than the actual token storage resources required, resulting in a waste of resources.

Summary of the Invention

Embodiments of the present invention provide a token management method, device, chip, and mobile platform, which are used to save resources required for storing tokens and save resource overhead.

In a first aspect, an embodiment of the present invention provides a token management method, including:

Divide the received instruction into K instruction fragments, and allocate tokens for each instruction fragment in turn;

Store K tokens in the assigned order and generate storage information for each token;

Reorder the K instruction fragments, and send the reordered K instruction fragments;

The storage information of each token is used to indicate whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the instruction in the instruction. The storage address of the token's next token.

In a second aspect, an embodiment of the present invention provides a token management method, including:

Receiving data read by each instruction fragment in a plurality of instruction fragments, the plurality of instruction fragments including K instruction fragments, and the K instruction fragments are all instruction fragments of a reordered instruction;

Determine the token corresponding to the data read by each instruction fragment;

Determine the storage address indicated by the token read information as the storage address of the first token of the instruction according to the token read information identified by the instruction to which the instruction belongs, and according to the first token The storage address of the first token;

Obtaining the storage information of the other K-1 tokens of the instruction in sequence according to the storage information of the first token;

Obtaining other K-1 tokens according to the storage information of the K-1 tokens in sequence;

According to the order of obtaining the K tokens of the instruction and the tokens corresponding to the data read by each instruction fragment, the data corresponding to the K tokens is reorganized to obtain the data read by the instruction ;

The storage information of each token is used to indicate whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the instruction in the instruction. The storage address of the next token of the token;

The token reading information is used to indicate a storage address of a first token currently stored corresponding to an instruction identifier to which the instruction belongs.

According to a third aspect, an embodiment of the present invention provides a token management apparatus, including: a memory and a processor;

The memory is used to store program code;

The processor is configured to call the code stored in the memory to execute the token management method according to the embodiment of the present invention in the first aspect, or the token according to the embodiment of the present invention in the second aspect Management methods.

According to a fourth aspect, an embodiment of the present invention provides a chip, including: a DDR controller and a DDR device;

The DDR controller is configured to execute the token management method according to the embodiment of the present invention in the first aspect, and the token management method according to the embodiment of the present invention in the second aspect;

The DDR device is configured to receive the reordered multiple instruction fragments sent by the DDR device, and send data read by each instruction fragment to the DDR controller.

In a fifth aspect, a movable platform according to an embodiment of the present invention includes a fuselage, a pan-tilt, and an imaging device. The imaging device is connected to the fuselage through the pan-tilt. Aspects of the chip according to the embodiment of the present invention;

The chip is also connected to the imaging device, and the chip is further configured to process image data collected by the imaging device.

According to a sixth aspect, an embodiment of the present invention provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, where the computer program includes at least one piece of code, and the at least one piece of code can be executed by a computer to control The computer executes the token management method according to the first aspect and / or the second aspect of the embodiments of the present invention.

In a seventh aspect, an embodiment of the present invention provides a computer program for implementing the token management method according to the first aspect and / or the second aspect of the embodiment of the present invention when the computer program is executed by a computer.

According to the token management method, device, chip and mobile platform provided by the embodiments of the present invention, after dividing an instruction into multiple instruction fragments, sequentially allocating multiple tokens, storing the tokens in sequence, and generating token storage Information, and the storage information of each token is used to indicate whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction ID to which the instruction belongs, and the instruction in the instruction. The storage address of the token's next token. Through the storage information, all tokens under the instruction can be obtained according to the order in which the tokens are allocated, thereby ensuring that data under the same instruction will not be wrong. There is no need to store tokens in the form of FIFO in accordance with the order in which the tokens are allocated, saving storage resources and avoiding waste of resources.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic architecture diagram of an unmanned flight system according to an embodiment of the present invention;

2 is a flowchart of a token management method according to an embodiment of the present invention;

3 is a schematic diagram of a token storage structure, an ID read pointer storage structure, and an ID write pointer storage structure according to an embodiment of the present invention;

4 is a schematic diagram of a token storage structure, an ID read pointer storage structure, and an ID write pointer storage structure during initialization according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of instruction division and allocation token provided by an embodiment of the present invention; FIG.

6 is a schematic diagram of four tokens of a storage instruction 0 according to an embodiment of the present invention;

7 is a schematic diagram of three tokens of a storage instruction 1 according to an embodiment of the present invention;

8 is a schematic diagram of five tokens of a storage instruction 2 provided by an embodiment of the present invention;

9 is a schematic diagram of a token reading sequence according to an embodiment of the present invention;

FIG. 10 is a schematic diagram of four tokens of read instruction 0 according to an embodiment of the present invention; FIG.

11 is a schematic diagram of three tokens of a read instruction 1 according to an embodiment of the present invention;

FIG. 12 is a schematic diagram of five tokens of a read instruction 2 according to an embodiment of the present invention; FIG.

13 is a schematic structural diagram of a token management apparatus according to an embodiment of the present invention;

14 is a schematic structural diagram of a token management apparatus according to another embodiment of the present invention;

15 is a schematic structural diagram of a chip according to an embodiment of the present invention;

FIG. 16 is a schematic structural diagram of a movable platform according to an embodiment of the present invention.

detailed description

In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of the embodiments of the present invention, but not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by a person of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

It should be noted that when a component is called "fixed to" another component, it may be directly on another component or a centered component may exist. When a component is considered to be “connected” to another component, it may be directly connected to another component or a centered component may coexist.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

Hereinafter, some embodiments of the present invention will be described in detail with reference to the drawings. In the case of no conflict, the following embodiments and features in the embodiments can be combined with each other.

Embodiments of the present invention provide a token management method, device, chip, and mobile platform. The movable platform may be, for example, a drone, an unmanned ship, an unmanned car, a robot, or the like. The drone may be a rotorcraft, for example, a multi-rotor aircraft propelled by multiple propulsion devices through air, and the embodiment of the present invention is not limited thereto.

FIG. 1 is a schematic architecture diagram of an unmanned flight system according to an embodiment of the present invention. This embodiment is described by taking a rotary wing drone as an example.

The unmanned aerial system 100 may include a drone 110, a display device 130, and a control terminal 140. The UAV 110 may include a power system 150, a flight control system 160, a rack, and a gimbal 120 carried on the rack. The drone 110 may perform wireless communication with the control terminal 140 and the display device 130.

The frame may include a fuselage and a tripod (also called a landing gear). The fuselage may include a center frame and one or more arms connected to the center frame, and one or more arms extend radially from the center frame. The tripod is connected to the fuselage, and is used to support the UAV 110 when landing.

The power system 150 may include one or more electronic governors (referred to as ESCs) 151, one or more propellers 153, and one or more electric motors 152 corresponding to the one or more propellers 153. The electric motors 152 are connected to Between the electronic governor 151 and the propeller 153, the motor 152 and the propeller 153 are arranged on the arm of the drone 110; the electronic governor 151 is used to receive the driving signal generated by the flight control system 160 and provide driving according to the driving signal Current is supplied to the motor 152 to control the rotation speed of the motor 152. The motor 152 is used to drive the propeller to rotate, so as to provide power for the flight of the drone 110, and the power enables the drone 110 to achieve one or more degrees of freedom. In some embodiments, the drone 110 may rotate about one or more rotation axes. For example, the rotation axis may include a roll axis (Roll), a yaw axis (Yaw), and a pitch axis (Pitch). It should be understood that the motor 152 may be a DC motor or an AC motor. In addition, the motor 152 may be a brushless motor or a brushed motor.

The flight control system 160 may include a flight controller 161 and a sensing system 162. The sensing system 162 is used to measure the attitude information of the drone, that is, the position information and status information of the drone 110 in space, such as three-dimensional position, three-dimensional angle, three-dimensional velocity, three-dimensional acceleration, and three-dimensional angular velocity. The sensing system 162 may include, for example, at least one of a gyroscope, an ultrasonic sensor, an electronic compass, an Inertial Measurement Unit (IMU), a vision sensor, a global navigation satellite system, and a barometer. For example, the global navigation satellite system may be a Global Positioning System (Global Positioning System, GPS). The flight controller 161 is used to control the flight of the drone 110. For example, the flight controller 161 may control the flight of the drone 110 according to the attitude information measured by the sensing system 162. It should be understood that the flight controller 161 may control the drone 110 according to a pre-programmed program instruction, and may also control the drone 110 by responding to one or more control instructions from the control terminal 140.

The gimbal 120 may include a motor 122. The pan / tilt is used to carry a photographing device 123 or a microphone (not shown in the figure). The flight controller 161 may control the movement of the gimbal 120 through the motor 122. Optionally, as another embodiment, the PTZ 120 may further include a controller for controlling the movement of the PTZ 120 by controlling the motor 122. It should be understood that the gimbal 120 may be independent of the drone 110 or may be a part of the drone 110. It should be understood that the motor 122 may be a DC motor or an AC motor. In addition, the motor 122 may be a brushless motor or a brushed motor. It should also be understood that the gimbal can be located on the top of the drone or on the bottom of the drone.

The photographing device 123 may be, for example, a device for capturing an image, such as a camera or a video camera. The photographing device 123 may communicate with the flight controller and perform shooting under the control of the flight controller. The photographing device 123 of this embodiment includes at least a photosensitive element. The photosensitive element is, for example, a complementary metal oxide semiconductor (CMOS) sensor or a charge-coupled device (CCD) sensor. It can be understood that the shooting device 123 can also be directly fixed on the drone 110, so that the PTZ 120 can be omitted.

The display device 130 is located on the ground side of the unmanned flight system 100, can communicate with the drone 110 wirelessly, and can be used to display attitude information of the drone 110. In addition, an image captured by the imaging device may be displayed on the display device 130. It should be understood that the display device 130 may be an independent device, or may be integrated in the control terminal 140.

The control terminal 140 is located on the ground side of the unmanned flight system 100 and can communicate with the unmanned aerial vehicle 110 in a wireless manner for remotely controlling the unmanned aerial vehicle 110.

In addition, the drone 110 may further include a speaker (not shown) for playing audio files. The speaker may be directly fixed on the drone 110 or may be mounted on the gimbal 120.

It should be understood that the above-mentioned naming of each component of the unmanned flight system is for identification purposes only, and should not be construed as limiting the embodiments of the present invention.

The above-mentioned flight controller 161 may include a chip, and the DDR controller in the chip may be used to implement a solution described below. The data mentioned in the following embodiments are, for example, image data obtained by the shooting device 123. .

FIG. 2 is a flowchart of a token management method according to an embodiment of the present invention. As shown in FIG. 2, the method of this embodiment can be applied to a chip (such as a system-level chip, and this embodiment is not limited to this). For example, In the chip's DDR controller, the method in this embodiment may include:

S201. Divide the received instruction into K instruction fragments, and assign tokens to each instruction fragment in turn.

S202. Store K tokens in the assigned order, and generate storage information of each token.

S203. Reorder the K instruction fragments, and send the reordered K instruction fragments.

In this embodiment, the DDR controller may receive at least one instruction through, for example, the AXI bus. For each instruction, the received instruction is divided into K instruction fragments, and the number of instruction fragments into which each instruction is divided may be different. . In this embodiment, tokens are assigned to each instruction segment in sequence. For example, an instruction is divided into instruction segment 1, instruction segment 2, and instruction segment 3 in sequence. Then, instruction segment 1 is assigned a token, instruction segment 2 is assigned a token, Instruction fragment 3 allocates a token.

In this embodiment, the tokens are stored according to the order in which the tokens are allocated, that is, the tokens that are allocated first are stored first, and the tokens that are allocated later are stored later, so as to ensure that the stored token order is the same as the order of the instruction fragments. And after each token is stored, the storage information of the token is generated. The storage information of each token is used to indicate whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the instruction in the instruction. The storage address of the token's next token. It should be noted that multiple different instructions may belong to the same instruction identifier.

In some embodiments, if the instruction is the first instruction currently stored corresponding to the instruction identifier, the token reading information of the instruction identifier is generated. The token reading information is used to indicate a storage address of a first token currently stored corresponding to an instruction identifier to which the instruction belongs. For information about reading the token, refer to the related description in S206.

In this embodiment, after storing all the tokens of the instruction, the DDR controller reorders these instruction fragments according to the current state of the DDR device. For example, the reordered instruction fragments are: instruction fragment 2, instruction fragment 3, instruction fragment 1, then send the reordered instruction fragment to the DDR device.

Optionally, the relative order of the reordered instruction fragments is also saved in this embodiment, for example, the relative order of the instruction fragments indicating reordering may be stored in the CAM. In one embodiment, the order of the reordered instruction fragments may be stored. Tokens are stored in the CAM in order according to the first-in, first-out rules.

S204. Receive data read by each of the multiple instruction fragments, where the multiple instruction fragments include K instruction fragments, and the K instruction fragments are all instruction fragments of a reordered instruction.

S205. Determine a token corresponding to the data read by each instruction segment.

In this embodiment, the DDR controller receives data sent by the DDR device according to the order of the received instruction fragments. The data may be data read by multiple instruction fragments, where the multiple instruction fragments may include all instruction fragments of at least one instruction. This embodiment takes one instruction as an example, and the instruction includes K instruction fragments, and The K instruction fragments are all instruction fragments after reordering the instruction, and other instructions are processed similarly. Then the token corresponding to the data read by each instruction fragment can be determined. Optionally, the relative order of the reordered instruction fragments is also stored in this embodiment. Therefore, according to the relative order of the received data, the relative order of the instruction fragments corresponding to each data can be determined, thereby obtaining the token corresponding to each data. . For example, the tokens of the reordered instruction fragments may be sequentially stored in the CAM according to the first-in, first-out rule. Therefore, in this embodiment, the token corresponding to the data may be obtained from the CAM according to the order of receiving data.

S206. Determine the storage address indicated by the token reading information as the storage address of the first token of the instruction according to the token reading information identified by the instruction to which the instruction belongs, and according to the first The storage address of the token to obtain the first token.

In this embodiment, one of the instructions is taken as an example. According to the token reading information identified by the instruction to which the instruction belongs, the storage address indicated by the token reading information is obtained, and the storage address is determined as the first of the instruction. The storage address of the token, and then obtain the token stored in the storage address according to the storage address, and the token is the first token of the instruction.

S207. According to the storage information of the first token, sequentially obtain the storage information of the other K-1 tokens of the instruction; and sequentially obtain the other K-1 according to the storage information of the K-1 tokens. Tokens.

In this embodiment, since the storage information of the token is generated when each token is stored, this embodiment can obtain the storage information of the first token. Since the storage information of each token is used for Indication: The storage address of the next token for this token. Therefore, according to the storage information of the first token, the storage information of the second token can be obtained, and according to the storage information of the second token, the second token can be obtained; Store the information, obtain the storage information of the third token, and obtain the third token based on the storage information of the third token, and so on, and will not repeat them here. It should be noted that, in this embodiment, the storage information of each token is also used to indicate whether the token is the last token of the instruction. Therefore, when the obtained storage information of one token indicates that the token is not the last token of the instruction, the storage information of the next token of the token is continuously obtained. When the obtained storage information of one of the tokens indicates that the token is the last token of the instruction, it means that the storage information of all tokens under the instruction has been obtained, and the token is the Kth token. .

In addition, the storage information of each token is also used to indicate whether the token is the last token identified by the instruction to which the instruction belongs. If yes, it means that all the tokens under the instruction ID have been acquired. If not, it means that there are still tokens for other instructions under the instruction ID that have not been acquired.

S208: Reorganize the data corresponding to the K tokens according to the sequence of acquiring the K tokens of the instruction and the data corresponding to the data read by each instruction fragment to obtain the data read by the instruction. The data.

In this embodiment, the division order of the instruction fragment of the instruction determines the reorganization order of the instruction data. After acquiring the K tokens of the instruction, according to the order of acquiring the K tokens (that is, the relative allocation order of the K tokens of the instruction) and the tokens corresponding to each data, the data corresponding to the tokens Reorganize according to the order of obtaining K tokens to obtain reorganized data, and the reorganized data is the correct data read by the instruction.

In this embodiment, after the instruction is divided into multiple instruction fragments, multiple tokens are allocated in sequence, and the storage information of the token is also generated in turn. The storage information of each token is used to indicate that: Whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the storage address of the next token of the token in the instruction. Through the storage information, all tokens under the instruction can be obtained according to the order in which the tokens are allocated, thereby ensuring that data under the same instruction will not be wrong. There is no need to store tokens in the form of FIFO in accordance with the order in which the tokens are allocated, saving storage resources and avoiding waste of resources.

In some embodiments, the storage information of each token is storage information of a first linked list structure; the first linked list structure includes: a first storage instruction, a second storage instruction, and a third storage instruction; A storage instruction is used to indicate whether the token is the last token of the instruction, and the second storage instruction is used to indicate whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the third storage instruction is used Used to indicate the storage address of the next token.

Optionally, the first linked list structure further includes: a fourth storage indication, where the fourth storage indication is used to indicate whether a corresponding storage resource has stored a token.

In some embodiments, a possible implementation manner of the above S202 may include S2021-S2023 as shown below.

S2021. For any two adjacent tokens of the K tokens, store the i-1th token of the K tokens in an idle first storage resource, and generate the i-th token Storage information of 1 token, a fourth storage indication in the storage information of the i-1th token is used to indicate that the corresponding storage resource has stored the token, and a first storage indication is used to indicate the i-th 1 token is not the last token of the instruction, the second storage instruction is used to indicate that the i-1th token is the last token corresponding to the instruction identifier of the instruction, and the third storage instruction The storage address used to indicate the next token is a preset storage address; where i is an integer greater than or equal to 2 and less than or equal to K.

S2022. The i-th token is stored in an idle second storage resource, and storage information of the i-th token is generated, and a fourth storage indication in the storage information of the i-th token is used to indicate The corresponding storage resource has stored a token, the second storage instruction is used to indicate that the i-th token is the last token corresponding to the instruction ID to which the instruction belongs, and the third storage instruction is used for the next token. The storage address is a preset storage address. When i is equal to K, the first storage instruction is used to indicate that the i-th token is the last token of the instruction. When i is not equal to K, the first storage instruction is used. Indicating that the i-th token is not the last token of the instruction;

S2023: Change the second storage instruction in the storage information of the i-1th token to indicate that the i-1th token is not the last token corresponding to the instruction ID of the instruction, and change the third storage instruction to Indicates that the storage address of the next token is the address of the second storage resource.

Optionally, the idle first storage resource and the idle second storage resource are idle adjacent storage resources.

Accordingly, after acquiring each token, the third storage indication in the storage information of the token is changed to a preset storage address.

Optionally, after obtaining the token, if the first storage in the storage information of the token indicates that the token is the last token of the instruction, then change the first storage instruction to indicate the The token is not the last token of the instruction.

Optionally, after acquiring the token, if the fourth storage indication in the storage information of the token is changed from indicating that the token is stored to indicating that the token is not stored.

Optionally, after the token is obtained, if the second storage in the storage information of the token indicates that the token is the last token corresponding to the instruction identifier, the second storage instruction is changed to Indicates that the token is not the last token corresponding to the instruction ID.

Therefore, this embodiment uses a linked list to save the storage information of the token, which can significantly reduce the token storage resources required in the prior art.

In some embodiments, the token reading information is token reading information of a second linked list structure, and the second linked list structure includes a first reading instruction, and the first reading instruction is used to indicate the reading The storage address of the first token corresponding to the instruction ID to which the instruction belongs.

Optionally, the second linked list structure further includes: a second read instruction, where the second read instruction is used to indicate whether a token corresponding to an instruction identifier to which the instruction belongs is stored.

Optionally, a possible implementation manner of generating the token reading information of the instruction is: after storing the first token of the instruction, generating the token reading information identified by the instruction; The second read instruction in the token read information is used to indicate that the instruction has been stored with the token, and the first read instruction is used to indicate the storage address of the first token. If the instruction is the first currently stored instruction corresponding to the corresponding instruction identifier, after storing the first token of the instruction, the token reading information of the instruction identifier is also generated, and after other tokens of the instruction are stored, Keep the token read information identified by this instruction unchanged.

Accordingly, when reading K tokens, if the second storage indication in the storage information of one of the K tokens indicates that the token is the last token corresponding to the instruction ID, The first read instruction is changed to indicate a preset storage address; if the second storage in the storage information of each of the K tokens indicates that the token is not the last token corresponding to the instruction ID, then Changing the first read instruction to indicate a first storage address; the first storage address is: a third storage instruction that belongs to the same storage information as the first storage instruction indicating that the token is the last token of the instruction, The indicated storage address.

Optionally, if the second storage indication in the storage information of one of the K tokens indicates that the token is the last token corresponding to the instruction identifier, changing the second reading indication to A token indicating that the instruction was not stored.

Therefore, in this embodiment, the form of a linked list is used to save the token reading information identified by the instruction, which can significantly reduce the token storage resources required in the prior art.

In some embodiments, after the i-1th token of the K tokens is stored in an idle first storage resource, token writing information identified by an instruction to which the instruction belongs is further generated, where The token write information is used to indicate that the storage address corresponding to the recently stored token corresponding to the instruction identifier is the address of the first storage resource. Accordingly, an implementation manner of storing the i-th token in the idle second storage resource is: determining the idle second storage resource according to the token write information; and storing the i-th token Tokens are stored in a spare second storage resource.

Furthermore, after the i-th token is stored in the idle second storage resource, the token write information is also updated, and the updated token write information is used to indicate that the instruction identifier corresponds to a recently stored token The storage address of is the address of the second storage resource.

Optionally, the token writing information is token writing information of a third linked list structure, and the third linked list structure includes a writing instruction, where the writing instruction is used to indicate that the instruction identifier corresponds to a recently stored The storage address of the token.

Accordingly, when reading K tokens, if the second storage indication in the storage information of one of the K tokens indicates that the token is the last token identified by the instruction, the The token write information of the instruction identifier is changed to indicate that the storage address of the recently stored token corresponding to the instruction identifier is a preset storage address.

Therefore, in this embodiment, a form of a linked list is used to save the token write information identified by the instruction, which can significantly reduce the token storage resources required in the prior art.

It should be noted that the first linked list structure, the second linked list structure, and the third linked list structure may be stored in the same linked list, or may be stored in different linked lists.

The following describes the token storage instructions and the process of obtaining the tokens of the instructions, respectively. The following uses the storage information of the tokens of the first linked list structure as the token storage structure, and the token reading information of the instruction identifier of the second linked list structure as the ID read pointer storage structure and the token identification of the instruction identifier of the third linked list structure. The information is an ID write pointer storage structure as an example for illustration. The ID described below indicates the instruction ID to which the instruction belongs.

3 is a schematic diagram of a token storage structure, an ID read pointer storage structure, and an ID write pointer storage structure according to an embodiment of the present invention. As shown in FIG. 3, in the token storage structure, the token storage structure includes: flag, ind , Last, token, next_ptr, where flag is used to indicate whether the token in the same row has a stored token, for example: flag equal to 1 means stored, and flag equal to 0 means not stored. ind indicates whether the token stored in the same row is the last token of an ID, for example: ind equals 0 means no, and ind equals 1 means yes. last indicates whether the token stored in the token in the same line is the last token of an instruction. token is used to store tokens. next_ptr indicates a pointer to the storage location of the next token of this token relative to the same ID.

In the ID read pointer storage structure, ID is used for addressing, that is, ID = 0 corresponds to 0 address, ID = 1 corresponds to 1 address, ID = 2 corresponds to 2 address, ID = 3 corresponds to 3 address, and so on. The ID read pointer storage structure includes: flag, rd_ptr. flag indicates whether a valid token exists before this ID, flag equals 1 to indicate existence, and flag equals 0 to indicate no existence. rd_ptr is used to point to the storage address of the token to be read next time with this ID.

In the ID write pointer storage structure, ID is used for addressing, that is, ID = 0 corresponds to 0 address, ID = 1 corresponds to 1 address, ID = 2 corresponds to 2 address, ID = 3 corresponds to 3 address, and so on. The ID write pointer storage structure includes: wr_ptr. wr_ptr is used to point to the storage address of the latest token of this ID.

It should be noted that, although different linked lists are used to illustrate the token storage structure, the ID read pointer storage structure, and the ID write pointer storage structure in FIG. 3, this embodiment is not limited thereto. In some embodiments, at least one of the token storage structure, the ID read pointer storage structure, and the ID write pointer storage structure may be stored in the same linked list.

Among them, the possible number of IDs can be determined in advance according to the system architecture and scheme, but the value of the ID cannot be controlled, so when the number of IDs is N, but its value is not fixed from 0 to N- 1, so you can map N different ID values to 0 ~ N-1 numbers, and use this number as the internal ID. When recovering the RID signal of the AXI read data channel, you need to map the ID number to the corresponding ID. value. The ID values mentioned below refer to the numbering information in 0 to N-1.

FIG. 4 is a schematic diagram of a token storage structure, an ID read pointer storage structure, and an ID write pointer storage structure during initialization according to an embodiment of the present invention. As shown in FIG. 4, the foregoing parameters are all 0.

The following describes the storage process of the token based on the initialization as shown in FIG. 4.

S301. Receive five read instructions, which are instructions 0-4, and divide each instruction. Instruction 0 is divided into 4 instruction fragments, corresponding to the generation of 4 tokens, which are 60-63. Instruction 1 is divided into 3 instruction fragments, corresponding to the generation of 3 tokens, 57-59 respectively. Instruction 2 is divided into 5 instruction fragments, corresponding to the generation of 5 tokens, respectively 52-56. Instruction 3 is divided into three instruction fragments, which correspondingly generate 3 tokens, which are 49-51. Instruction 4 is divided into four instruction fragments, which correspondingly generate 4 tokens, which are 45-48, as shown in FIG. 5 for example.

S302. The four tokens of instruction 0 are stored, and the ID of instruction 0 is, for example, 0. As shown in FIG. 6, FIG. 6 is a schematic diagram of four tokens of a storage instruction 0 according to an embodiment of the present invention. Among them, instruction 0 is divided into 4 DDR device instructions, and the allocated tokens are 63, 62, 61, 60, respectively.

S3021, Store token 63: According to the flag in the token storage structure, find the flag equal to 0, write the token 63 into the token in the token storage structure of this row, and set the flag of this row to 1, due to the ID If only 0 is written temporarily, the token 63 can be regarded as the last token of the ID stored, and the int is set to 1. And the token 63 is not the last token of the instruction, so last is set to zero. Then according to flag = 0 in the read pointer storage structure pointed to by ID = 0, the flag is set to 1, and the rd_ptr of this line is updated to the pointer position stored by the token 63, which is 0. Since the position pointer of the token 63 is 0, the wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 0; the schematic result after execution is shown in step A in FIG. 6.

S3022. Store the token 62: find the next position pointer whose position pointer of the token 63 is 0 (that is, the position pointer is 1), and the flag corresponding to this position pointer is 0, so the token storage of the second line can be found Structure, then write the token 62 into the token storage structure of the second line, and set the flag in the token storage structure from 0 to 1, and set the ind in the second line to 1. And the token 62 is not the last token of the instruction, so last is set to zero. According to the write pointer storage structure pointed to by ID = 0, wr_ptr is 0, indicating that the token corresponding to ID = 0 is being stored, and the token in the first line is also a token with ID = 0, and the command in the second line is The card is the latest token with ID = 0, so set the ind in the first line to zero. In addition, the token in the second line is the next token of the token in the first line, and the position pointer of the second line is 1, so the next_ptr in the first line is updated to 1. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 0, the read pointer storage structure is not updated. Because the token with ID = 0 is stored and the position pointer of the token is 1, the wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 1. After execution, it is performed as step B.

S3023. Store the token 61: find the next position pointer whose position pointer of the token 62 is 1 (that is, the position pointer is 2), and the flag corresponding to this position pointer is 0, so the token storage of the third line can be found Structure, and then write the token 61 into the token storage structure of the third line, and set the flag in the token storage structure from 0 to 1, and set the ind in the third line to 1. And the token 61 is not the last token of the instruction, so last is set to zero. According to the write pointer storage structure pointed to by ID = 0, wr_ptr is 0, indicating that the token corresponding to ID = 0 is being stored, and the token in the second line is also a token with ID = 0, and the command in the third line is The card is the latest token with ID = 0, so set the ind in the second line to zero. In addition, the token in the third line is the next token of the token in the second line, and the pointer in the third line is 2, so the next_ptr in the second line is updated to 2. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 0, the read pointer storage structure is not updated. Because the token with ID = 0 is stored, and the position pointer of the token is 2, the wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 2. After execution, it is performed as step C.

S3024. Store the token 60: find the next position pointer whose position pointer of the token 61 is 1 (that is, the position pointer is 3), and the flag corresponding to this position pointer is 0, so the token storage of the fourth line can be found Structure, and then write the token 60 into the token storage structure of the fourth line, and set the flag in the token storage structure from 0 to 1, and set the ind in the fourth line to 1. And the token 60 is the last token of the instruction, so set last to 1. According to the write pointer storage structure pointed to by ID = 0, wr_ptr is 0, indicating that the token corresponding to ID = 0 is being stored, and the token in the third line is also a token with ID = 0, and the command in the fourth line is The card is the latest token with ID = 0, so set ind in the third line to 0. In addition, the token in the fourth line is the next token of the token in the third line, and the pointer in the fourth line is 3, so the next_ptr in the third line is updated to 3. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 0, the read pointer storage structure is not updated. Because the token with ID = 0 is stored and the position pointer of the token is 3, the wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 3, and then executed as step D.

S303. The three tokens of instruction 1 are stored, and the ID of instruction 1 is, for example, 1. As shown in FIG. 7, FIG. 7 is a schematic diagram of three tokens of a storage instruction 1 according to an embodiment of the present invention. Among them, instruction 1 is divided into three instruction fragments, and the allocated tokens are 59, 58, 57 respectively.

S3031, Store token 59: According to the flag in the token storage structure, find the free position pointer as 4, write the token 59 into the token in the token storage structure of this row, and set the flag of this row to 1. Since only one token is temporarily written with ID 1, the token 59 can be regarded as the last token of the ID stored, and the int is set to 1. And the token 59 is not the last token of the instruction, so last is set to 0. Then according to flag = 0 in the read pointer storage structure pointed to by ID = 1, set the flag to 1, and update the rd_ptr of this line to the pointer position stored by token 59, which is 4. Update wr_ptr in the write pointer storage structure pointed to by ID = 1 to 4; the schematic result after execution is shown in step A in FIG. 7.

S3032. Store token 58: find the next position pointer whose position pointer of token 59 is 4 (that is, the position pointer is 5), and the flag corresponding to this position pointer is 0, so the token storage of the sixth row can be found Structure, and then write the token 58 into the token storage structure of the sixth line, and set the flag in the token storage structure from 0 to 1, and set the ind in the sixth line to 1. And the token 58 is not the last token of the instruction, so last is set to zero. According to the write pointer storage structure pointed to by ID = 1, wr_ptr is 4, indicating that the token was written in the position where the position pointer was 4 last time, so the ind in the fifth row is set to 0. In addition, the token in the sixth line is the next token of the token in the fifth line, and the position pointer of the sixth line is 5, so the next_ptr in the fifth line is updated to 5. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 1, the read pointer storage structure is not updated. Since the token with ID = 1 is stored and the position pointer of the token is 5, the wr_ptr in the write pointer storage structure pointed to by ID = 1 is updated to 5, and after execution, it is performed as step B.

S3033. Store the token 57: find the next position pointer whose position pointer of the token 58 is 5 (that is, the position pointer is 6), and the flag corresponding to this position pointer is 0, so the token storage of the seventh line can be found Structure, and then write the token 57 into the token storage structure of the seventh line, and set the flag in the token storage structure from 0 to 1, and set the ind in the seventh line to 1. And this token 57 is the last token of the instruction, so set last to 1. According to the write pointer storage structure pointed to by ID = 1, wr_ptr is 5, and the ind in the sixth row is set to 0. In addition, the token in the seventh line is the next token of the token in the sixth line, and the pointer in the seventh line is 6, so the next_ptr in the sixth line is updated to 6. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 1, the read pointer storage structure is not updated. Because the token with ID = 1 is stored, and the position pointer of the token is 6, the wr_ptr in the write pointer storage structure pointed to by ID = 1 is updated to 6, and after execution, it is performed as step C.

S304. The five tokens of instruction 2 are stored, and the ID of instruction 2 is, for example, 0. As shown in FIG. 8, FIG. 8 is a schematic diagram of five tokens of a storage instruction 2 according to an embodiment of the present invention. Among them, instruction 2 is divided into 5 instruction fragments, and the allocated tokens are 56, 55, 54, 53, 52, respectively.

S3041, Store token 56: According to the flag in the token storage structure, find an idle position pointer as 7, write the token 56 to the token in the token storage structure of this row, and set the flag of this row to 1 Since only one token is temporarily written with ID 1, the token 56 can be considered as the last token of the ID stored, and the int is set to 1. And the token 56 is not the last token of the instruction, so last is set to zero. According to the write pointer storage structure pointed to by ID = 0, wr_ptr is 3, indicating that the ID token was written in the position pointer of 3 last time, so the ind in the fourth line is set to 0. In addition, the token in the eighth line is the next token of the token in the fourth line, and the position pointer of the eighth line is 7, so the next_ptr in the fourth line is updated to 7. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 0, the read pointer storage structure is not updated. Because the token with ID = 0 is stored and the position pointer of the token is 7, the wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 7, and after execution, it is performed as step A.

S3042. Store token 55: find the next position pointer whose position pointer of token 56 is 7 (that is, the position pointer is 8), and the flag corresponding to this position pointer is 0, so the token storage of the ninth row can be found Structure, and then write the token 55 into the token storage structure of the ninth line, and set the flag in the token storage structure from 0 to 1, and set the ind in the ninth line to 1. And the token 55 is not the last token of the instruction, so set last to 0. According to the write pointer storage structure pointed to by ID = 0, wr_ptr is 7, and ind in the eighth row is set to 0. In addition, the token in the ninth line is the next token of the token in the eighth line, and the position pointer of the ninth line is 8, so the next_ptr in the eighth line is updated to 8. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 0, the read pointer storage structure is not updated. Because the token with ID = 0 is stored, and the position pointer of the token is 8, the wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 8. After the execution, step B is performed.

S3043. Store the token 54: find the next position pointer whose position pointer of the token 55 is 8 (that is, the position pointer is 9), and the flag corresponding to this position pointer is 0, so the token storage of the tenth row can be found Structure, and then write the token 54 into the token storage structure of the tenth line, and set the flag in the token storage structure from 0 to 1, and set the ind in the tenth line to 1. And the token 54 is not the last token of the instruction, so last is set to zero. According to the write pointer storage structure pointed to by ID = 0, wr_ptr is 8, and ind in the ninth row is set to 0. In addition, the token in the tenth line is the next token of the token in the ninth line, and the position pointer of the tenth line is 9, so the next_ptr in the ninth line is updated to 9. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 0, the read pointer storage structure is not updated. Because the token with ID = 0 is stored, and the position pointer of the token is 9, the wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 9, and after execution, it is performed as step C.

S3044. Store the token 53: find the next position pointer whose position pointer of the token 54 is 9 (that is, the position pointer is 10), and the flag corresponding to this position pointer is 0, so the token of the eleventh line can be found Storage structure, and then write token 53 into the token storage structure of the eleventh line, and set the flag in the token storage structure from 0 to 1, and set the ind in the eleventh line to 1. And the token 53 is not the last token of the instruction, so last is set to zero. According to the write pointer storage structure pointed to by ID = 0, wr_ptr is 9, and ind in the tenth row is set to 0. In addition, the token in the eleventh line is the next token of the token in the tenth line, and the position pointer of the eleventh line is 10, so the next_ptr in the tenth line is updated to 10. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 0, the read pointer storage structure is not updated. Because the token with ID = 0 is stored, and the position pointer of the token is 10, wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 10, and the execution is performed as step D.

S3045. Store the token 52: find the next position pointer whose position pointer of the token 53 is 10 (that is, the position pointer is 11), and the flag corresponding to this position pointer is 0, so the token of the twelfth line can be found Storage structure, and then write the token 52 into the token storage structure of the twelfth line, and set the flag in the token storage structure from 0 to 1, and set the ind in the twelfth line to 1. And the token 52 is the last token of the instruction, so set last to 1. According to the write pointer storage structure pointed to by ID = 0, wr_ptr is 10, and ind in the eleventh row is set to 0. In addition, the token in the twelfth line is the next token of the token in the eleventh line, and the position pointer of the twelfth line is 11, so the next_ptr in the eleventh line is updated to 11. In addition, if flag = 1 in the read pointer storage structure pointed to by ID = 0, the read pointer storage structure is not updated. Because the token with ID = 0 is stored, and the position pointer of the token is 11, the wr_ptr in the write pointer storage structure pointed to by ID = 0 is updated to 11. After execution, it is performed as step D.

The token storage of other instructions is similarly processed, and is not repeated here.

After the DDR controller saves the token, the instruction is optimized and sent to the DDR device. When the DDR device returns the read data in the order of instruction execution after the DDR controller is optimized, the token order is also adjusted in the DDR controller at this time. The DDR controller recovers the corresponding token information for each returned data.

Write the returned read data into Read Reorder Buff first, and save the write address information with token information for subsequent read data, and make mark information for the returned token, for example, as shown in Figure 9.

The following describes the process of the DDR controller after receiving the above instructions 0-2, dividing it into instruction fragments and sending it to the DDR device, and then receiving the data returned by the DDR device.

S401. Receive the return data corresponding to the instruction 0, as shown in FIG.

Since the ID of the instruction 0 is 0, the read pointer storage structure is queried to determine that the initial read pointer is 0 (ie, rd_ptr in the first row of the read pointer storage structure). Since the initial read pointer of ID = 0 is 0, and the pointer to the position in the token storage structure = 0, that is, the token storage structure in the first line, the corresponding token can be obtained as 63, and the DDR controller judges that the token 63 is related The corresponding data of the token (token 63-60, that is, the token belonging to the same instruction) has been returned, and then starts to read Read Reorder Buff. According to the token storage structure and the read pointer information of ID = 0, it is 0, and the fourth Last in the line is 1, you can infer that the position pointer of the token storage structure needs to be read is 0, 1, 2, 3, so that the token 63, 62, 61, 60 is read, and then the position pointer in the token storage structure is read The flags corresponding to 0, 1, 2, and 3 are set to 0, that is, the flags in the first, second, third, and fourth rows are set to 0, the token storage structure is released, and then the tokens 63, 62, 61, and 60 are queried. Get the read address of the ReadReorder buff, so you can get the correct data for instruction 0. Since the int in the token storage structure corresponding to tokens 63, 62, 61, 60 is not 1, indicating that the token with ID = 0 has not been released, the flag corresponding to ID = 0 in the ID read pointer storage structure cannot be cleared. . After reading the data corresponding to the token 63, 62, 61, 60, the rd_ptr corresponding to ID = 0 in the ID read pointer storage structure needs to be updated to 7 for the next reading of the ID in the token storage structure. The initial pointer position of the operation.

S402. Receive the return data corresponding to the instruction 1, as shown in FIG. 11.

Since the ID of the instruction 1 is 1, the read pointer storage structure is queried to determine that the initial read pointer is 4 (that is, rd_ptr in the second row of the read pointer storage structure). Since the initial read pointer of ID = 1 is 4, and the pointer to the position in the token storage structure = 4, that is, the token storage structure in the fifth line, the corresponding token can be obtained, and the DDR controller judges that the token 59 is related. The data corresponding to the token (token 59-57, that is, the token belonging to the same instruction) has been returned, and then the read Read Reorder Buff is started. According to the token storage structure and the read pointer information of ID = 1, it is 4, and the seventh line Last in is 1 and it can be inferred that the position pointer of the token storage structure needs to be read is 4, 5, 6 so as to read the token 59, 58, 57 and then the position pointer 4, 5, 6 in the token storage structure The corresponding flag is set to 0, that is, the flags in the fifth, sixth, and seventh rows are set to 0, the token storage structure is released, and then the token 59, 58, 57 is queried to obtain the read address of the Read, Reorder, and Buff. The data of the correct instruction 1. Since the int corresponding to token 57 in the token storage structure corresponding to tokens 59, 58, 57 is 1, indicating that all the tokens with ID = 1 have been released, the flag corresponding to ID = 1 in the ID read pointer storage structure needs to be cleared. Zero and change rd_ptr to 0. In addition, wr_ptr with ID = 1 in the ID write pointer storage structure should also be cleared.

S403. Receive the return data corresponding to the instruction 2 as shown in FIG.

Since the ID of instruction 2 is 0, the read pointer storage structure is queried to determine that the initial read pointer is 7 (that is, rd_ptr in the eighth row in the read pointer storage structure). Since the initial read pointer of ID = 0 is 7, pointing to the position pointer in the token storage structure = 7, that is, the token storage structure in the eighth line, the corresponding token can be obtained, and the DDR controller judges that the token 56 is related The corresponding data of the token (token 56-52, that is, the token belonging to the same instruction) has been returned, and then read Read Reorder Buff, according to the token storage structure and the read pointer information of ID = 0 is 7, and the tenth Last in the two lines is 1, you can infer that the position pointer that needs to read the token storage structure is 7, 8, 9, 10, 11 to read the token 56, 55, 54, 53, 52, and then store the token The flags corresponding to the position pointers 7, 8, 9, 10, and 11 in the structure are set to 0, that is, the flags in the eighth, nine, ten, eleven, and twelve rows are set to zero. The token storage structure is released, and then used again. The tokens 56, 55, 54, 53, 52 are queried to obtain the read address of Read, Reorder, and Buff, so that the correct command 2 data can be obtained. Since the int corresponding to 52 in the token storage structure corresponding to the tokens 56, 55, 54, 53, and 52 is 1, indicating that all the tokens with ID = 0 have been released, the flag corresponding to ID = 0 in the ID read pointer storage structure is released. Need to be cleared and set rd_ptr to 0. After reading the data corresponding to the tokens 56, 55, 54, 53, 52, the rd_ptr corresponding to ID = 0 in the ID read pointer storage structure needs to be cleared.

Optionally, if a new ID or existing ID has a new token to be saved during the token read release process, you need to modify the read and write pointer information of the corresponding ID at the same time, and update the pointer to the command The flag, ind, and next_ptr information in the card storage resource.

With the above solution, the storage resources required to store the token storage structure, ID read pointer storage structure, and ID write pointer storage structure according to the embodiments of the present invention are as follows:

Token storage resource: 2 ^M * (1 + 1 + 1 + M + M) bit

ID read pointer storage resource: N * M bit

ID write pointer storage resource: N * M bit

However, the storage resources required by the existing technology provided in the background technology are: N * 2 ^M * M bit.

Among them, N is the number of IDs of the above instructions, and 2 ^M is the instruction CAM depth of the DDR device.

The values in the following table are the storage resources saved by this proposal compared with the prior art:

N / M	4	5	6	7
4	48bit	184bit	528bit	1352bit
8	272bit	784bit	2016bit	4880bit
16	720bit	1984bit	4992bit	11936bit
32	1616bit	4384bit	10994bit	26048bit

Compared with the prior art, the embodiments of the present invention have the following improvements: multiple IDs share token storage resources, and compared with the prior art, tokens are stored by ID alone, which greatly reduces logical storage resources.

FIG. 13 is a schematic structural diagram of a token management apparatus according to an embodiment of the present invention. As shown in FIG. 13, the token management apparatus 1300 in this embodiment may include a memory 1301 and a processor 1302.

The memory 1301 is configured to store program code;

The processor 1302 is configured to call the code stored in the memory 1301 and execute:

Divide the received instruction into K instruction fragments, and allocate tokens for each instruction fragment in turn;

Store K tokens in the assigned order and generate storage information for each token;

Reorder the K instruction fragments, and send the reordered K instruction fragments;

The storage information of each token is used to indicate whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the instruction in the instruction. The storage address of the token's next token.

In some embodiments, the processor 1302 is further configured to:

If the instruction is the first instruction currently stored corresponding to the corresponding instruction identifier, generating token reading information of the instruction identifier;

The token reading information is used to indicate a storage address of a first token currently stored corresponding to an instruction identifier to which the instruction belongs.

In some embodiments, the storage information of each token is storage information of a first linked list structure; the first linked list structure includes: a first storage instruction, a second storage instruction, and a third storage instruction; A storage instruction is used to indicate whether the token is the last token of the instruction, and the second storage instruction is used to indicate whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the third storage instruction is used Used to indicate the storage address of the next token.

In some embodiments, the first linked list structure further includes: a fourth storage indication, where the fourth storage indication is used to indicate whether a corresponding storage resource has stored a token.

In some embodiments, the processor 1302 is specifically configured to:

For any two adjacent tokens of the K tokens, the i-1th token of the K tokens is stored in an idle first storage resource, and the i-1th token is generated The storage information of the token, the fourth storage indication in the storage information of the i-1th token is used to indicate that the corresponding storage resource has stored the token, and the first storage indication is used to indicate the i-1th The token is not the last token of the instruction, the second storage instruction is used to indicate that the i-1th token is the last token corresponding to the instruction ID to which the instruction belongs, and the third storage instruction is used to Indicates that the storage address of the next token is a preset storage address;

Storing the i-th token in an idle second storage resource, and generating storage information of the i-th token, and a fourth storage indication in the storage information of the i-th token is used to indicate a corresponding The storage resource has stored a token, the second storage instruction is used to indicate that the i-th token is the last token corresponding to the instruction ID to which the instruction belongs, and the third storage instruction is a storage address for the next token It is a preset storage address. When i is equal to K, the first storage instruction is used to indicate that the i-th token is the last token of the instruction. When i is not equal to K, the first storage instruction is used to indicate The i-th token is not the last token of the instruction;

Changing the second storage instruction in the storage information of the i-1th token to indicate that the i-1th token is not the last token corresponding to the instruction ID to which the instruction belongs, and changing the third storage instruction to the instruction A storage address of a token is an address of the second storage resource;

Here, i is an integer of 2 or more and K or less.

In some embodiments, the idle first storage resource and the idle second storage resource are idle adjacent storage resources.

In some embodiments, the token reading information is token reading information of a second linked list structure, and the second linked list structure includes a first reading instruction, and the first reading instruction is used to indicate the reading The storage address of the first token corresponding to the instruction ID to which the instruction belongs.

In some embodiments, the second linked list structure further includes: a second read instruction, which is used to indicate whether a token corresponding to an instruction identifier to which the instruction belongs is stored.

In some embodiments, the processor 1302 is specifically configured to:

Generating the token reading information identified by the instruction after storing the first token of the instruction;

The second read instruction in the token read information is used to indicate that the instruction has been stored with the token, and the first read instruction is used to indicate the storage address of the first token.

In some embodiments, after the processor 1302 stores the (i-1) th token out of the K tokens in an idle first storage resource, it is further configured to: Token write information, where the token write information is used to indicate that a storage address corresponding to a recently stored token corresponding to the instruction identifier is an address of the first storage resource;

When the processor 1302 stores the i-th token in an idle second storage resource, the processor 1302 is specifically configured to: determine the idle second storage resource according to the token write information; and store the i-th token Tokens are stored in an idle second storage resource;

The processor 1302 is further configured to update the token write information after storing the i-th token in an idle second storage resource, and the updated token write information is used to indicate the instruction identifier The storage address corresponding to the recently stored token is the address of the second storage resource.

In some embodiments, the token writing information is token writing information of a third linked list structure, and the third linked list structure includes a writing instruction, where the writing instruction is used to indicate that the instruction identifier corresponds to the most recent The storage address of the stored token.

The token management apparatus of this embodiment may be used to execute the technical solution related to storing tokens in the foregoing method embodiments of the present invention, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

FIG. 14 is a schematic structural diagram of a token management apparatus according to another embodiment of the present invention. As shown in FIG. 14, the token management apparatus 1400 of this embodiment may include a memory 1401 and a processor 1402.

The memory 1401 is configured to store program code;

The processor 1402 is configured to call the code stored in the memory 1401, and execute:

Receiving data read by each instruction fragment in a plurality of instruction fragments, the plurality of instruction fragments including K instruction fragments, and the K instruction fragments are all instruction fragments of a reordered instruction;

Determine the token corresponding to the data read by each instruction fragment;

Determine the storage address indicated by the token read information as the storage address of the first token of the instruction according to the token read information identified by the instruction to which the instruction belongs, and according to the first token The storage address of the first token;

Obtaining the storage information of the other K-1 tokens of the instruction in sequence according to the storage information of the first token;

Obtaining other K-1 tokens according to the storage information of the K-1 tokens in sequence;

According to the order of obtaining the K tokens of the instruction and the tokens corresponding to the data read by each instruction fragment, the data corresponding to the K tokens is reorganized to obtain the data read by the instruction ;

The storage information of each token is used to indicate whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the instruction in the instruction. The storage address of the next token of the token;

The token reading information is used to indicate a storage address of a first token currently stored corresponding to an instruction identifier to which the instruction belongs.

In some embodiments, the storage information of each token is storage information of a first linked list structure; the first linked list structure includes: a first storage instruction, a second storage instruction, and a third storage instruction; A storage indication is used to indicate whether the token is the last token of the instruction, and the second storage indication is used to indicate whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the third storage instruction is used Used to indicate the storage address of the next token.

In some embodiments, the processor 1402 is further configured to: for the storage information of each token, after obtaining the token, change the third storage indication in the storage information of the token to a pre- Set the storage address.

In some embodiments, the processor 1402 is further configured to: for the storage information of each token, after obtaining the token, if the first storage in the storage information of the token indicates the token Is the last token of the instruction, the first storage instruction is changed to indicate that the token is not the last token of the instruction.

In some embodiments, the first linked list structure further includes: a fourth storage indication, where the fourth storage indication is used to indicate whether a corresponding storage resource has stored a token.

In some embodiments, the processor 1402 is further configured to: for the storage information of each token, after acquiring the token, if the fourth storage indication in the storage information of the token is indicated by The stored token changed to indicate that the token was not stored.

In some embodiments, the processor 1402 is further configured to: for the storage information of each token, after obtaining the token, if the second storage in the storage information of the token indicates the token For the last token corresponding to the instruction identifier, the second storage instruction is changed to indicate that the token is not the last token corresponding to the instruction identifier.

In some embodiments, the token reading information is token reading information of a second linked list structure, and the second linked list structure includes a first reading instruction, and the first reading instruction is used to indicate the reading The storage address of the first token of the instruction.

In some embodiments, the processor 1402 is further configured to:

If the second storage indication in the storage information of one of the K tokens indicates that the token is the last token corresponding to the instruction ID, changing the first reading indication to indicate a preset storage address;

If the second storage in the storage information of each of the K tokens indicates that the token is not the last token corresponding to the instruction identifier, changing the first read instruction to indicate the first storage An address; the first storage address is a storage address indicated by a third storage instruction that belongs to the same storage information as the first storage instruction indicating that the token is the last token of the instruction.

In some embodiments, the second linked list structure further includes a second read instruction, where the second read instruction is used to indicate whether a token of the instruction has been stored.

In some embodiments, the processor 1402 is further configured to: if the second storage indication in the storage information of one of the K tokens indicates that the token is the last token corresponding to the instruction identifier, Then changing the second read instruction to a token indicating that the instruction is not stored.

In some embodiments, the processor 1402 is further configured to: if the second storage indication in the storage information of one of the K tokens indicates that the token is the last token identified by the instruction, then Changing the token write information of the instruction identifier to indicate that the storage address of the most recently stored token corresponding to the instruction identifier is a preset storage address.

In some embodiments, the token writing information is token writing information of a third linked list structure, and the third linked list structure includes a writing instruction, where the writing instruction is used to indicate that the instruction identifier corresponds to The storage address of the most recently stored token.

The token management apparatus of this embodiment may be used to execute the technical solution related to reading tokens in the foregoing method embodiment of the present invention, and the implementation principles and technical effects thereof are similar, and details are not described herein again.

FIG. 15 is a schematic structural diagram of a chip according to an embodiment of the present invention. As shown in FIG. 15, the chip 1500 in this embodiment may include a DDR controller 1501 and a DDR device 1502.

The DDR controller 1501 is configured to execute the token management method in the foregoing method embodiments;

The DDR device 1502 is configured to receive the reordered multiple instruction fragments sent by the DDR device, and send data read by each instruction fragment to the DDR controller.

The chip in this embodiment may be used to implement the technical solutions of the foregoing method embodiments of the present invention. The implementation principles and technical effects are similar, and are not described herein again.

FIG. 16 is a schematic structural diagram of a movable platform according to an embodiment of the present invention. As shown in FIG. 16, the movable platform 1600 in this embodiment may include a fuselage 1601, a pan-tilt 1602, and an imaging device 1603. The imaging device 1603 is connected to the body 1601 through the pan / tilt 1602.

The fuselage 1601 includes a chip 1604. The chip may adopt the structure of the embodiment shown in FIG. 15 and correspond to the technical solutions that can implement the foregoing method embodiments of the present invention. The implementation principles and technical effects are similar. Here, No longer.

The chip 1604 is also connected to the imaging device 1603, and the chip 1604 is further configured to process image data collected by the imaging device 1603.

A person of ordinary skill in the art may understand that all or part of the steps of the foregoing method embodiments may be completed by a program instructing related hardware. The foregoing program may be stored in a computer-readable storage medium. When the program is executed, the program is executed. Including the steps of the above method embodiment; and the foregoing storage medium includes: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc. The medium.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, but not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: The technical solutions described in the foregoing embodiments can still be modified, or some or all of the technical features can be equivalently replaced; and these modifications or replacements do not depart from the essence of the corresponding technical solutions of the technical solutions of the embodiments of the present invention. range.

Claims (27)

1. A token management method, comprising:

   Divide the received instruction into K instruction fragments, and allocate tokens for each instruction fragment in turn;

   Store K tokens in the assigned order and generate storage information for each token;

   Reorder the K instruction fragments, and send the reordered K instruction fragments;

   The storage information of each token is used to indicate whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the instruction in the instruction. The storage address of the token's next token.
2. The method according to claim 1, further comprising:

   If the instruction is the first instruction currently stored corresponding to the corresponding instruction identifier, generating token reading information of the instruction identifier;

   The token reading information is used to indicate a storage address of a first token currently stored corresponding to an instruction identifier to which the instruction belongs.
3. The method according to claim 1 or 2, wherein the storage information of each token is storage information of a first linked list structure; and the first linked list structure comprises: a first storage instruction and a second storage instruction And a third storage instruction; the first storage instruction is used to indicate whether the token is the last token of the instruction, and the second storage instruction is used to indicate whether the token is the last instruction corresponding to the instruction identifier to which the instruction belongs The third storage instruction is used to indicate a storage address of a next token.
4. The method according to claim 3, wherein the first linked list structure further comprises: a fourth storage indication, where the fourth storage indication is used to indicate whether a corresponding storage resource has stored a token.
5. The method according to claim 4, wherein said storing K tokens in an allocation order and generating storage information of each token comprises:

   For any two adjacent tokens of the K tokens, the i-1th token of the K tokens is stored in an idle first storage resource, and the i-1th token is generated The storage information of the token, the fourth storage indication in the storage information of the i-1th token is used to indicate that the corresponding storage resource has stored the token, and the first storage indication is used to indicate the i-1th The token is not the last token of the instruction, the second storage instruction is used to indicate that the i-1th token is the last token corresponding to the instruction ID to which the instruction belongs, and the third storage instruction is used to Indicates that the storage address of the next token is a preset storage address;

   Storing the i-th token in an idle second storage resource, and generating storage information of the i-th token, and a fourth storage indication in the storage information of the i-th token is used to indicate a corresponding The storage resource has stored a token, the second storage instruction is used to indicate that the i-th token is the last token corresponding to the instruction ID to which the instruction belongs, and the third storage instruction is a storage address for the next token It is a preset storage address. When i is equal to K, the first storage instruction is used to indicate that the i-th token is the last token of the instruction. When i is not equal to K, the first storage instruction is used to indicate The i-th token is not the last token of the instruction;

   Changing the second storage instruction in the storage information of the i-1th token to indicate that the i-1th token is not the last token corresponding to the instruction ID to which the instruction belongs, and changing the third storage instruction to the instruction A storage address of a token is an address of the second storage resource;

   Here, i is an integer of 2 or more and K or less.
6. The method according to claim 5, wherein the idle first storage resource and the idle second storage resource are idle adjacent storage resources.
7. The method according to claim 2, wherein the token reading information is token reading information of a second linked list structure, and the second linked list structure includes a first reading instruction, and the first reading The fetch instruction is used to indicate a storage address of a first token corresponding to an instruction identifier to which the instruction belongs.
8. The method according to claim 7, wherein the second linked list structure further comprises: a second read instruction, the second read instruction is used to indicate whether a command corresponding to an instruction identifier to which the instruction belongs is stored brand.
9. The method according to claim 8, wherein generating the token reading information identified by the instruction comprises:

   Generating the token reading information identified by the instruction after storing the first token of the instruction;

   The second read instruction in the token read information is used to indicate that the instruction has been stored with the token, and the first read instruction is used to indicate the storage address of the first token.
10. The method according to claim 6, wherein after storing the (i-1) th token out of the K tokens in an idle first storage resource, the method further comprises:

    Generating token writing information of an instruction identifier to which the instruction belongs, where the token writing information is used to indicate that a storage address corresponding to the recently stored token corresponding to the instruction identifier is an address of the first storage resource;

    The storing the i-th token in an idle second storage resource includes: determining an idle second storage resource according to the token write information; and storing the i-th token in an idle second storage resource. In the second storage resource;

    After the i-th token is stored in the idle second storage resource, the method further includes: updating the token writing information, and the updated token writing information is used to indicate that the instruction identifier corresponds to a recently stored The storage address of the token is an address of the second storage resource.
11. The method according to claim 10, wherein the token write information is token write information of a third linked list structure, and the third linked list structure includes a write instruction, and the write instruction is used for Instruct the instruction identifier to correspond to a storage address of a recently stored token.
12. A token management method, comprising:

    Receiving data read by each instruction fragment in a plurality of instruction fragments, the plurality of instruction fragments including K instruction fragments, and the K instruction fragments are all instruction fragments of a reordered instruction;

    Determine the token corresponding to the data read by each instruction fragment;

    Determine the storage address indicated by the token read information as the storage address of the first token of the instruction according to the token read information identified by the instruction to which the instruction belongs, and according to the first token The storage address of the first token;

    Obtaining the storage information of the other K-1 tokens of the instruction in sequence according to the storage information of the first token;

    Obtaining other K-1 tokens according to the storage information of the K-1 tokens in sequence;

    According to the order of obtaining the K tokens of the instruction and the tokens corresponding to the data read by each instruction fragment, the data corresponding to the K tokens is reorganized to obtain the data read by the instruction ;

    The storage information of each token is used to indicate whether the token is the last token of the instruction, whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, and the instruction in the instruction. The storage address of the next token of the token;

    The token reading information is used to indicate a storage address of a first token currently stored corresponding to an instruction identifier to which the instruction belongs.
13. The method according to claim 12, wherein the storage information of each token is storage information of a first linked list structure; the first linked list structure comprises: a first storage instruction, a second storage instruction, and a first Three storage instructions; the first storage instruction is used to indicate whether the token is the last token of the instruction, and the second storage instruction is used to indicate whether the token is the last token corresponding to the instruction identifier to which the instruction belongs, The third storage instruction is used to indicate a storage address of a next token.
14. The method according to claim 13, further comprising:

    For the storage information of each token, after acquiring the token, the third storage indication in the storage information of the token is changed to a preset storage address.
15. The method according to claim 13 or 14, further comprising:

    For the storage information of each token, after obtaining the token, if the first storage in the storage information of the token indicates that the token is the last token of the instruction, the first storage is stored. The instruction changes to indicate that the token is not the last token of the instruction.
16. The method according to any one of claims 13-15, wherein the first linked list structure further comprises: a fourth storage indication, where the fourth storage indication is used to indicate whether a corresponding storage resource has stored a token .
17. The method according to claim 16, further comprising:

    For the storage information of each token, after obtaining the token, if the fourth storage indication in the storage information of the token is changed from indicating that the token is stored to indicating that the token is not stored.
18. The method according to any one of claims 13-17, further comprising:

    For the storage information of each token, after obtaining the token, if the second storage in the storage information of the token indicates that the token is the last token corresponding to the instruction ID, the The second storage instruction is changed to indicate that the token is not the last token corresponding to the instruction ID.
19. The method according to any one of claims 13 to 18, wherein the token reading information is token reading information of a second linked list structure, and the second linked list structure includes a first reading instruction, The first read instruction is used to indicate a storage address of a first token of the instruction.
20. The method according to claim 19, further comprising:

    If the second storage indication in the storage information of one of the K tokens indicates that the token is the last token corresponding to the instruction ID, changing the first reading indication to indicate a preset storage address;

    If the second storage in the storage information of each of the K tokens indicates that the token is not the last token corresponding to the instruction identifier, changing the first read instruction to indicate the first storage An address; the first storage address is a storage address indicated by a third storage instruction that belongs to the same storage information as the first storage instruction indicating that the token is the last token of the instruction.
21. The method according to claim 19 or 20, wherein the second linked list structure further comprises: a second read instruction, the second read instruction is used to indicate whether a token of the instruction has been stored.
22. The method according to claim 21, further comprising:

    If the second storage indication in the storage information of one of the K tokens indicates that the token is the last token corresponding to the instruction ID, changing the second read indication to indicate that no The token of the instruction.
23. The method according to any one of claims 13 to 22, further comprising:

    If the second storage indication in the storage information of one of the K tokens indicates that the token is the last token identified by the instruction, change the token writing information identified by the instruction to the indicated The storage address of the recently stored token corresponding to the instruction identifier is a preset storage address.
24. The method according to claim 23, wherein the token writing information is token writing information of a third linked list structure, and the third linked list structure includes a writing instruction, and the writing instruction is used for The storage address of the recently stored token corresponding to the instruction identifier is indicated.
25. A token management device, comprising: a memory and a processor;

    The memory is used to store program code;

    The processor is configured to call the code stored in the memory and execute the token management method according to any one of claims 1-11, or the token management method according to any one of claims 12-24 Token management method.
26. A chip characterized by comprising: a double data rate DDR controller and a DDR device;

    The DDR controller is configured to execute the token management method according to any one of claims 1-11, and the token management method according to any one of claims 12-24;

    The DDR device is configured to receive the reordered multiple instruction fragments sent by the DDR device, and send data read by each instruction fragment to the DDR controller.
27. A movable platform, characterized by a fuselage, a gimbal and an imaging device, wherein the imaging device is connected to the fuselage through the gimbal, and the fuselage includes the chip according to claim 26;

    The chip is also connected to the imaging device, and the chip is further configured to process image data collected by the imaging device.

Images