CN115955441A - Management scheduling method and device based on TSN queue - Google Patents

Abstract

The invention discloses a management scheduling method and a device based on a TSN queue, wherein the management scheduling method based on the TSN queue comprises the following steps: acquiring application information of a data frame to be enqueued; judging whether the application information of the data frame to be enqueued meets enqueuing conditions or not, if so, enqueuing the data frame to be enqueued and updating queue information; and judging whether to execute dequeue scheduling on the enqueue according to the queue information, if so, executing dequeue scheduling, and dequeuing the enqueue in the queue according to a dequeue scheduling result. The invention obtains the queue scheduling information according to the application information and the current queue information of the data frame to be enqueued, and then enqueues or dequeues the data frame according to the scheduling information, and realizes the improvement of the characteristics of the data stream such as the delay characteristic, the frame loss rate and the like by managing and scheduling the enqueue and dequeue processes of the TSN queue.

Description

Management scheduling method and device based on TSN queue

Technical Field

The invention relates to the technical field of TSN queue scheduling, in particular to a management scheduling method based on a TSN queue and a management scheduling device based on the TSN queue.

Background

With the continuous development of automobile science and technology, intellectualization and networking, the latest applications and functions in automobiles are continuously increasing the requirements for bandwidth, delay reduction, high availability, qoS and cost reduction. Current and traditional automotive network protocols are not sufficient to meet these upcoming demands, and under such circumstances, there is an urgent need for a vehicle-mounted network that has high bandwidth, is also open, extensible, has strong compatibility, and is convenient for network aggregation, and that simultaneously meets vehicle-mounted strict regulatory requirements, vehicle-mounted electrical environment, and high reliability requirements.

The vehicle-mounted Ethernet is a novel local area network technology for connecting electronic units in a vehicle through the Ethernet, and different from the traditional Ethernet which uses 4 pairs of unshielded twisted pair cables, the vehicle-mounted Ethernet can realize the transmission rate of 100Mbit/s or even 1000Mbit/s on a single pair of unshielded twisted pair cables, and simultaneously, the vehicle-mounted Ethernet also meets the requirements of the automobile industry on high reliability, low electromagnetic radiation, low power consumption, bandwidth allocation, low delay, synchronous real-time property and the like.

The TSN (time sensitive network), i.e. the protocol family that implements deterministic minimum time delay in non-deterministic ethernet, is a set of protocol standards developed by the TSN working group in the IEEE802.1 working group, defines a time sensitive mechanism for ethernet data transmission, and adds determinacy and reliability to the conventional ethernet to ensure real-time, deterministic and reliable data transmission. The circular queue forwarding mechanism is an important component of a bounded low-delay technology in the TSN, and since a credit-based shaper is adopted in the TSN to shape data traffic, the credit-based shaper can only implement soft real-time performance on a data stream, and as the topology of a data stream transmission path increases, the delay of the data stream will accumulate and increase. And in the worst case, the delay of the data flow is related to the network topology, the number of the data flow and the queue buffer condition of the network node.

Disclosure of Invention

The present invention is directed to a management scheduling method based on a TSN queue and a management scheduling apparatus based on a TSN queue, to solve at least one of the above-mentioned problems.

The invention provides the following scheme:

according to an aspect of the present invention, a management scheduling method based on a TSN queue is provided, including:

acquiring application information of a data frame to be enqueued;

judging whether the application information of the data frame to be enqueued meets enqueuing conditions or not, if so, judging that the application information of the data frame to be enqueued meets the enqueuing conditions

Enqueuing the data frame to be enqueued and updating queue information;

judging whether to execute dequeue scheduling on the enqueue queue according to the queue information, if so, determining whether to execute dequeue scheduling on the enqueue queue

And performing dequeue scheduling, and dequeuing the enqueue in the queue according to a dequeue scheduling result.

Optionally, the determining whether the application information of the data frame to be enqueued satisfies enqueuing conditions includes:

acquiring an initial allocation queue number according to the application information of the data frame to be enqueued;

judging whether the initial distribution queue is a circular queue according to the initial distribution queue number, if so, judging whether the initial distribution queue is the circular queue

Acquiring a storage space and gating information of an initial distribution queue;

judging whether the storage space of the initial distribution queue can store the data frame to be enqueued and whether the gating information of the initial distribution queue is in an open state, if so, judging whether the gating information of the initial distribution queue is in the open state

Performing circular queue logic enqueuing on the data frame to be enqueued and updating logic enqueuing queue information;

and performing circular queue physical enqueuing on the data frames to be enqueued according to the logical enqueue information and updating the physical enqueue information.

Optionally, the determining whether the application information of the data frame to be enqueued satisfies enqueuing conditions includes:

judging whether the initial distribution queue is a common queue according to the initial distribution queue number, if so, judging whether the initial distribution queue is the common queue

Acquiring queue information of a common queue;

judging whether the common queue can store the data frame to be enqueued according to the queue information of the common queue and the application information of the data frame to be enqueued, if so, judging whether the common queue can store the data frame to be enqueued, and if so, storing the data frame to be enqueued in the common queue

Performing common queue logic enqueuing on the data frames to be enqueued and updating logic enqueuing queue information;

and performing common queue physical enqueue on the data frames to be enqueued according to the logical enqueue information and updating the physical enqueue information.

Optionally, the determining whether to perform dequeue scheduling on the enqueue queue according to the queue information includes:

acquiring a non-empty queue according to the logic enqueue information;

acquiring queue information of the non-empty queue;

judging whether the non-empty queue can be dequeued or not according to the queue information of the non-empty queue, if so, determining that the non-empty queue can be dequeued

According to the queue information of the non-empty queue, carrying out dequeue priority ordering on the non-empty queue to obtain a logic dequeue queue;

logically dequeuing the non-empty queue according to the logical dequeue queue, and updating a physical dequeue queue;

and physically dequeuing the non-empty queue according to the physical dequeue queue.

Optionally, the queue information of the non-empty queue includes a queue number of the non-empty queue, a queue number status, frame length information of a data frame of the non-empty queue, a buffer address of the data frame of the non-empty queue, a source port number of the data frame of the non-empty queue, an output port number of the non-empty queue, and a port status of an output port of the non-empty queue.

Optionally, the determining, according to the queue information of the non-empty queue, whether the non-empty queue can be dequeued includes:

judging whether the port state of the non-empty queue output port is idle or not, if so, judging whether the port state of the non-empty queue output port is idle or not, and if not, judging whether the port state of the non-empty queue output port is idle or not

And carrying out dequeue priority sequencing on the non-empty queue according to the queue number state of the non-empty queue and the output port number of the non-empty queue to obtain a logic dequeue queue.

Optionally, the physically dequeuing the non-empty queue according to the physical dequeue queue includes:

judging whether the non-empty queue is a circular queue according to queue information of the non-empty queue, if so, judging whether the non-empty queue is the circular queue

Judging whether the port state of the non-empty queue output port is an open state or not and the first receiving end corresponding to the non-empty queue output port can normally receive the non-empty queue, if so, judging that the port state of the non-empty queue output port is an open state and the first receiving end corresponding to the non-empty queue output port can normally receive the non-empty queue

And physically dequeuing the non-empty queue according to the physical dequeue queue.

Optionally, the physically dequeuing the non-empty queue according to the physical dequeue queue includes:

judging whether the non-empty queue is a common queue according to the queue information of the non-empty queue, if so, judging whether the non-empty queue is the common queue

Judging whether a second receiving end corresponding to the non-empty queue output port can receive the non-empty queue according to the queue information of the non-empty queue, if so, judging that the second receiving end corresponding to the non-empty queue output port can receive the non-empty queue

And physically dequeuing the non-empty queue according to the physical dequeue queue.

Optionally, the management scheduling method based on the TSN queue further includes:

judging whether the number of the data frames to be enqueued meets the preset protection condition or not when the data frames to be enqueued are enqueued, if so, determining that the number of the data frames to be enqueued meets the preset protection condition

And updating queue information when the data frames to be enqueued are enqueued.

The invention also provides a management scheduling device based on the TSN queue, which comprises:

the application information acquisition module is used for acquiring application information of a data frame to be enqueued;

an enqueue judging module, configured to judge whether the application information of the data frame to be enqueued satisfies an enqueue condition, and if so, determine whether the application information of the data frame to be enqueued satisfies the enqueue condition

The enqueuing module is used for enqueuing the data frames to be enqueued and updating queue information;

a dequeue judgment module, configured to judge whether to perform dequeue scheduling on an enqueue queue according to the queue information, and if yes, determine whether to perform dequeue scheduling on the enqueue queue

And the dequeue module is used for executing dequeue scheduling and dequeuing the enqueue in the queue according to a dequeue scheduling result.

Compared with the prior art, the invention has the following advantages:

the method comprises the steps of obtaining queue scheduling information according to application information and current queue information of a data frame to be enqueued, and further carrying out enqueuing and dequeuing operations on the data frame according to the scheduling information; by scheduling the core functions of buffer allocation of the queues, enqueue operation and dequeue operation of the queues and the like, the invention can simultaneously satisfy the characteristics of improving the time delay characteristic and frame loss rate and the like of the data flow on the limited resources of the physical hardware platform, and can better realize the queue management of the data flow with different requirements by the compatibility of two buffer allocation modes.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a flow chart of a management scheduling method based on TSN queues according to the present invention;

fig. 2 is a structural diagram of a shared bus switch based on a TSN queue management scheduling method according to an embodiment of the present invention;

fig. 3 is a schematic diagram illustrating an overall design of queue management scheduling based on a TSN queue management scheduling method according to an embodiment of the present invention;

FIG. 4 is a flow chart illustrating a logical enqueue procedure of a TSN queue-based management scheduling method according to an embodiment of the present invention;

fig. 5 is a flowchart illustrating a general queue enqueue determining method based on a TSN queue management scheduling method according to an embodiment of the present invention;

FIG. 6 is a flow chart of a physical enqueue process of a management scheduling method based on TSN queues according to an embodiment of the present invention;

FIG. 7 is a flow chart of scheduling control of a management scheduling method based on TSN queues according to an embodiment of the present invention;

FIG. 8 is a flow chart illustrating dequeue scheduling in a TSN queue-based management scheduling method according to an embodiment of the present invention;

FIG. 9 is a flow chart illustrating logical dequeue of a method for managing scheduling based on TSN queues according to an embodiment of the present invention;

FIG. 10 is a flowchart illustrating a method for managing and scheduling round-robin queue physical dequeuing based on TSN queues according to an embodiment of the present invention;

FIG. 11 is a diagram illustrating a normal queue physical dequeue flow of a management scheduling method based on a TSN queue according to an embodiment of the present invention;

fig. 12 is a schematic structural diagram of a management scheduling apparatus based on TSN queues according to an embodiment of the present invention;

fig. 13 is a structural diagram of an electronic device that can implement the TSN queue-based management scheduling method of the present invention.

Detailed Description

The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

FIG. 1 is a flow chart of a management scheduling method based on TSN queues according to the present invention;

as shown in fig. 1, a management scheduling method based on a TSN queue includes:

step 1: acquiring application information of a data frame to be enqueued;

step 2: judging whether the application information of the data frame to be enqueued meets the enqueuing condition, if so, judging that the application information of the data frame to be enqueued meets the enqueuing condition

And step 3: enqueuing the data frames to be enqueued and updating queue information;

and 4, step 4: judging whether to execute dequeue scheduling on the enqueue queue according to the queue information, if so, determining whether to execute dequeue scheduling on the enqueue queue

And 5: and performing dequeue scheduling, and dequeuing the enqueue in the queue according to a dequeue scheduling result.

The method comprises the steps of obtaining queue scheduling information according to application information and current queue information of a data frame to be enqueued, and further carrying out enqueuing and dequeuing operations on the data frame according to the scheduling information; by scheduling the core functions of buffer allocation of the queues, enqueue operation and dequeue operation of the queues and the like, the invention can simultaneously satisfy the characteristics of improving the time delay characteristic and frame loss rate and the like of the data flow on the limited resources of the physical hardware platform, and can better realize the queue management of the data flow with different requirements by the compatibility of two buffer allocation modes.

In this embodiment, the application information of the data frame to be enqueued includes the length of the data frame;

in this embodiment, the determining whether the application information of the data frame to be enqueued satisfies the enqueuing condition includes:

acquiring an initial allocation queue number according to application information of a data frame to be enqueued;

judging whether the initial distribution queue is a circular queue according to the initial distribution queue number, if so, judging whether the initial distribution queue is the circular queue

Acquiring a storage space and gating information of an initial distribution queue;

judging whether the storage space of the initial distribution queue can store the data frame to be enqueued and whether the gating information of the initial distribution queue is in an open state, if so, judging that the data frame to be enqueued can be stored in the storage space of the initial distribution queue and the gating information of the initial distribution queue is in the open state

Performing circular queue logic enqueuing on the data frame to be enqueued and updating logic enqueuing queue information;

and performing circular queue physical enqueuing on the data frame to be enqueued according to the logical enqueue information and updating the physical enqueue information.

In this embodiment, determining whether the application information of the data frame to be enqueued satisfies the enqueuing condition includes:

judging whether the initial distribution queue is a common queue according to the initial distribution queue number, if so, judging whether the initial distribution queue is the common queue

Acquiring queue information of a common queue, wherein the queue information comprises a queue length, a minimum threshold value, a maximum threshold value and a residual public buffer space;

judging whether the common queue can store the data frame to be enqueued according to the queue information of the common queue and the application information of the data frame to be enqueued, if so, judging whether the common queue can store the data frame to be enqueued

Performing common queue logic enqueue on a data frame to be enqueued and updating logic enqueue information;

and performing common queue physical enqueuing on the data frames to be enqueued according to the logical enqueue information and updating the physical enqueue information.

In this embodiment, determining whether to perform dequeue scheduling on the enqueue according to the queue information includes:

acquiring a non-empty queue according to the logic enqueue information;

acquiring queue information of a non-empty queue;

judging whether the non-empty queue can be dequeued or not according to the queue information of the non-empty queue, and if so, determining that the non-empty queue can be dequeued

According to the queue information of the non-empty queue, carrying out dequeue priority ordering on the non-empty queue to obtain a logic dequeue queue;

logically dequeuing the non-empty queue according to the logical dequeue queue, and updating the physical dequeue queue;

and physically dequeuing the non-empty queue according to the physical dequeue queue.

In this embodiment, the queue information of the non-empty queue includes a queue number of the non-empty queue, a queue number state, frame length information of a data frame of the non-empty queue, a buffer address of the data frame of the non-empty queue, a source port number of the data frame of the non-empty queue, an output port number of the non-empty queue, and a port state of an output port of the non-empty queue.

In this embodiment, determining whether the non-empty queue can be dequeued according to the queue information of the non-empty queue includes:

judging whether the port state of the non-empty queue output port is idle or not, if so, judging whether the port state of the non-empty queue output port is idle or not, and if not, judging whether the port state of the non-empty queue output port is idle or not

And carrying out dequeue priority sequencing on the non-empty queue according to the queue number state of the non-empty queue and the output port number of the non-empty queue to obtain a logic dequeue queue.

In this embodiment, the physically dequeuing the non-empty queue according to the physical dequeue queue includes:

judging whether the non-empty queue is a circular queue according to the queue information of the non-empty queue, if so, judging whether the non-empty queue is the circular queue

Judging whether the port state of the non-empty queue output port is an open state or not and the first receiving end corresponding to the non-empty queue output port can normally receive the non-empty queue, if so, judging whether the port state of the non-empty queue output port is an open state or not, and if not, judging whether the first receiving end corresponding to the non-empty queue output port can normally receive the non-empty queue

And physically dequeuing the non-empty queue according to the physical dequeue queue.

In this embodiment, the physically dequeuing the non-empty queue according to the physical dequeue queue includes:

judging whether the non-empty queue is a common queue according to the queue information of the non-empty queue, if so, judging whether the non-empty queue is the common queue

Judging whether a second receiving end corresponding to the non-empty queue output port can receive the non-empty queue according to the queue information of the non-empty queue, if so, judging whether the second receiving end corresponding to the non-empty queue output port can receive the non-empty queue

And physically dequeuing the non-empty queue according to the physical dequeue queue.

In this embodiment, the management scheduling method based on the TSN queue further includes:

judging whether the number of the data frames to be enqueued meets the preset protection condition when the data frames to be enqueued are enqueued, if so, determining that the number of the data frames to be enqueued meets the preset protection condition

And updating the queue information when the data frame to be enqueued is enqueued.

Fig. 2 is a structural diagram of a shared bus switch based on a TSN queue management scheduling method according to an embodiment of the present invention;

as shown in fig. 2, the management scheduling method based on the TSN queue of the present invention is an important component of a network switching node, where a data stream is input into an MAC IP core from a PHY side interface, the MAC IP core accesses an input polling module after performing a receiving process on the data stream, and then the data stream is imported into a packet process to perform stream classification, forwarding table lookup and learning, etc., so as to generate enqueue information and scheduling information. And the queue management scheduling module performs enqueue and dequeue operations on the data frames according to the enqueue information and the scheduling information, and accumulates the enqueue/dequeue delays through receiving/sending processing delay of the MAC IP core, flow classification/forwarding table searching delay of packet processing and enqueue/dequeue delay of queue management.

Fig. 3 is a schematic diagram illustrating an overall design of a queue management scheduling module of a management scheduling method based on a TSN queue according to an embodiment of the present invention;

as shown in fig. 3, in an embodiment of the present invention, a queue management scheduling module is used for design and description, wherein the queue management scheduling module is mainly divided into a cache management module, an enqueue scheduling module, a dequeue scheduling module, and a queue information management module. In this embodiment, the queue management scheduling module adopts a shared bus structure and a design idea of a pipeline, and needs to satisfy the following design requirements:

the method supports 6 output ports, the bandwidth of each output port is 1Gbps, and the total line rate is 6 Gbps;

support 6-level priority based on output ports;

a cyclic queue forwarding mechanism is supported;

quality of service guarantees based on different priority queues are supported.

When the input data frame of each port is the shortest frame and the port rate is 1Gbps, the bearing pressure of the network node is the largest. According to calculation, when the system clock is 125MHz and the data bit width is 128bits, in the worst case, each functional module in the queue management and scheduling module must complete the functional response of the response within 10 clocks.

In this embodiment, the enqueue scheduling module is mainly divided into a logical enqueue control module and a physical enqueue control module. The logical enqueuing control module finishes logical enqueuing of the data frames, finishes judgment on whether the data frames can be enqueued, applies a buffer space for the data frames which are successfully enqueued, and provides a data frame marking signal which fails to be enqueued for the physical enqueuing module. And the physical enqueuing control module receives the logical enqueuing information provided by the logical enqueuing control module, discards the data frame failed in the logical enqueuing, and moves the data frame successfully subjected to the logical enqueuing to the cache management module to complete the physical enqueuing of the data frame.

In this embodiment, the queue information management module is divided into a head-of-line update module, a tail-of-line update module, and a captain information update module. When the queue is logically enqueued and logically dequeued, information such as a head address, a tail address and a queue length of the queue is provided, and when the queue is physically enqueued and physically dequeued, the information such as the head address, the tail address and the queue length of the queue is updated.

In this embodiment, the cache management module is mainly divided into three parts, namely a storage information management module, a circular queue storage space module, and a common queue storage space module.

The storage information management module is mainly responsible for distribution and recovery of storage space;

the circular queue storage space module provides storage space for a circular queue;

the common queue storage space module provides a chain type storage structure based on a fixed-length cache space for a common queue;

the buffer management module provides a space entity for the storage of the data frame, so that the storage mode of the data frame and the distribution mode of the buffer space determine the overall logic design of the queue management module; the number of circular queues is small, the occupied buffer space is relatively small, and the circular queues have determinacy, so that a space storage mode with fixed length can be adopted to reduce the design complexity. The number of queues of the ordinary queues is large, a fixed buffer space is allocated for each queue, great hardware overhead is needed, and the randomness of the ordinary queues can cause great buffer waste, so that a chain type storage structure based on a fixed-length buffer space is adopted.

In this embodiment, the dequeue scheduling module is mainly divided into a scheduling control module, a logical dequeue control module, and a physical dequeue control module. The logic scheduling control module schedules queues with different priorities and provides dequeuing scheduling information to the logic dequeuing control module. The logical dequeue control module receives the logically dequeued queue number and provides it to the physical dequeue control module. And the physical dequeue control module receives the frame length, the buffer address and the like of the dequeue data frame, and moves the data frame from the buffer space to the MAC core for output.

The scheduling control module sends dequeued queue numbers to the scheduling information FIFO in sequence according to a scheduling algorithm, the logic dequeuing control module reads the queue numbers after detecting that the scheduling information FIFO is not empty, then inquires the queue head information of the queues, then inquires the residual cache addresses of the data frames from the cache management module in sequence, sends the dequeuing information to the physical dequeuing information FIFO, and the physical dequeuing control module carries out physical dequeuing according to the information in the physical dequeuing information FIFO.

FIG. 4 is a flow chart illustrating logical enqueuing for a method for managing and scheduling based on TSN queues according to an embodiment of the present invention;

as shown in fig. 4, the logical enqueue of the TSN queue-based management scheduling method mainly completes the judgment of the logical enqueue application of the data frame, if the enqueue application is successful, applies for a buffer space for the data frame, and if the enqueue application is failed, sends an empty data instruction to the last processing node of the data frame.

The logical enqueue states are detailed as follows:

IDLE state, when the initialization completion mark is low, jumping into the initialization state, when the initialization completion mark is raised and the data frame carries out enqueue application, jumping into the queue information reading state;

INIT: initializing the state, initializing linked list information and queue information, jumping to an IDLE state after initialization is completed, and pulling up an initialization completion identifier;

READ _ QUEUE _ INFO: reading a QUEUE information reading state, reading an initial allocation QUEUE number and a frame length of a data frame applying for enqueuing, reading information such as a QUEUE length, a minimum threshold value, a maximum threshold value, a buffer space use state and the like of the QUEUE, judging whether the QUEUE is a circular QUEUE or a common QUEUE according to the QUEUE number, jumping the circular QUEUE into a circular _ QUEUE _ JUDGE state, and jumping the common QUEUE into a GENERAL _ QUEUE _ JUDGE state;

CYCLE _ quese _ JUDGE: a circular queue enqueuing application judgment state, if the storage space of the current circular queue can store data frames and the queue gating is opened, the data frames are judged to be enqueued, and the data frames jump into a WR _ ENQUEUE _ RESULT state; otherwise, judging that enqueuing FAILs, and jumping into a JUDGE _ FAIL state;

GENERAL _ QUEUE _ JUDGE: judging the state of the common queue enqueue application, judging whether the data frames can be enqueued according to the length of the current data frame, the queue length, the minimum threshold value, the maximum threshold value and the public cache space of the residual space, and skipping to enter if judging that the data frames can be enqueued

WR _ ENQUEUE _ RESULT state, if it is determined that ENQUEUE FAILs, jump to JUDGE _ FAIL state.

Fig. 5 is a flowchart illustrating a general queue enqueue determining method based on a TSN queue management scheduling method according to an embodiment of the present invention;

as shown in fig. 5, the general queue enqueue determining process of the TSN queue-based management scheduling method includes:

WR _ ENQUEUE _ RESULT: updating the state of logic ENQUEUE information, if enqueued is a common queue, updating the head address, tail address, queue length information and the like of the logically enqueued queue for an enqueued data frame, if enqueued is a circular queue, updating an information bit code number table in the circular queue, simultaneously writing the enqueued information into the enqueued information, and jumping into an ENQUEUE _ FINISH state after all operations are finished;

judgge _ FAIL: the enqueue failure state is that the data frame information of the enqueue failure is sent to a processing node grouping processing module on the data frame so that the grouping processing module can empty the data frame and then jump into an IDLE state;

ENQUEUE _ FINISH: and (4) enqueuing to be completed, and jumping to an IDLE state from the next state.

The logical enqueue can complete the data frame enqueue request processing of the circular queue within 6 clocks, the processing time depends on the frame length of the enqueue data frame for the enqueue request of the common queue, and the processing time is 8 clocks for the shortest data frame of 64 Bytes.

FIG. 6 is a flow chart illustrating a physical enqueue procedure of a TSN queue-based management scheduling method according to an embodiment of the present invention;

as shown in fig. 6, the physical enqueue of the TSN queue-based management scheduling method mainly moves the logically enqueued data frame from the previous processing node packet processing module to the buffer management module, completes the physical enqueue of the data frame, and sends a physical enqueue completion flag to the dequeue scheduling module.

The physical enqueue states are illustrated as follows:

IDLE: an idle state, jumping into a READ _ INFO state when the enqueue information is not empty;

READ _ INFO: reading the state of the physical enqueue information, judging the type of the QUEUE according to the QUEUE number, and jumping into a WRITE _ CYCLE _ QUEUE state if the QUEUE is a circular QUEUE, or jumping into a WRITE _ GENERAL _ QUEUE state if the QUEUE is a common QUEUE;

WRITE _ CYCLE _ QUEUE: a circular queue moving state, namely moving the current enqueue data frame from the grouping processing module to a circular queue caching module, and jumping into a WRITE _ FINISH state after the moving is finished;

WRITE _ GENERAL _ quese: a common queue moving state, namely moving the current enqueue data frame from the grouping processing module to a common queue caching module, and jumping into a WRITE _ FINISH state after the moving is finished;

WRITE _ FINISH: and in the physical enqueue ending state, sending an enqueue number and an enqueue completion mark of the current physical enqueue data frame to the dequeue scheduling module, and then jumping to the IDLE state.

After the data frame is physically enqueued, the data frame can be really moved from the buffer for packet processing to the buffer management for queue management, and the scheduling control module can start dequeuing the data frame to prevent the phenomenon of false dequeuing.

In another embodiment, the information of the head of the queue, the tail of the queue, and the length of the queue is also updated when data is logically enqueued and logically dequeued, specifically, the information of the head of the queue address, the tail of the queue address, the length of the queue, and the like, of the queue is provided when the queue is logically enqueued and logically dequeued, and the information of the head of the queue address, the tail of the queue address, the length of the queue, and the like, of the queue is updated when the queue is physically enqueued and physically dequeued.

The first-queue updating specifically updates the first BD address of the first data frame of the queue and the frame length of the first data frame, so that the dequeue scheduling module dequeues the first frame of the queue.

The method comprises the steps that a true double-port RAM with the bit width of 32bits and the address depth of 256 is maintained by the queue head updating, each address corresponds to the queue head state of one queue, the queue head updating needs to solve the problem that only one data frame exists in the queue, and the queue head is lost when the queue is subjected to enqueuing and dequeuing simultaneously. After the dequeue scheduling module finishes dequeuing, all data frames in the queue are considered to be dequeued, so that the queue head information is updated to zero, and the queue head is lost.

The solution to the loss of the head of the queue is to add interactive signals during the enqueue and dequeue processes. When an enqueue operation is carried out, only one frame in the queue is detected, and when the queue is dequeuing, the queue head information is actively updated. And after the updated queue head information is detected during dequeuing, zero clearing operation is not carried out on the queue head information.

The queue tail updating specifically is to update the queue tail BD address of the last data frame in the queue, and link the data frame to the queue tail when enqueuing. The queue tail updating is responsible for maintaining a true double-port RAM with 16bits of bit width and 256 address depths, each address corresponds to the BD state of the queue tail of a queue, and similar to the phenomenon of queue head loss, when only one data frame exists in the queue and the queue performs enqueuing and dequeuing simultaneously, the phenomenon of queue tail loss can occur.

The solution to the phenomenon of queue tail loss is to detect that there is only one frame in the queue when the enqueue operation is performed, and actively update the queue tail information when the queue is dequeuing. And after the updated queue tail information is detected during dequeuing, zero clearing operation is not carried out on the queue tail information.

The queue length updating specifically is to update the number of BDs occupied by data frames already enqueued in each queue, and when applying for enqueuing, provide queue length information to perform enqueuing judgment.

The queue length update is responsible for maintaining a true double-port RAM with the bit width of 32bits and the address depth of 256.

FIG. 7 is a flow chart of scheduling control of a management scheduling method based on TSN queues according to an embodiment of the present invention;

as shown in fig. 7, the scheduling control of the TSN queue-based management scheduling method includes:

and planning the dequeuing sequence of the non-empty queue according to a scheduling algorithm. Specifically, the scheduling control is responsible for quickly determining the non-empty queues and scheduling the non-empty queues according to a scheduling algorithm.

The scheduling control states are explained as follows:

INIT: initializing a state, and assigning initial values to a scheduling information RAM and a bit code table, wherein the initial values are 0;

IDLE: and in the IDLE state, when the ready signal is pulled high, the GET state is jumped into, and otherwise, the IDLE state is maintained. The dual-port RAM and the bit code table are not in an updated state, and when the physical enqueue completion identifier is pulled high, the ready signal is pulled high;

GET: acquiring a queue information state, acquiring the number of data frames physically enqueued in the queue according to the queue number of the physical enqueue, and jumping to an UPDATE state after the queue information is acquired;

UPDATE: updating the queue information, updating the number of the frames of the entered physical enqueue data stored in the scheduling information RAM, setting a corresponding bit in a bit code table to be 1, and then jumping to a WAIT state;

WAIT: a pause state, namely waiting for the information in the scheduling information RAM to be updated actually, and then jumping into a FINISH state;

FINISH: and updating the completion state and jumping into an IDLE state.

The dequeue scheduling is mainly responsible for dequeue scheduling, and sends a queue number for scheduling dequeue to the logic dequeue control module for logic dequeue, and the description of each state is as follows:

READ _ INFO: and reading the queue information state, wherein the state is the core state for finishing queue scheduling. The design adopts a two-stage scheduling mode based on ports and priority levels for scheduling. Firstly, inquiring a bit code table, wherein the priority of a queue corresponding to multicast/unicast is the highest, and secondly, polling is carried out according to the unicast queue of each port. And jumping to a JUDGE state after information extraction is completed.

JUDGE: and judging the state, wherein in the IDLE state, the port state and the output port number of the scheduling queue cannot be predicted, so that scheduling judgment needs to be carried out after the queue number is determined. And judging whether the corresponding output port of the currently dequeued queue is idle, if so, judging that the scheduling is successful, and jumping into a WRITE _ TXFIFO state, otherwise, jumping into a READ _ NEXT state.

READ _ NEXT: and reading the information state of the next queue, jumping to the READ _ INFO state after the scheduling judgment fails, skipping the current port, and polling the output port corresponding to the next non-empty queue.

WRITE _ TXFIFO: and updating the state of the scheduling result, sending the queue number of the queue successfully scheduled to the logic dequeuing control module to perform logic dequeuing, updating the bit code table and the scheduling information RAM, and jumping to the IDLE state.

FIG. 8 is a flow chart illustrating dequeue scheduling in a TSN queue-based management scheduling method according to an embodiment of the present invention;

as shown in fig. 8, the dequeue schedule is mainly responsible for dequeue scheduling, and sends the queue number scheduled to dequeue to the logical dequeue control module for logical dequeue, and the descriptions of the states are as follows:

READ _ INFO: and reading the queue information state, wherein the state is the core state for finishing queue scheduling. The design adopts a two-stage scheduling mode based on ports and priority levels for scheduling. Firstly, inquiring a bit code table, wherein the queue corresponding to multicast/unicast has the highest priority, and secondly, polling is carried out according to the unicast queue of each port. And jumping to a JUDGE state after information extraction is completed.

JUDGE: and judging the state, wherein in the IDLE state, the port state and the output port number of the scheduling queue cannot be predicted, so that scheduling judgment needs to be carried out after the queue number is determined. And judging whether the corresponding output port of the currently dequeued queue is idle, if so, judging that the scheduling is successful, and jumping into a WRITE _ TXFIFO state, otherwise, jumping into a READ _ NEXT state.

READ _ NEXT: and reading the information state of the next queue, jumping to the READ _ INFO state after the scheduling judgment fails, skipping the current port, and polling the output port corresponding to the next non-empty queue.

WRITE _ TXFIFO: and updating the state of the scheduling result, sending the queue number of the successfully scheduled queue to a logic dequeuing control module to perform logic dequeuing, updating the bit code table and the scheduling information RAM, and jumping to the IDLE state.

FIG. 9 is a flow chart illustrating logical dequeue of a method for managing scheduling based on TSN queues according to an embodiment of the present invention;

as shown in fig. 9, the logical dequeue acquires the buffer address and the frame length information of the dequeued data frame according to the received dequeue number, sequentially queries the buffer address of the data frame, and writes the physical dequeue information into the physical dequeue information FIFO to complete the physical dequeue.

The logic dequeues the various states as follows:

IDLE: idle state when detecting the storage of a dequeue number from a scheduling control module

Jumping into a READ _ TXFIFO when the FIFO is not empty;

READ _ TXFIFO: reading a logic dequeue queue number state, wherein the state is divided into 8 sub-states, the 8 sub-states are 8 sub-states which are in a dequeue information FIFO data format, each sub-state corresponds to a priority, and jumping to a READ _ FRAME _ INFO state after reading a dequeue FRAME queue number;

READ _ FRAME _ INFO: reading the dequeue frame information state, wherein the information comprises frame length information of the data frame, a cache address of the data frame, a source port number of the data frame, an output port number and the like;

WRITE _ SR: the method comprises the steps that cache query and physical dequeue information FIFO updating states are realized, if a dequeue queue is a circular queue, physical dequeue information is directly written into the physical dequeue information FIFO, if the dequeue queue is a common queue, a cache query application is sent to a cache management module according to the first BD block address of a data frame, all BD block addresses and BD block information of the data frame are read out and are respectively written into 8 physical dequeue information FIFOs according to priorities, and a physical module dequeue control module carries out physical dequeue;

TX _ DONE: the logic dequeues complete state and the next clock jumps into the IDLE state.

FIG. 10 is a flowchart illustrating a method for managing and scheduling round-robin queue physical dequeuing based on TSN queues according to an embodiment of the present invention;

as shown in fig. 10, the physical dequeuing is specifically to read physical dequeuing information in the physical dequeuing information FIFO, and move the data frame from the buffer to the MAC IP core, thereby completing the physical dequeuing of the data frame. Because the priorities of the circular queue and the common queue are different and the output channels are different, a double-state machine design is adopted. The output channel of the circular queue corresponds to an eMAC interface of the MAC IP core, and the output channel of the common queue corresponds to a pMAC interface of the MAC IP core.

The states of the circular queue physical dequeue are illustrated as follows:

IDLE: in the initial state, jumping to a PRE _ READ _ INFO state when a physical dequeue information FIFO corresponding to the circular queue is not empty, and otherwise, keeping an IDLE state;

PRE _ READ _ INFO: pre-reading the dequeue information state, acquiring key information such as a dequeue port number, a frame length and the like in the physical dequeue information FIFO, and then jumping into the JUDGE state, wherein the FIFO is arranged by adopting the first character, so that the data in the dequeue information FIFO cannot be clearly lost in the state;

JUDGE: judging the physical dequeue state of the circular queue, judging whether dequeue can be realized or not when the gating of the output port is changed into an open state and the MAC IP core can receive a dequeue data frame, and jumping into a WRITE _ CYCLE state or not;

WRITE _ CYCLE: and in the physical dequeue state of the circular queue, the dequeued data frame is moved to the MAC IP core from the circular queue cache, the dequeue information in the corresponding physical dequeue information FIFO is emptied at the same time, the physical dequeue of the data frame is completed, and then the IDLE state is jumped into.

FIG. 11 is a diagram illustrating a normal queue physical dequeue flow of a management scheduling method based on a TSN queue according to an embodiment of the present invention;

as shown in fig. 11, the states of the physical dequeue of the normal queue are illustrated as follows:

IDLE2: an initial state, namely jumping into a PRE _ READ _ INFO2 state according to a strict priority order when detecting that a physical dequeue information FIFO corresponding to a common queue is not empty;

PRE _ READ _ INFO2: pre-reading the state of the common queue physical dequeuing information, pre-reading the queue dequeuing information in the physical dequeuing information FIFO, and then switching to the JUDGE2 state;

judge2: judging the state of the physical dequeue of the common queue, judging whether the data frame can be directly sent to an MAC (media access control) core according to the frame length of the data frame, the output port number and the residual cache space of the MAC core corresponding to the output port, if the data frame can be sent to the MAC core, returning to an IDLE state for judging again, and otherwise, jumping to a WRITE _ FULL _ BD state;

WRITE _ FULL _ BD: in a common data transfer state, because the bit width of a bus adopted by the module is 128bits, data in one BD block is transferred for 4 times at most; if the current state is finished, the BD blocks still remain and the data in the next BD block still needs 4 times, the WRITE _ FULL _ BD still jumps back; after the current state is finished, if no residual data exists, jumping to a WRITE _ PART _ BD state; when the current state is finished once, the BD block address needs to be sent to the cache management module for cache release.

WRITE _ PART _ BD: in the ordinary data moving state, the last data less than 4 times of moving of the data frame is moved out, and the cache release is still needed after the moving is finished.

WRITE _ FINISH: the physical dequeue end state and the next state jumps to the IDLE state.

Fig. 12 is a schematic structural diagram of a management scheduling apparatus based on TSN queues according to an embodiment of the present invention;

as shown in fig. 12, the management scheduling apparatus based on the TSN queue includes an application information obtaining module, an enqueue determining module, an enqueue module, a dequeue determining module, and a dequeue module; wherein,

the application information acquisition module is used for acquiring application information of the data frame to be enqueued;

the enqueue judging module is used for judging whether the application information of the data frame to be enqueued meets enqueue conditions or not, if so, the enqueue judging module judges whether the application information of the data frame to be enqueued meets the enqueue conditions or not;

the enqueue module is used for enqueuing the data frames to be enqueued and updating queue information;

the dequeue judging module is used for judging whether to execute dequeue scheduling on the enqueue according to the queue information, and if so, determining whether to execute dequeue scheduling on the enqueue;

and the dequeuing module is used for executing dequeuing scheduling and dequeuing the enqueue in the queue according to a dequeuing scheduling result.

It should be noted that, although the present system only discloses the basic function modules such as the application information obtaining module, the enqueue judging module, the enqueue module, the dequeue judging module, and the dequeue module, the present invention is not limited to the above basic function modules, and on the basis of the above basic function modules, a person skilled in the art can arbitrarily add one or more function modules in combination with the prior art to form an infinite number of embodiments or technical solutions, that is, the present system is open rather than closed, and the protection scope of the present invention is considered to be limited to the above disclosed basic function modules because the present embodiment only discloses individual basic function modules.

Fig. 13 is a structural diagram of an electronic device that can implement the TSN queue-based management scheduling method of the present invention.

As shown in fig. 13, the electronic apparatus includes: the system comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory are communicated with each other through the communication bus; the memory has stored therein a computer program that, when executed by the processor, causes the processor to perform the steps of the TSN queue based management scheduling method.

The present application further provides a computer-readable storage medium storing a computer program executable by an electronic device, the computer program, when run on the electronic device, causing the electronic device to perform the steps of the TSN queue based management scheduling method.

The communication bus mentioned in the electronic device may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.

The electronic device includes a hardware layer, an operating system layer running on top of the hardware layer, and an application layer running on top of the operating system. The hardware layer includes hardware such as a Central Processing Unit (CPU), a Memory Management Unit (MMU), and a Memory. The operating system may be any one or more computer operating systems that implement control of the electronic device through a Process (Process), such as a Linux operating system, a Unix operating system, an Android operating system, an iOS operating system, or a Windows operating system. In the embodiment of the present invention, the electronic device may be a handheld device such as a smart phone and a tablet computer, or an electronic device such as a desktop computer and a portable computer, which is not particularly limited in the embodiment of the present invention.

The execution main body of the electronic device control in the embodiment of the present invention may be the electronic device, or a functional module capable of calling a program and executing the program in the electronic device. The electronic device may acquire the firmware corresponding to the storage medium, the firmware corresponding to the storage medium is provided by a vendor, and the firmware corresponding to different storage media may be the same or different, which is not limited herein. After the electronic device acquires the firmware corresponding to the storage medium, the firmware corresponding to the storage medium may be written into the storage medium, specifically, the firmware corresponding to the storage medium is burned into the storage medium. The process of burning the firmware into the storage medium can be realized by adopting the prior art, and details are not described in the embodiment of the present invention.

The electronic device may further acquire a reset command corresponding to the storage medium, where the reset command corresponding to the storage medium is provided by a vendor, and the reset commands corresponding to different storage media may be the same or different, and are not limited herein.

At this time, the storage medium of the electronic device is a storage medium in which the corresponding firmware is written, and the electronic device may respond to the reset command corresponding to the storage medium in which the corresponding firmware is written, so that the electronic device resets the storage medium in which the corresponding firmware is written according to the reset command corresponding to the storage medium. The process of resetting the storage medium according to the reset command can be implemented by the prior art, and is not described in detail in the embodiment of the present invention.

For convenience of description, the above devices are described as being functionally divided into various units and modules. Of course, the functions of the units and modules may be implemented in one or more software and/or hardware when implementing the present application.

It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

For simplicity of explanation, the method embodiments are described as a series of acts or combinations, but those skilled in the art will appreciate that the embodiments are not limited by the order of acts described, as some steps may occur in other orders or concurrently with other steps in accordance with the embodiments of the invention. Further, those of skill in the art will appreciate that the embodiments described in the specification are presently preferred and that no particular act is required to implement the embodiments of the invention.

From the above description of the embodiments, it is clear to those skilled in the art that the present application can be implemented by software plus necessary general hardware platform. Based on such understanding, the technical solutions of the present application may be essentially or partially implemented in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the embodiments or some parts of the embodiments.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and these modifications or substitutions do not depart from the spirit of the corresponding technical solutions of the embodiments of the present invention.

Claims (10)

1. A management scheduling method based on TSN queue is characterized by comprising the following steps:

acquiring application information of a data frame to be enqueued;

judging whether the application information of the data frame to be enqueued meets enqueuing conditions or not, if so, judging that the application information of the data frame to be enqueued meets the enqueuing conditions

Enqueuing the data frame to be enqueued and updating queue information;

judging whether to execute dequeue scheduling on the enqueue according to the queue information, if so, determining whether to execute dequeue scheduling on the enqueue

And performing dequeue scheduling, and dequeuing the enqueue in the queue according to a dequeue scheduling result.

2. The TSN queue-based management scheduling method of claim 1, wherein the determining whether the application information of the data frame to be enqueued satisfies enqueuing conditions includes:

acquiring an initial allocation queue number according to the application information of the data frame to be enqueued;

judging whether the initial distribution queue is a circular queue according to the initial distribution queue number, if so, judging whether the initial distribution queue is the circular queue

Acquiring a storage space and gating information of an initial distribution queue;

judging whether the storage space of the initial distribution queue can store the data frame to be enqueued and whether the gating information of the initial distribution queue is in an open state, if so, judging that the data frame to be enqueued can be stored in the storage space of the initial distribution queue and the gating information of the initial distribution queue is in the open state

Performing circular queue logic enqueuing on the data frame to be enqueued and updating logic enqueuing queue information;

and performing circular queue physical enqueuing on the data frames to be enqueued according to the logical enqueue information and updating the physical enqueue information.

3. The TSN queue-based management scheduling method of claim 2, wherein the determining whether the application information of the data frame to be enqueued satisfies enqueuing conditions includes:

judging whether the initial distribution queue is a common queue according to the initial distribution queue number, if so, judging whether the initial distribution queue is the common queue

Acquiring queue information of a common queue;

judging whether the common queue can store the data frame to be enqueued according to the queue information of the common queue and the application information of the data frame to be enqueued, if so, judging whether the common queue can store the data frame to be enqueued, and if so, storing the data frame to be enqueued in the common queue

Performing common queue logic enqueue on a data frame to be enqueued and updating logic enqueue information;

and performing common queue physical enqueue on the data frames to be enqueued according to the logical enqueue information and updating the physical enqueue information.

4. The TSN queue-based management scheduling method of claim 3, wherein the determining whether to perform dequeue scheduling on the enqueue according to the queue information comprises:

acquiring a non-empty queue according to the logic enqueue information;

acquiring queue information of the non-empty queue;

judging whether the non-empty queue can be dequeued or not according to the queue information of the non-empty queue, if so, determining that the non-empty queue can be dequeued

According to the queue information of the non-empty queue, carrying out dequeue priority ordering on the non-empty queue to obtain a logic dequeue queue;

logically dequeuing the non-empty queue according to the logical dequeue queue, and updating a physical dequeue queue;

and physically dequeuing the non-empty queue according to the physical dequeue queue.

5. The method for managing and scheduling based on the TSN queue according to claim 4, wherein the queue information of the non-empty queue includes a queue number of the non-empty queue, a queue number status, frame length information of a data frame of the non-empty queue, a buffer address of the data frame of the non-empty queue, a source port number of the data frame of the non-empty queue, an output port number of the non-empty queue, and a port status of an output port of the non-empty queue.

6. The TSN queue-based management scheduling method of claim 5, wherein said determining whether the non-empty queue can be dequeued according to the queue information of the non-empty queue comprises:

judging whether the port state of the non-empty queue output port is idle or not, if so, judging whether the port state of the non-empty queue output port is idle or not, and if not, judging whether the port state of the non-empty queue output port is idle or not

And carrying out dequeue priority sequencing on the non-empty queue according to the queue number state of the non-empty queue and the output port number of the non-empty queue to obtain a logic dequeue queue.

7. The method for managing and scheduling based on a TSN queue of claim 6, wherein said physically dequeuing said non-empty queue according to said physical dequeue queue comprises:

judging whether the non-empty queue is a circular queue according to the queue information of the non-empty queue, if so, judging whether the non-empty queue is the circular queue

Judging whether the port state of the non-empty queue output port is in an open state or not and a first receiving end corresponding to the non-empty queue output port can normally receive the non-empty queue, if so, judging that the port state of the non-empty queue output port is in the open state and a first receiving end corresponding to the non-empty queue output port can normally receive the non-empty queue

And physically dequeuing the non-empty queue according to the physical dequeue queue.

8. The TSN queue-based management scheduling method of claim 7, wherein the physically dequeuing the non-empty queue from the physical dequeue queue comprises:

judging whether the non-empty queue is a common queue according to the queue information of the non-empty queue, if so, judging whether the non-empty queue is the common queue

Judging whether a second receiving end corresponding to the non-empty queue output port can receive the non-empty queue according to the queue information of the non-empty queue, if so, judging that the second receiving end corresponding to the non-empty queue output port can receive the non-empty queue

And physically dequeuing the non-empty queue according to the physical dequeue queue.

9. The TSN queue-based management scheduling method of claim 8, wherein the TSN queue-based management scheduling method further comprises:

judging whether the number of the data frames to be enqueued meets the preset protection condition or not when the data frames to be enqueued are enqueued, if so, determining that the number of the data frames to be enqueued meets the preset protection condition

And updating the queue information when the data frame to be enqueued is enqueued.

10. A management scheduler apparatus based on TSN queue, comprising:

the application information acquisition module is used for acquiring application information of a data frame to be enqueued;

an enqueue judging module, configured to judge whether the application information of the data frame to be enqueued satisfies an enqueue condition, and if so, determine whether the application information of the data frame to be enqueued satisfies the enqueue condition

The enqueuing module is used for enqueuing the data frames to be enqueued and updating queue information;

a dequeue judging module, configured to judge whether to perform dequeue scheduling on the enqueue according to the queue information, and if so, determine whether to perform dequeue scheduling on the enqueue

And the dequeuing module is used for executing dequeuing scheduling and dequeuing the enqueue in the queue according to a dequeuing scheduling result.

Images