WO2021259321A1

WO2021259321A1 - Storage scheduling method, device, and storage medium

Info

Publication number: WO2021259321A1
Application number: PCT/CN2021/101809
Authority: WO
Inventors: 付行双; 陈昌胜; 仲建锋; 胡达
Original assignee: 中兴通讯股份有限公司
Priority date: 2020-06-23
Filing date: 2021-06-23
Publication date: 2021-12-30
Also published as: CN113835611A

Abstract

A storage scheduling method, a device, and a storage medium, relating to the technical field of communications. The method comprises: determining the sum of the length of a received target packet and the current on-chip queue depth of a target queue corresponding to the target packet as an update depth of the target queue; when the update depth of the target queue is greater than a migration threshold of the target queue, determining that the target packet needs to be migrated to an off-chip storage device for storage; if the current off-chip storage device bandwidth is sufficient and the off-chip storage device capacity is sufficient, migrating the target packet out of a chip and storing the target packet in the off-chip storage device, and the on-chip queue depth of the target queue remaining unchanged; if the current off-chip storage device bandwidth is not sufficient or the off-chip storage device capacity is not sufficient and the current on-chip storage occupancy value is less than or equal to a preset storage occupancy threshold, storing the target packet on an on-chip storage device, and determining the current update depth of the target queue as the new current on-chip queue depth of the target queue; and if the current on-chip storage occupancy value is greater than the preset storage occupancy threshold, migrating and discarding the target packet, and the on-chip queue depth of the target queue remaining unchanged. The storage scheduling method improves the utilization rate of an on-chip storage space, thereby improving the overall performance of a chip.

Description

Storage scheduling method, equipment and storage medium

Cross-references to related applications

This disclosure requires the priority of a Chinese patent application filed with the State Intellectual Property Office, the application number is 202010582287.9, and the invention title is "Storage Scheduling Method, Device, and Storage Medium" on June 23, 2020. The entire content of the application is incorporated by reference. In this disclosure.

Technical field

The embodiments of the present disclosure relate to the field of communication technologies, and in particular, to a storage scheduling method, device, and storage medium.

Background technique

In the network processing engine, on the one hand, it is necessary to continuously improve the performance index, and on the other hand, it is limited by the chip area, resources, and technology. It is impossible to increase the main frequency without limitation or to meet the performance requirements by piling up resources. Therefore, in system design, the performance index is often achieved by increasing the number of queues and correspondingly increasing the off-chip storage space. The messages in each queue can be stored in on-chip storage space or off-chip storage devices. When a new message is received, the storage location of the message needs to be scheduled.

Due to the limited bandwidth and capacity of off-chip storage devices, additional power consumption, resources, and circuit area will be added. Therefore, it is impossible to increase off-chip storage devices indefinitely to meet performance requirements. Therefore, the storage scheduling method for scheduling the storage location of the message, in addition to meeting the function of moving the message from the on-chip to the off-chip storage device, also needs to improve the utilization of on-chip resources, minimize the discarding of messages, and improve the chip Overall performance.

Summary of the invention

The main purpose of the embodiments of the present disclosure is to propose a storage scheduling method, device, and storage medium, aiming to realize the function of improving the utilization of on-chip storage space through storage scheduling, thereby improving the overall performance of the chip.

In order to achieve the above objective, according to the first aspect of the present disclosure, the embodiments of the present disclosure provide a storage scheduling method. The method includes the following steps: combining the length of the received target message and the target corresponding to the target message. The sum of the current on-chip queue depths of the queue is determined as the update depth of the target queue; when the update depth of the target queue is greater than the movement threshold of the target queue, it is determined that the target message needs to be moved to an off-chip storage device Storage; if the bandwidth of the current off-chip storage device is sufficient and the capacity of the off-chip storage device is sufficient, the target message is moved to the off-chip and stored on the off-chip storage device, and the on-chip queue depth of the target queue remains unchanged; if the current chip The bandwidth of the external storage device is insufficient or the capacity of the off-chip storage device is insufficient, and the current on-chip storage occupancy value is less than or equal to the preset storage occupancy threshold, the target message is stored in the on-chip storage device, and the current update depth of the target queue , That is, the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message is determined as the new current on-chip queue depth of the target queue; if the current on-chip storage occupancy value If it is greater than the preset storage occupancy threshold, the target message is moved and discarded, and the on-chip queue depth of the target queue remains unchanged.

In order to achieve the above objective, according to the second aspect of the present disclosure, the embodiments of the present disclosure also propose a storage scheduling device, the device includes a memory, a processor, is stored on the memory and can run on the processor The program and the data bus used to realize the connection and communication between the processor and the memory, and when the program is executed by the processor, the steps of the foregoing method are implemented.

In order to achieve the foregoing objective, according to the third aspect of the present disclosure, embodiments of the present disclosure provide a storage medium for computer-readable storage, the storage medium stores one or more programs, and the one or more programs It may be executed by one or more processors to implement the steps of the foregoing method.

The storage scheduling method, device and storage medium proposed in this embodiment include: determining the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message as the update depth of the target queue. When the update depth of the target queue is greater than the moving threshold of the target queue, it is determined that the target message needs to be moved to an off-chip storage device for storage. If the bandwidth of the current off-chip storage device is sufficient and the capacity of the off-chip storage device is sufficient, the target message is moved to the off-chip and stored on the off-chip storage device, and the on-chip queue depth of the target queue remains unchanged; if the current off-chip storage is The device bandwidth is insufficient or the off-chip storage device capacity is insufficient, and the current on-chip storage occupancy value is less than or equal to the preset storage occupancy threshold, the target message is stored in the on-chip storage device, and the current update depth of the target queue is , The sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message is determined as the new current on-chip queue depth of the target queue; if the current on-chip storage occupancy value is greater than the preset If the storage occupancy threshold is set, the target message is moved and discarded, and the on-chip queue depth of the target queue remains unchanged. In the storage scheduling method, when the update depth of the target queue is greater than the moving threshold of the target queue, on the one hand, when the off-chip storage device cannot successfully store the target message and the current on-chip storage occupancy value is less than the preset storage occupancy threshold, the target queue The message is stored in the on-chip space again, which improves the utilization of the on-chip storage space, thereby improving the overall performance of the chip; on the other hand, the off-chip storage device cannot successfully store the target message and the current on-chip storage occupancy value is greater than the preset Only when the storage occupancy threshold is triggered, the move and discard of the target message is triggered to discard the message, so as to reduce the probability of discarding the message as much as possible, and to prevent the storage scheduling method from affecting the normal operation of the business.

Description of the drawings

FIG. 1 is a flowchart of a storage scheduling method provided by an embodiment;

FIG. 2 is a flowchart of a storage scheduling method provided by another embodiment;

FIG. 3 is a schematic structural diagram of a storage scheduling device provided by an embodiment;

4 is a schematic structural diagram of a storage scheduling device provided by another embodiment;

FIG. 5 is a schematic structural diagram of a storage scheduling device provided by another embodiment;

Fig. 6 is a schematic structural diagram of a storage scheduling device provided by an embodiment.

detailed description

It should be understood that the specific embodiments described herein are only used to explain the embodiments of the present disclosure, and are not used to limit the embodiments of the present disclosure.

In the following description, the use of suffixes such as “module”, “component” or “unit” used to indicate elements is only for facilitating the description of the embodiments of the present disclosure, and has no special meaning in itself. Therefore, "module", "part" or "unit" can be used in a mixed manner.

The embodiment of the present disclosure provides a storage scheduling method, including: determining the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message as the update depth of the target queue. When the update depth is greater than the moving threshold of the target queue, it is determined that the target message needs to be moved to an off-chip storage device for storage. If the bandwidth of the current off-chip storage device is sufficient and the capacity of the off-chip storage device is sufficient, the target message is moved off-chip and stored on the off-chip storage device, and the on-chip queue depth of the target queue remains unchanged; if the current off-chip The bandwidth of the storage device is insufficient or the capacity of the off-chip storage device is insufficient, and the current on-chip storage occupancy value is less than or equal to the preset storage occupancy threshold, then the target message is stored in the on-chip storage device, and the current update depth of the target queue , That is, the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message is determined as the new current on-chip queue depth of the target queue; if the current on-chip storage occupancy value If it is greater than the preset storage occupancy threshold, the target message is moved and discarded, and the on-chip queue depth of the target queue remains unchanged. In the storage scheduling method, when the update depth of the target queue is greater than the moving threshold of the target queue, on the one hand, when the off-chip storage device cannot successfully store the target message and the current on-chip storage occupancy value is less than the preset storage occupancy threshold, the target queue The message is stored in the on-chip space again, which improves the utilization of the on-chip storage space, thereby improving the overall performance of the chip; on the other hand, the off-chip storage device cannot successfully store the target message and the current on-chip storage occupancy value is greater than the preset Only when the storage occupancy threshold is triggered, the move and discard of the target message is triggered, and the message is discarded, so as to reduce the probability of discarding the message as much as possible, and to prevent the storage scheduling method from affecting the normal operation of the business.

FIG. 1 is a flowchart of a storage scheduling method provided by an embodiment of the disclosure. This embodiment is applicable to a scenario where the storage location of a message is scheduled. This embodiment may be executed by a storage scheduling apparatus, which may be implemented by software and/or hardware, and the storage scheduling apparatus may be integrated into a communication device. As shown in Figure 1, the storage scheduling method provided in this embodiment includes the following steps:

Step 101: The sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message is determined as the update depth of the target queue.

The queues in this embodiment can be divided according to users, that is, messages of different users belong to different queues, or they can be divided according to service types, that is, messages of different services belong to different queues, or queues can be divided according to other factors. , This implementation does not impose restrictions on this.

After receiving the target message, the storage scheduling apparatus needs to determine whether the target message is stored in the on-chip storage space or in the off-chip storage device. The on-chip storage space in this embodiment refers to the storage space on the chip in the storage scheduling device. The on-chip storage space in this embodiment may be cache space, random access memory (Random Access Memory, RAM), etc. The off-chip storage device in this embodiment may be a storage device provided outside the chip. The chip in this embodiment may be a chip in a network processing engine.

In an embodiment, the target message may carry the queue number. After receiving the target message, the queue corresponding to the target message is determined according to the queue number of the target message. For the convenience of description, the queue corresponding to the target message is called the target queue. The current on-chip queue depth of the target queue refers to the length of the message stored in the on-chip storage space in the target queue. The current on-chip queue depth of the target queue can be obtained from the module that manages the on-chip queue depth in the storage scheduling device after receiving the target message, or it can be obtained according to the current on-chip queue depth after receiving the target message. The length of the message stored in the on-chip storage space is directly calculated.

Optionally, the length of the message in this embodiment may refer to the length under different metrics such as the number of bytes of the message and the number of slices of the message.

The sum of the length of the received target message and the current on-chip queue depth of the target queue is determined as the update depth of the target queue. For example, assuming that the length of the target message is c0 and the current on-chip queue depth of the target queue is b0, then c0+b0 is determined as the update depth of the target queue.

Step 102: When the update depth of the target queue is greater than the moving threshold of the target queue, it is determined that the target message needs to be moved to an off-chip storage device for storage.

In an embodiment, each target queue corresponds to a moving threshold, and the moving threshold may be preset. The moving threshold of the target queue can be set according to the needs of the user, or it can be sent to the storage scheduling device by other devices.

In one implementation, when the update depth of the target queue is greater than the moving threshold of the target queue, it means that the target queue is congested at this time, that is, in the scenario where the message enters the queue fast and the dequeue is slow, it is determined that the target message needs Move to an off-chip storage device for storage.

In another implementation, when the update depth of the target queue is less than or equal to the moving threshold of the target queue, the target message is stored in the on-chip storage space; the current on-chip queue depth of the target queue is added to the length of the target message And, determine the new current on-chip queue depth of the target queue. In this implementation, when the target queue is not congested, that is, in a scenario where messages are slow to enter the queue and quickly dequeue, after storing the target message in the on-chip storage space, update the current on-chip queue depth of the target queue. Improve the accuracy of the current on-chip queue depth of the target queue, thereby improving the accuracy of subsequent storage scheduling and further improving the performance of the chip.

It should be noted that the way the messages in the queue are stored in the on-chip storage space can be shared storage, that is, multiple queues share on-chip storage space, or exclusive storage, that is, each queue has its own corresponding The on-chip storage space can also be hybrid storage, that is, multiple queues share a part of the on-chip storage space, and each queue also has its own corresponding on-chip storage space. This embodiment does not limit the storage mode of the queue message in the on-chip storage space.

Step 103: Determine whether the bandwidth of the current off-chip storage device is sufficient, and whether the cache capacity of the off-chip storage device is sufficient.

In one embodiment, before the target message is moved, the status information sent by the off-chip storage device needs to be received to determine whether the off-chip storage device can store the target message.

Optionally, the off-chip storage device may send the status information to the storage scheduling apparatus at a preset frequency. The status information is used to indicate whether the off-chip storage device has the ability to store target messages. Exemplarily, the status information may be information such as the current bandwidth and/or current remaining storage space of the off-chip storage device.

Step 104: If the bandwidth of the current off-chip storage device is sufficient and the capacity of the off-chip storage device is sufficient, move the target message to the off-chip and store in the off-chip storage device, and perform step 105;

In one embodiment, when the bandwidth of the off-chip storage device meets the bandwidth requirement of the target message, and the storage space of the off-chip storage device is sufficient, the storage scheduling apparatus determines that the target message needs to be moved to the off-chip storage device for storage. , The target message will be moved to the off-chip storage device for storage.

Step 105: The on-chip queue depth of the target queue remains unchanged.

In one embodiment, when the target message is moved off-chip and stored in an off-chip off-chip storage device, the on-chip queue depth of the target queue corresponding to the target message does not change, and the current on-chip storage occupancy value also does not change.

The current on-chip storage occupancy value refers to the sum of the current on-chip queue depths of all queues.

In an embodiment, the current on-chip storage occupancy value may be sent to the storage scheduling device by other modules.

Step 106: If the bandwidth of the current off-chip storage device is insufficient, or the storage capacity of the off-chip storage device is sufficient, step 107 is executed.

In an embodiment, according to the status information sent by the off-chip storage device, the current off-chip storage device does not have the ability to store the target message.

It is certain that when the storage capacity of the off-chip storage device is insufficient, the target message will not be moved, and will not be moved to the off-chip storage device.

Step 107: Whether the current on-chip storage occupancy value is greater than the on-chip preset storage occupancy threshold.

In one embodiment, the target message that needs to be moved off-chip, when the off-chip storage device does not have the ability to store the target message, it is determined according to the comparison result of the current on-chip storage occupancy value and the on-chip preset storage occupancy threshold Whether the current on-chip storage device has the ability to store target messages.

The storage scheduling device needs to determine the current on-chip storage occupancy value: the sum of the current on-chip queue depths of all queues is determined as the current on-chip storage occupancy value. In this implementation manner, the sum of the current on-chip queue depths of all queues is determined as the current on-chip storage occupancy value, which can improve the accuracy of the current on-chip storage occupancy value as much as possible.

Step 108: When the current on-chip storage occupancy value is greater than the on-chip preset storage occupancy threshold, the target message that needs to be moved off-chip is stored on the chip again.

The current on-chip storage occupancy value is less than the preset storage occupancy threshold, indicating that there is excess storage space in the on-chip storage device at this time, and the current on-chip storage device has the ability to store the target message.

The off-chip storage device cannot successfully store the target message. However, when the current on-chip storage device has the ability to store the target message, the message in the target queue is re-stored in the on-chip space, which improves the utilization of the on-chip storage space. , Improve the overall performance of the chip.

Step 109: Determine the current update depth of the target queue, that is, the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message as the new current on-chip queue depth of the target queue.

In one embodiment, the current on-chip queue depth of the target queue needs to be updated. The specific update method is: the sum of the current on-chip queue depth of the target queue plus the length of the target message is determined as the new current on-chip queue depth of the target queue.

Step 110: When the previous on-chip storage occupancy value is less than or equal to the on-chip preset storage occupancy threshold, the target message is moved and discarded.

When it is determined that the off-chip storage device cannot store the target message, and the on-chip storage device cannot store the target message, the target message is marked as being moved and discarded, and the storage scheduling device moves and discards the target message.

When the off-chip storage device cannot successfully store the target message, and the current on-chip storage device cannot store the target message, the move and discard of the target message is triggered to discard the message, so as to reduce the number of discarded messages as much as possible. The probability of avoiding the storage scheduling method from affecting the normal operation of the business.

The foregoing method of determining whether a message is moved or discarded is simple to implement and highly efficient.

Step 111: The on-chip queue depth of the target queue remains unchanged.

Before executing the storage scheduling method provided in this embodiment, information configuration is also required in advance. The information that needs to be configured may include: the number of queues, the movement threshold corresponding to each queue, and the preset storage occupancy threshold, etc.

The embodiment of the present disclosure provides a storage scheduling method, including: determining the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message as the update depth of the target queue. When the update depth is greater than the move threshold of the target queue, it is determined that the target message needs to be moved to the off-chip storage device for storage. If the current off-chip storage device has sufficient bandwidth and the off-chip storage device capacity is sufficient, move the target message off-chip and store it in the off-chip. For off-chip storage devices, the on-chip queue depth of the target queue remains unchanged; if the bandwidth of the current off-chip storage device is insufficient or the capacity of the off-chip storage device is insufficient, and the current on-chip storage occupancy value is less than or equal to the preset storage occupancy threshold, the target The message is stored in the on-chip storage device, and the current update depth of the target queue, that is, the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message, is determined as the new target queue The current on-chip queue depth; if the current on-chip storage occupancy value is greater than the preset storage occupancy threshold, the target message is moved and discarded, and the on-chip queue depth of the target queue remains unchanged. In the storage scheduling method, when the update depth of the target queue is greater than the moving threshold of the target queue, on the one hand, when the off-chip storage device cannot successfully store the target message and the current on-chip storage occupancy value is less than the preset storage occupancy threshold, the target queue The message is stored in the on-chip space again, which improves the utilization of the on-chip storage space, thereby improving the overall performance of the chip; on the other hand, the off-chip storage device cannot successfully store the target message and the current on-chip storage occupancy value is greater than the preset Only when the storage occupancy threshold is triggered, the move and discard of the target message is triggered to discard the message, so as to reduce the probability of discarding the message as much as possible, and to prevent the storage scheduling method from affecting the normal operation of the business.

FIG. 2 is a flowchart of a storage scheduling method provided by another embodiment of the present disclosure. This embodiment provides a detailed description of other steps included in the storage scheduling method on the basis of the embodiment shown in FIG. 1 and various alternative solutions. As shown in Figure 2, the storage scheduling method provided in this embodiment includes the following steps:

Step 201: Determine the attributes of the target message.

Step 202: When it is judged that the attribute of the target message is an on-chip message, the target message is stored in the on-chip storage.

Step 203: When it is determined that the attribute of the target message is an off-chip message, the target message is stored in an off-chip storage device.

Step 204: When it is determined that the attribute of the target message is a mixed type, it is determined to execute step 205.

That is, when it is judged that the attribute of the target message is a mixed type, the step of determining the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message is determined to be the update depth of the target queue. .

In an embodiment, the target message will carry its own attribute information. The attributes of the message in this embodiment may include: mixed-type messages, on-chip messages, and off-chip messages. Among them, mixed type messages refer to messages that can be stored in off-chip storage devices or in on-chip storage space. On-chip messages refer to messages that can only be stored in on-chip storage space. Messages refer to messages that can only be stored in off-chip storage devices.

In this embodiment, after receiving the target message, the attributes of the target message are first judged. Select subsequent operations based on the attributes of the target message. The first case: when it is determined that the target message is an on-chip message, no storage scheduling is required, and the target message is stored in the on-chip storage space. The second case: when it is determined that the attributes of the target message are of mixed types, it is necessary to determine whether the target message needs to be stored on-chip or off-chip at this time, that is, storage scheduling needs to be performed, and step 205 is determined to be executed. The third case: when it is determined that the target message is an off-chip message, there is no need to perform storage scheduling, and the target message is stored in an off-chip storage device.

This first determines the attributes of the target message, and then determines the processing method for subsequent operations. Storage scheduling can be performed when storage scheduling is needed, avoiding the business processing efficiency caused by blind scheduling to store on-chip messages to off-chip storage devices Reduce, or, store off-chip messages to the on-chip storage space to occupy the on-chip storage space.

It should be noted that the messages in the same queue in this embodiment may be messages of the same attribute or messages of different attributes. This embodiment is not limited to this.

Step 205: The sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message is determined as the update depth of the target queue.

The implementation process and technical principles of step 205 are similar to that of step 101, and will not be repeated here.

Step 206: Determine whether the update depth of the target queue is greater than the moving threshold of the target queue.

Step 207: When the update depth of the target queue is less than or equal to the moving threshold of the target queue, store the target message in the on-chip storage space.

In step 207, the update depth of the target queue is less than or equal to the moving threshold of the target queue, indicating that the target message can be stored in the on-chip storage space. Therefore, the target message is stored in the on-chip storage space.

Step 208: The sum of the current on-chip queue depth of the target queue plus the length of the target message is determined as the new current on-chip queue depth of the target queue.

In step 208, after storing the target message in the on-chip storage space, the current on-chip queue depth of the target queue needs to be updated to improve the accuracy of subsequent storage scheduling. The update method is to determine the new current on-chip queue depth of the target queue by adding the sum of the current on-chip queue depth of the target queue to the length of the target message.

Step 209: When the update depth of the target queue is greater than the moving threshold of the target queue, it is determined that the target message needs to be moved to an off-chip storage device for storage.

The implementation process and technical principles of step 209 and step 102 are similar, and will not be repeated here.

Step 210: Determine whether the bandwidth of the current off-chip storage device is sufficient, and whether the cache capacity of the off-chip storage device is sufficient.

Before the target message is moved, the status information sent by the off-chip storage device needs to be received to determine whether the off-chip storage device can store the target message.

Step 211: If the bandwidth of the current off-chip storage device is sufficient and the capacity of the off-chip storage device is sufficient, move the target message to the off-chip and store in the off-chip storage device.

In an embodiment, when the bandwidth of the off-chip storage device meets the bandwidth requirement of the target message, and the storage space of the off-chip storage device is sufficient, that is, when the off-chip storage device has sufficient capacity to store the target message, After determining that the target message needs to be moved to the off-chip storage device for storage, the storage scheduling device moves the target message to the off-chip storage device for storage.

Step 212: Move the target message to the off-chip and store it in the off-chip storage device.

Step 213: The on-chip queue depth of the target queue remains unchanged.

After the storage scheduling device successfully moves the target message to the off-chip storage device for storage, it does not need to update the current on-chip queue depth of the target queue. Since the current on-chip queue depth of the target queue has not changed in this scenario, the current on-chip queue depth of the target queue is determined as the new current on-chip queue depth of the target queue.

The implementation process and technical principles of step 211 and step 105 are similar, and will not be repeated here.

Step 214: If the bandwidth of the current off-chip storage device is insufficient, or the storage capacity of the off-chip storage device is sufficient, step 215 is executed.

The implementation process and technical principles of step 214 and step 106 are similar, and will not be repeated here.

Step 215: Whether the current on-chip storage occupancy value is greater than the on-chip preset storage occupancy threshold.

The implementation process and technical principles of step 215 are similar to step 107, and will not be repeated here.

Step 216: When the current on-chip storage occupancy value is greater than the on-chip preset storage occupancy threshold, the target message that needs to be moved off-chip is stored on the chip again.

Step 217: Determine the current update depth of the target queue, that is, the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message as the new current on-chip queue depth of the target queue.

The implementation process and technical principles of step 217 are similar to step 109, and will not be repeated here.

Step 218: When the previous on-chip storage occupancy value is less than or equal to the on-chip preset storage occupancy threshold, the target message is moved and discarded.

Step 219: The on-chip queue depth of the target queue remains unchanged.

After the storage scheduling device moves and discards the target message, it does not need to update the current on-chip queue depth of the target queue. Since the current on-chip queue depth of the target queue has not changed in this scenario, the current on-chip queue depth of the target queue is determined as the new current on-chip queue depth of the target queue.

The solution of this embodiment will be described in detail below in conjunction with a specific implementation of the storage scheduling device. FIG. 3 is a schematic structural diagram of a storage scheduling device provided by an embodiment. As shown in FIG. 3, the storage scheduling device provided in this embodiment includes the following modules: a threshold management module 31, a movement state counting module 32, and a movement arbitration module 33. Among them, the threshold management module 31 and the movement state counting module 32 are both connected to the movement arbitration module 33.

The threshold management module 31 can implement the movement threshold of the query queue and the query on-chip storage occupancy threshold. Assuming that the number of queues is two, numbered 0 and 1, then the threshold management module 31 can query the corresponding queue movement threshold according to the queue number, and store it in the shared temporary storage register for the movement arbitration module 33 to use, or according to the queue number Query the corresponding storage occupancy threshold. Assume that the shift domain value corresponding to queue 0 is a0, and the corresponding storage occupancy threshold value is s0. The moving threshold corresponding to queue 1 is a1, and the corresponding storage occupancy threshold is s1.

The moving state counting module 32 can determine the current on-chip queue depth of the queue, determine the update depth of the queue, and determine the new current on-chip queue depth of the queue. Further, the moving state counting module 32 can also determine the message attributes of the target message. Fig. 3 includes N+1 moving state counting modules 32. Each moving state counting module 32 can manage the current on-chip queue depth of queues in different dimensions. For example, the movement state counting module 0 manages the current on-chip queue depth of Q+1 queues when the queue is divided from dimension A, and the movement state counting module N manages the current on-chip queue depth of Q+1 queues when the queue is divided from dimension M. Suppose the current on-chip queue depth of queue 0 is b0, and the current chip queue depth of queue 1 is b1. Here, queue 0 and queue 1 can be queues in any dimension.

The movement arbitration module 33 includes an on-chip resource judgment sub-module 331 and a scheduler 332. Among them, the number of schedulers 332 may be multiple. Optionally, the configuration value of scheduler 0 is 2'b00, which means that scheduler 0 performs system-level scheduling. The scheduler 0 is used to calculate the on-chip storage space consumption status in real time, that is, to calculate the current on-chip storage occupancy value. Other schedulers are used to schedule the queue. Each queue can correspond to a scheduler. Illustratively, queue 0 corresponds to scheduler 1, and queue 1 corresponds to scheduler 2. The configuration value of scheduler 1 and scheduler 2 is 2'b01, which indicates that the scheduling strategy is flow queue level scheduling.

The length of the message received by queue 0 is c0, the on-chip storage occupancy value at this time is c0', and the current on-chip queue depth of queue 0 is b0. Determine b0+c0 as the update depth of queue 0. If scheduler 1 determines that b0+c0>a0, then the currently received message corresponding to queue 0 needs to be moved to the off-chip storage device; if scheduler 1 determines that b0+c0≤a0, then the current received message corresponding to queue 0 The received message is stored in the on-chip storage space.

Similarly, the length of the message received by queue 1 is c1, the on-chip buffer occupancy value at this time is c0'+c1', and the current on-chip queue depth of queue 1 is b1. Determine b1+c1 as the update depth of queue 1. If scheduler 2 determines that b1+c1>a1, then the currently received message corresponding to queue 1 needs to be moved to the off-chip storage device; if scheduler 2 determines that b1+c1≤a1, then the currently received message corresponding to queue 1 The received message is stored in the on-chip storage space.

If the currently received message corresponding to queue 1 does not need to be moved to an off-chip storage device, the message will be stored in the on-chip space. If the currently received message corresponding to queue 1 needs to be moved to an off-chip storage device, the status information fed back by the off-chip storage device shows that the current off-chip storage device does not have enough storage capacity, and the message will not be moved. Move to an off-chip storage device. If the on-chip resource judging submodule 331 determines that the on-chip space is sufficient and there is enough space to store the currently received message, that is, c0'+c1'≤s1, the message that needs to be moved will be stored in the on-chip space again; The internal resource judgment submodule 331 determines that the on-chip space is insufficient to store the currently received message, that is, c0'+c1'>s1, it will trigger the transfer and discard of the message, and discard the message that needs to be moved.

As shown in Figure 3, if any one of the multiple schedulers determines that it needs to be moved, it needs to be moved. Therefore, the multiple schedulers are connected by a "logical OR". If any scheduler determines that the message needs to be moved to an off-chip storage device, and (that is, the "logical AND" on the right in Figure 3), the off-chip storage device has insufficient storage capacity and sufficient on-chip space, the message will be Re-store in the on-chip space; if any scheduler determines that the message needs to be moved to an off-chip storage device, and (ie the "logical AND" on the right in Figure 3), the off-chip storage device has insufficient storage capacity and on-chip space When it is insufficient, it will trigger the transfer and discard of the message, and the message will be discarded.

The moving state counting module 32 can update the current on-chip queue depth of the queue according to the storage scheduling result. Therefore, the moving state counting module can be connected with the storage scheduling result, that is, the on-chip/off-chip moving result, and determine whether the on-chip queue depth needs to be updated according to the moving result.

When the messages currently received in queue 1 need to be moved to an off-chip storage device, if the off-chip storage device does not have enough storage capacity, there is enough space on the chip to store the currently received messages, and the messages that need to be moved will be stored again In the on-chip space, the free on-chip storage space is reused to improve the utilization of the on-chip storage space, thereby improving the overall performance of the chip. When the message currently received in queue 1 needs to be moved to an off-chip storage device, if the off-chip storage device does not have enough storage capacity and there is or does not have enough space on the chip to store the currently received message, the message will be triggered. Moving and discarding, discarding the message, reducing the probability of discarding the message as much as possible, and avoiding the storage scheduling method from affecting the normal operation of the business.

Based on the storage scheduling method provided by this embodiment, there can be a very high degree of freedom when designing input multi-queues. The number of queues and scheduling strategies can be parameterized and scaled. It can adapt to various system structures and data flow scheduling requirements. Greatly reduce the design difficulty of the system and improve the reliability of the structure. At the same time, it also helps with the portability and inheritability of the structure.

In the storage scheduling method provided in this embodiment, on the one hand, before performing storage scheduling, first determine the attributes of the target message, and then determine subsequent operations. Storage scheduling can be performed when storage scheduling is needed, avoiding blind scheduling and reporting the chip. The service processing efficiency is reduced due to the storage of files to off-chip storage devices, or the storage of off-chip messages in the on-chip storage space occupies the on-chip storage space; on the other hand, this embodiment may use different storage scheduling results based on different storage scheduling results. The calculation process updates the current on-chip queue depth of the target queue to provide an accurate basis for the subsequent storage scheduling process, improve the accuracy of storage scheduling, and further improve the performance of the chip.

FIG. 4 is a schematic structural diagram of a storage scheduling device provided by another embodiment of the present disclosure. As shown in FIG. 4, the storage scheduling apparatus provided in this embodiment includes the following modules: a first determining module 41, a second determining module 42, a third determining module 43, and a triggering module 44.

The first determining module 41 is configured to determine the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message as the update depth of the target queue.

The second determining module 42 is configured to determine that the target message needs to be moved to an off-chip storage device for storage when the update depth of the target queue is greater than the moving threshold of the target queue.

The third determining module 43 is configured to determine that if the target message needs to be moved to an off-chip storage device, the off-chip storage device does not have sufficient storage capacity, and when there is enough space on the chip, it is determined that the target message will be stored in the on-chip space again.

The trigger module 44 is configured to trigger the removal and discarding of the target message in the target queue if the target message needs to be moved to an off-chip storage device, and the off-chip storage device does not have sufficient storage capacity, and the on-chip space is insufficient.

Optionally, the device further includes: a receiving module, a first storage module, and a discarding module.

The receiving module is configured to receive the status information sent by the off-chip storage device.

The first storage module is configured to move the target message to the off-chip storage device for storage if it is determined according to the status information that the off-chip storage device can store the target message.

The discarding module is configured to discard the target message if it is determined according to the status information that the off-chip storage device cannot store the target message, and there is no space in the on-chip storage to store the target message.

In an embodiment, the device further includes: a fourth determining module 45 configured to determine the current on-chip queue depth of the target queue minus the length of the dequeue message as the new current on-chip queue depth of the target queue .

The fourth determining module 45 is also configured to re-store the target message that needs to be moved to the on-chip space, the current on-chip queue depth of the target queue plus the length of the target message is determined as the new current on-chip of the target queue Queue depth.

In an embodiment, the device further includes: a fifth determining module configured to determine the sum of the current on-chip queue depths of all queues as the current on-chip storage occupancy value.

In one embodiment, in terms of triggering the message transfer and discarding of the target queue, the trigger module 44 is specifically configured to: add a discard tag to the target queue; in queue dequeue scheduling, if it is determined that the target queue has a discard tag, trigger The packets in the target queue are dequeued, and the dequeued packets are discarded.

The storage scheduling device provided in this embodiment is used to execute the storage scheduling method of any of the foregoing embodiments. The implementation principles and technical effects of the storage scheduling device provided in this embodiment are similar, and will not be repeated here.

FIG. 5 is a schematic structural diagram of a storage scheduling device provided by another embodiment of the present disclosure. This embodiment provides a detailed description of other modules included in the storage scheduling device on the basis of the embodiment shown in FIG. 4 and various alternative solutions. As shown in FIG. 5, the storage scheduling device provided in this embodiment further includes the following modules: a second storage module 51, a judgment module 52, a sixth determination module 53, a third storage module 54 and a fourth storage module 55.

The second storage module 51 is configured to store the target message in the on-chip storage space when the update depth of the target queue is less than or equal to the movement threshold of the target queue.

The fourth determining module 45 is also configured to determine the current on-chip queue depth of the target queue plus the length of the target message as the new current on-chip queue depth of the target queue.

The fourth determining module 45 is further configured to determine the current on-chip queue depth of the target queue as the new current on-chip queue depth of the target queue if the target message is successfully moved to the off-chip storage device for storage.

Optionally, the preset storage occupancy threshold is a storage occupancy threshold corresponding to the target queue. Correspondingly, the device further includes: a fifth determining module, configured to determine a storage occupancy threshold corresponding to the priority of the target queue according to the priority of the target queue when the target message is moved to the off-chip storage device .

The judging module 52 is configured to judge the attributes of the target message.

The sixth determining module 53 is configured to determine the sum of the length of the target message to be received and the current on-chip queue depth of the target queue corresponding to the target message when it is determined that the attribute of the target message is a mixed type, and it is determined as Steps to update the depth of the target queue.

In an embodiment, the third storage module 54 is configured to store the target message in the on-chip storage when it is determined that the attribute of the target message is an on-chip message.

The fourth storage module 55 is configured to store the target message in the off-chip storage device when it is determined that the attribute of the target message is an off-chip message.

FIG. 6 is a schematic structural diagram of a storage scheduling device provided by an embodiment of the disclosure. As shown in Figure 6, the storage scheduling device includes a processor 61 and a memory 62; the number of processors 61 in the storage scheduling device can be one or more. In Figure 6, one processor 61 is taken as an example; The processor 61 and the memory 62; can be connected by a bus or in other ways. In FIG. 6, the connection by a bus is taken as an example.

As a computer-readable storage medium, the memory 62 can be used to store software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the storage scheduling method in the embodiment of the present application (for example, the first storage scheduling device in the storage scheduling device). The determining module 41, the second determining module 42, the third determining module 43, the triggering module 44, and the fourth determining module 45). The processor 61 stores various functional applications and data processing of the scheduling device by running the software programs, instructions, and modules stored in the memory 62, that is, realizes the above-mentioned storage scheduling method.

The memory 62 may mainly include a program storage area and a data storage area. The program storage area may store an operating system and an application program required by at least one function; the data storage area may store data created according to the use of the storage scheduling device and the like. In addition, the memory 62 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices.

An embodiment of the present application also provides a storage medium containing computer-executable instructions. The computer-executable instructions are used to execute a storage scheduling method when executed by a computer processor, and the method includes:

Determining the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message as the update depth of the target queue;

When the update depth of the target queue is greater than the moving threshold of the target queue, determining that the target message needs to be moved to an off-chip storage device for storage;

If the bandwidth of the current off-chip storage device is sufficient and the capacity of the off-chip storage device is sufficient, the target message is moved off-chip and stored in the off-chip storage device, and the on-chip queue depth of the target queue remains unchanged;

If the bandwidth of the current off-chip storage device is insufficient or the capacity of the off-chip storage device is insufficient, and the current on-chip storage occupancy value is less than or equal to the preset storage occupancy threshold, the target message is stored in the on-chip storage device, and the target queue The current update depth of, that is, the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message is determined as the new current on-chip queue depth of the target queue;

If the current on-chip storage occupancy value is greater than the preset storage occupancy threshold, the target message is moved and discarded, and the on-chip queue depth of the target queue remains unchanged.

Of course, a storage medium containing computer-executable instructions provided in this application is not limited to the method operations described above, and can also perform related operations in the storage scheduling method provided in any embodiment of this application. .

The above are only exemplary embodiments of the present application, and are not used to limit the protection scope of the present application.

Generally speaking, the various embodiments of the present application can be implemented in hardware or dedicated circuits, software, logic or any combination thereof. For example, some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software that may be executed by a controller, microprocessor, or other computing device, although the application is not limited thereto.

A person of ordinary skill in the art can understand that all or some of the steps, systems, and functional modules/units in the device disclosed above can be implemented as software, firmware, hardware, and appropriate combinations thereof.

In the hardware implementation, the division between functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may consist of several physical components. The components are executed cooperatively. Certain physical components or all physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit . Such software may be distributed on a computer-readable medium, and the computer-readable medium may include a computer storage medium (or a non-transitory medium) and a communication medium (or a transitory medium). As is well known by those of ordinary skill in the art, the term computer storage medium includes volatile and non-volatile data implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Sexual, removable and non-removable media. Computer storage media include but are not limited to RAM, ROM, EEPROM, flash memory or other storage technologies, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or Any other medium used to store desired information and that can be accessed by a computer. In addition, as is well known to those of ordinary skill in the art, communication media usually contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as carrier waves or other transmission mechanisms, and may include any information delivery media. .

The preferred embodiments of the embodiments of the present disclosure are described above with reference to the accompanying drawings, and the scope of rights of the embodiments of the present disclosure is not limited thereby. Any modification, equivalent replacement and improvement made by those skilled in the art without departing from the scope and essence of the embodiments of the present disclosure shall fall within the scope of rights of the embodiments of the present disclosure.

Claims

A storage scheduling method, wherein the method includes:

Determining the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message as the update depth of the target queue;

When the update depth of the target queue is greater than the moving threshold of the target queue, determining that the target message needs to be moved to an off-chip storage device for storage;

If the bandwidth of the current off-chip storage device is sufficient and the capacity of the off-chip storage device is sufficient, the target message is moved off-chip and stored on the off-chip storage device, and the on-chip queue depth of the target queue remains unchanged;

If the bandwidth of the current off-chip storage device is insufficient or the capacity of the off-chip storage device is insufficient, and the current on-chip storage occupancy value is less than or equal to the preset storage occupancy threshold, the target packet is stored in the on-chip storage device, and the current update depth of the target queue , Determined as the new current on-chip queue depth of the target queue;

If the current on-chip storage occupancy value is greater than the preset storage occupancy threshold, the target packet is moved and discarded, and the on-chip queue depth of the target queue remains unchanged.
The method according to claim 1, wherein the method further comprises:

When the update depth of the target queue is less than or equal to the movement threshold of the target queue, storing the target message in the on-chip storage space;

The sum of the current on-chip queue depth of the target queue plus the length of the target message is determined as the new current on-chip queue depth of the target queue.
The method according to claim 1, wherein the method further comprises:

If the target message is successfully moved to the off-chip storage device for storage, the current on-chip queue depth of the target queue is determined as the new current on-chip queue depth of the target queue.
The method according to claim 1, wherein the method further comprises:

If the bandwidth of the off-chip storage device is insufficient or the capacity of the off-chip storage device is insufficient, and the current on-chip storage occupancy value is less than or equal to the preset storage occupancy threshold, the length of the target message and the current target queue corresponding to the target message The sum of the on-chip queue depths is determined as the new current on-chip queue depth of the target queue.
The method according to claim 1, wherein the method further comprises:

If the current on-chip storage occupancy value is greater than the preset storage occupancy threshold, and the target message is moved and discarded, the current on-chip queue depth of the target queue is determined as the new current on-chip queue depth of the target queue.
The method according to claim 1, wherein if the bandwidth of the off-chip storage device is insufficient or the capacity of the off-chip storage device is insufficient, and the current on-chip storage occupancy value is less than or equal to a preset storage occupancy threshold, the method further comprises:

The sum of the current on-chip queue depths of all queues is determined as the current on-chip storage occupancy value.
The method according to any one of claims 1 to 6, wherein the preset storage occupancy threshold is a storage occupancy threshold corresponding to the target queue.
The method according to any one of claims 1 to 6, wherein the sum of the length of the received target message and the current on-chip queue depth of the target queue corresponding to the target message is determined as the target Before the update depth of the queue, the method further includes:

Judging the attributes of the target message;

When it is determined that the attribute of the target message is of mixed type, the sum of the length of the target message to be received and the current on-chip queue depth of the target queue corresponding to the target message is determined to be the target The update depth step of the queue.
The method according to claim 8, wherein after said determining the attribute of the target message, the method further comprises:

When it is determined that the attribute of the target message is an on-chip message, storing the target message in on-chip storage;

When it is determined that the attribute of the target message is an off-chip message, the target message is stored in the off-chip storage device.
The method according to any one of claims 1 to 6, wherein before the target message needs to be moved to an off-chip storage device for storage, the method further comprises:

Receiving status information sent by the off-chip storage device;

If it is determined based on the off-chip storage device status information that the off-chip storage device can store the target message, move the target message to the off-chip storage device for storage;

If it is determined according to the off-chip storage device status information that the off-chip storage device cannot store the target message, and the current on-chip storage occupancy value is less than or equal to the preset storage occupancy threshold, the target message is stored in Chip;

If it is determined according to the off-chip storage device status information that the off-chip storage device cannot store the target message, and the current on-chip storage occupancy value is greater than a preset storage occupancy threshold, the target message is moved and discarded.
A storage scheduling device, wherein the device includes a memory, a processor, a program that is stored on the memory and can run on the processor, and is used to realize the connection between the processor and the memory A data bus for communication, when the program is executed by the processor, the steps of the storage scheduling method according to any one of claims 1 to 10 are implemented.
A storage medium used for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement claims 1 to 10 The steps of any one of the storage scheduling method.