CN108984427A - The processing method and processing device of memory in a kind of equipment based on Shen prestige framework - Google Patents

Abstract

Embodiments of the present invention provide a method and device for processing memory in a device based on the Shenwei architecture. The method includes: obtaining the amount of memory used by the device to send and receive messages during the initialization process of the large page memory; The usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage; after each large page memory block merge is completed, the memory usage and the preset memory limit address are cut according to the memory usage and the preset memory limit address. A large block of large page memory to obtain a target memory block; filter in the memory to be allocated of the device according to the PKB identifier or DMA identifier to obtain a target memory block; allocate the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory. The apparatus performs the method described above. The method and device provided by the embodiments of the present invention can avoid the abnormal situation that the equipment based on the Shenwei architecture cannot send and receive messages.

Description

Method and device for processing memory in equipment based on Shenwei architecture

technical field

The embodiment of the present invention relates to the technical field of hardware device failure processing, and in particular to a method and device for processing memory in a device based on the Shenwei architecture.

Background technique

During the initialization process of Shenwei architecture-based devices (including hardware devices such as firewalls), the huge page memory pages will be sorted according to the physical address from small to large, and the large page memory pages with continuous physical addresses belonging to the same physical node will be merged. Thereby merging into a large memory block of large pages, which is managed by a memory descriptor table (rte_memseg), the physical addresses of different large memory blocks of large pages may be separated and discontinuous.

In the prior art, when the above-mentioned large page memory is initialized, as long as the large page memory pages with continuous physical addresses are connected and integrated into a large large page memory block, no interface related to memory address limitation is provided. Similarly, the packet buffer (hereinafter referred to as "packetbuffer" PKB") and memory allocation for sending and receiving messages by DMA do not provide memory address limit related interfaces, so the memory allocated for PKB and DMA sending and receiving messages cannot be guaranteed to be within the preset memory limit address, if it is not in the preset memory Within the restricted address, it will cause the device to have an abnormal situation of sending and receiving messages. For example, when the input/output memory management unit (input/output memory management unit hereinafter referred to as "IOMMU") mode of the device is turned on, the network card will not be able to send and receive messages normally. The situation of the text appears.

Therefore, how to avoid the above-mentioned defects and avoid the abnormal situation that the equipment based on the Shenwei architecture cannot send and receive messages has become an urgent problem to be solved.

Contents of the invention

Aiming at the problems existing in the prior art, embodiments of the present invention provide a method and device for processing memory in a device based on the Shenwei architecture.

In the first aspect, an embodiment of the present invention provides a method for processing memory in a device based on the Shenwei architecture, the method including:

In the process of initializing the large page memory, the memory usage required by the device to send and receive messages is obtained; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage;

After each large block of large page memory is merged, according to the memory usage and the preset memory limit address, the large memory block of the large page is cut to obtain the target memory block;

Filtering the memory to be allocated of the device according to the PKB identifier or the DMA identifier to obtain the target memory block; wherein the PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; The memory to be allocated includes several target memory blocks;

Allocating the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory.

In the second aspect, an embodiment of the present invention provides a device for processing memory in a device based on the Shenwei architecture, and the device includes:

The acquisition unit is used to obtain the memory usage required by the device to send and receive messages during the initialization process of the large page memory; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage ;

A cutting unit, configured to cut the large memory block of the huge page according to the memory usage and the preset memory limit address after each large block merge of the large page memory is completed, so as to obtain the target memory block;

A filtering unit, configured to filter in the memory to be allocated of the device according to the PKB identifier or the DMA identifier, so as to obtain the target memory block; wherein, the PKB identifier or the DMA identifier is used to mark the PKB memory or The DMA memory; the memory to be allocated includes several target memory blocks;

An allocating unit, configured to allocate the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory.

In a third aspect, an embodiment of the present invention provides an electronic device based on the Shenwei architecture, including: a Shenwei processor, a memory, and a bus, wherein,

The Shenwei processor and the memory complete mutual communication through the bus;

The memory stores program instructions executable by the processor, and the processor calls the program instructions to perform the following methods:

In the process of initializing the large page memory, the memory usage required by the device to send and receive messages is obtained; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage;

After each large block of large page memory is merged, according to the memory usage and the preset memory limit address, the large memory block of the large page is cut to obtain the target memory block;

Filtering the memory to be allocated of the device according to the PKB identifier or the DMA identifier to obtain the target memory block; wherein the PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; The memory to be allocated includes several target memory blocks;

Allocating the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory.

In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, including:

The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to perform the following method:

In the process of initializing the large page memory, the memory usage required by the device to send and receive messages is obtained; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage;

After each large block of large page memory is merged, according to the memory usage and the preset memory limit address, the large memory block of the large page is cut to obtain the target memory block;

Filtering the memory to be allocated of the device according to the PKB identifier or the DMA identifier to obtain the target memory block; wherein the PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; The memory to be allocated includes several target memory blocks;

Allocating the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory.

The method and device for processing the memory in the device based on the Shenwei architecture provided by the embodiments of the present invention cut and obtain the large page memory according to the memory usage required for the device to send and receive messages and the preset memory limit address during the initialization process of the large page memory. to the target memory block; in the process of memory allocation, filter the memory to be allocated through the PKB memory ID or DMA memory ID to obtain and allocate the target memory block, which can avoid the failure of sending and receiving messages in the device based on Shenwei architecture abnormal situation.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

FIG. 1 is a schematic flowchart of a processing method for memory in a device based on the Shenwei architecture according to an embodiment of the present invention;

2 is a schematic structural diagram of a memory processing device in a device based on the Shenwei architecture according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a physical structure of an electronic device provided by an embodiment of the present invention.

Detailed ways

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

Fig. 1 is a schematic flowchart of a memory processing method in a device based on the Shenwei architecture according to an embodiment of the present invention. As shown in Fig. 1 , a method for processing memory in a device based on the Shenwei architecture provided by an embodiment of the present invention includes the following steps :

S101: In the process of initializing the large page memory, acquire the memory usage required by the device to send and receive messages; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage.

Specifically, in the process of initializing the large page memory, the device obtains the memory usage required by the device to send and receive messages; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage. A device may be understood as a component in the device having a logic operation function, and is used to execute the method of the embodiment of the present invention. The huge page memory can be initialized through the rte_eal_hugepage_init() function in the Data Plane Development Kit (hereinafter referred to as "DPDK"), and then the rte_eal_hugepage_init() function can be used to call the memory_get_pkb_dma_size() function to obtain the Memory usage, what needs to be explained is: the memory_get_pkb_dma_size() function is based on the installed memory of the device, and the corresponding reserved PKB+DMA memory is packaged. For example, the maximum memory capacity that the device can support is 8G, and the installed memory capacity If it is 4G (reserved PKB+DMA memory is deducted from 4G), the reserved PKB+DMA memory can correspond to custom 128M; the installed memory capacity is 8G (reserved PKB+DMA memory is deducted from 8G) , the reserved PKB+DMA memory can correspond to a custom 256M. In order to facilitate subsequent comparison calculations, the memory usage can be converted into bytes.

S102: Cut the huge page memory block according to the memory usage and the preset memory limit address after each huge page memory block merge is completed, so as to obtain the target memory block.

Specifically, after each huge page memory chunk merge is completed, the device cuts the huge page memory chunk according to the memory usage and the preset memory limit address, so as to obtain the target memory block. Huge page memory block merging can be understood as being carried out during the above huge page memory initialization process. The specific value of the preset memory limit address can be determined according to the index parameters of the hardware information of the device. The hardware information can include CPU information and/or Or the PCI bus information, the index parameter corresponding to the CPU information can include the architecture and/or main frequency of the CPU, but not specifically limited; the index parameter corresponding to the PCI bus information can include The parameters of the hanging peripheral information, but not specifically limited. The preset memory limit address can be set independently according to the actual situation, and it can be 3G (0xC0000000). The step of cutting may specifically include: cutting the large block of the huge page memory at a time, so that the total amount of memory corresponding to at least one part of the two parts after cutting is greater than or equal to the amount of memory used, and the start and end addresses of the total amount of memory are less than The preset memory limit address; using the at least a part as the target memory block. Normally, it is only necessary to cut the large page memory block whose start address is smaller than the memory limit address and whose end address is greater than the memory limit address, using the memory limit address as the critical value. For example, assuming that the default memory limit address is selected as 3G (0xC0000000), the amount of sending and receiving memory is selected as 128M, and the memory size of large block A of huge page memory is 7G. The value is cut, and the first part is from the start address of the huge page memory block A to the end address of the memory limit address 3G (0xC0000000); and the start address of the memory limit address 3G (0xC0000000) to the large page memory block A block of the end address is used as the second part. If the total memory of the first part after cutting is greater than the required sending and receiving memory usage of 128M, the first part after cutting will be used as the target memory block; while the second part after cutting cannot be used as the target memory block. target memory block. For another situation, if the memory size of large block B of the huge page memory is 1G, it is a large block whose start and end addresses are smaller than the memory limit address, and the total amount of memory is greater than the memory usage of sending and receiving 128M, then the large memory block of the large page B can be directly used as the target memory block without cutting.

S103: Filter the memory to be allocated of the device according to the PKB identifier or the DMA identifier to obtain the target memory block; wherein the PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; the memory to be allocated includes several target memory blocks.

Specifically, the apparatus performs filtering in the memory to be allocated of the device according to the PKB identifier or the DMA identifier, so as to obtain the target memory block; wherein, the PKB identifier or the DMA identifier is used to mark the PKB memory or the The DMA memory; the memory to be allocated includes several target memory blocks. The PKB identifier or DMA identifier can be the respective name string of the PKB memory or the DMA memory, and the name string can be written in the corresponding position of the malloc_heap_alloc() function called by DPDK. The malloc_heap_alloc() function is used to describe the remaining memory Find (call find_suitable_element() function) available memory required for allocation (call malloc_elem_alloc() function) in character table (free_memseg). That is, when the calling function detects the name string, it can know that the required and available memory is related to the PKB memory or the DMA memory. The specific step of obtaining the target memory block may be: according to the PKB identifier or the DMA identifier, directly skip the target large page memory block in the memory to be allocated until the target memory block is obtained; wherein, the target large page The start and end physical addresses of the page memory block are greater than the preset memory limit address. Refer to the above example: if the start and end physical addresses are greater than the memory limit address 3G, skip directly (the memory is used according to the physical address from high to low).

Furthermore, the find_suitable_element() function can be called in DPDK, so that the find_suitable_element() function can jump directly to the large memory block of the large page in the memory to be allocated that does not meet the filtering condition according to the name string through until the target memory block is acquired. It should be noted that the processing flow of other non-DMA memory and non-PKB memory is not considered in the scope of the embodiment of the present invention.

S104: Allocate the target memory block to the PKB memory and the DMA memory, so as to be used by the PKB memory and the DMA memory.

Specifically, the device allocates the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory. Can look up (call find_suitable_element () function) in the remaining memory descriptor table (free_memseg) by malloc_heap_alloc () function in the above-mentioned embodiment and allocate (call malloc_elem_alloc () function) required available memory, for described PKB memory and the DMA memory usage.

The embodiment of the present invention provides a method for processing memory in a device based on the Shenwei architecture. During the initialization process of the large page memory, according to the memory usage required by the device to send and receive messages and the preset memory limit address, the target is cut and obtained. Memory block; in the process of memory allocation, filter the memory to be allocated through the PKB memory ID or DMA memory ID to obtain and allocate the target memory block, which can avoid the abnormality that the device based on Shenwei architecture cannot send and receive messages situation.

On the basis of the above-mentioned embodiments, the cutting the huge page memory block according to the memory usage and the preset memory limit address to obtain the target memory block includes:

Cut the huge page memory block at a time, so that the total amount of memory corresponding to at least one of the two parts after cutting is greater than or equal to the amount of memory used, and the start and end addresses of the total amount of memory are less than the preset memory limit address.

Specifically, the device cuts the large block of huge page memory at a time, so that the total amount of memory corresponding to at least one of the two parts after cutting is greater than or equal to the amount of memory used, and the start and end addresses of the total amount of memory are smaller than the Default memory limit address. Reference may be made to the foregoing embodiments, and details are not repeated here.

The at least one part is used as the memory block.

Specifically, the device uses the at least a part as the memory block. Reference may be made to the foregoing embodiments, and details are not repeated here.

The embodiment of the present invention provides a method for processing memory in a device based on the Shenwei architecture, so that the total amount of memory corresponding to the divided memory block is greater than or equal to the amount of memory used, and the start and end addresses of the total amount of memory are less than the preset memory limit address, It can effectively avoid the abnormal situation that the equipment based on Shenwei architecture cannot send and receive messages.

On the basis of the foregoing embodiments, the filtering in the memory to be allocated of the device according to the PKB identifier or the DMA identifier to obtain the target memory block includes:

According to the PKB identifier or the DMA identifier, directly skip the target huge page memory block in the memory to be allocated until the target memory block is obtained; wherein, the start and end of the target huge page memory block The physical address is greater than the preset memory limit address.

Specifically, according to the PKB identifier or the DMA identifier, the device directly skips the target huge page memory block in the memory to be allocated until the target memory block is obtained; wherein, the target huge page memory The starting and ending physical addresses of the large block are greater than the preset memory limit address. Reference may be made to the foregoing embodiments, and details are not repeated here.

The method for processing memory in a device based on the Shenwei architecture provided by the embodiment of the present invention can quickly obtain the target memory block by filtering out large blocks of target large page memory in the memory to be allocated.

On the basis of the above embodiments, the PKB identifier or the DMA identifier is the respective name string of the PKB memory or the DMA memory; correspondingly, according to the PKB identifier or the DMA identifier, Directly skip the target large page memory block in the memory to be allocated until the target memory block is obtained, including:

By calling the find_suitable_element() function in the data plane development kit DPDK, the find_suitable_element() function will directly skip the target large page memory block in the memory to be allocated according to the name string until it is obtained The target memory block.

Specifically, the device calls the find_suitable_element() function in the data plane development kit DPDK, so that the find_suitable_element() function directly skips the target large page memory block in the memory to be allocated according to the name string , until the target memory block is acquired. Reference may be made to the foregoing embodiments, and details are not repeated here.

The method for processing memory in a device based on the Shenwei architecture provided by the embodiment of the present invention can filter out large blocks of target large page memory in the memory to be allocated by calling corresponding functions, and can further obtain the target memory block conveniently, effectively and quickly .

On the basis of the above embodiments, the acquisition of the memory usage required by the device includes:

The huge page memory initialization is performed through the rte_eal_hugepage_init() function in the data plane development kit DPDK.

Specifically, the device initializes the huge page memory through the rte_eal_hugepage_init() function in the data plane development kit DPDK. Reference may be made to the foregoing embodiments, and details are not repeated here.

Get the PKB and DMA memory usage by making the rte_eal_hugepage_init() function call the memory_get_pkb_dma_size() function.

Specifically, the device obtains the PKB and DMA memory usage by making the rte_eal_hugepage_init() function call the memory_get_pkb_dma_size() function. Reference may be made to the foregoing embodiments, and details are not repeated here.

The method for processing memory in a device based on the Shenwei architecture provided by the embodiment of the present invention can quickly and efficiently obtain memory usage by calling corresponding functions.

On the basis of the foregoing embodiments, the method further includes:

Get hardware information for the device.

Specifically, the device acquires hardware information of the device. Reference may be made to the foregoing embodiments, and details are not repeated here.

The specific value of the preset memory limit address is determined according to the index parameter of the hardware information.

Specifically, the device determines the specific value of the preset memory limit address according to the index parameter of the hardware information. Reference may be made to the foregoing embodiments, and details are not repeated here.

The method for processing the memory in the device based on the Shenwei architecture provided by the embodiment of the present invention can reasonably and accurately determine the specific value of the preset memory limit address.

On the basis of the foregoing embodiments, the hardware information includes CPU information and/or PCI bus information; correspondingly, the index parameter corresponding to the CPU information includes CPU architecture and/or main frequency; The index parameters corresponding to the PCI bus information include parameters of peripheral device information connected to the PCI bus.

Specifically, the index parameter corresponding to the CPU information in the device includes the system architecture and/or main frequency of the CPU; the index parameter corresponding to the PCI bus information includes the peripheral information of the PCI bus parameter. Reference may be made to the foregoing embodiments, and details are not repeated here.

The method for processing memory in a device based on the Shenwei architecture provided by the embodiment of the present invention, through CPU information and/or PCI bus information, and corresponding specific index parameters; can further reasonably and accurately determine the specific address of the preset memory limit address value.

Fig. 2 is a schematic structural diagram of a memory processing device in a device based on the Shenwei architecture according to an embodiment of the present invention. As shown in Fig. 2, an embodiment of the present invention provides a memory processing device in a device based on the Shenwei architecture, including an acquisition unit 201, cutting unit 202, filtering unit 203 and distribution unit 204, wherein:

The obtaining unit 201 is used to obtain the memory usage required by the device to send and receive messages during the initialization process of the large page memory; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage ; The cutting unit 202 is used to cut the large memory block of the large page according to the memory usage and the preset memory limit address to obtain the target memory block after each large memory block merge is completed; the filtering unit 203 is used to filter the to-be-allocated memory of the device according to the PKB identifier or the DMA identifier, so as to obtain the target memory block; wherein, the PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; the memory to be allocated includes several target memory blocks; the allocation unit 204 is used to allocate the target memory block to the PKB memory and the DMA memory for the PKB memory and the DMA memory use.

Specifically, the obtaining unit 201 is used to obtain the memory usage required by the device to send and receive messages during the initialization process of the large page memory; the memory usage is the data packet cache PKB memory usage and the direct memory access DMA memory usage The sum of the amount; the cutting unit 202 is used to cut the large memory block of the huge page according to the memory usage and the preset memory limit address after completing the large block merge of the large page memory, so as to obtain the target memory block ; The filtering unit 203 is used to filter in the memory to be allocated of the device according to the PKB identifier or the DMA identifier, so as to obtain the target memory block; wherein, the PKB identifier or the DMA identifier is used to mark the PKB memory Or the DMA memory; the memory to be allocated includes several target memory blocks; the allocation unit 204 is used to allocate the target memory block to the PKB memory and the DMA memory, for the PKB memory and the DMA memory usage described above.

The memory processing device in the device based on the Shenwei architecture provided by the embodiment of the present invention cuts and obtains the target memory block according to the memory usage required by the device and the preset memory limit address during the initialization process of the large page memory; In the process of memory allocation, filter the memory to be allocated through PKB memory identification or DMA memory identification to obtain and allocate the target memory block, which can avoid the abnormal situation that the Shenwei architecture-based device cannot send and receive messages.

The memory processing device in the device based on the Shenwei architecture provided by the embodiment of the present invention can be specifically used to execute the processing procedures of the above-mentioned method embodiments, and its functions will not be described in detail here, and the detailed description of the above-mentioned method embodiments can be referred to.

FIG. 3 is a schematic diagram of the physical structure of the electronic device provided by the embodiment of the present invention. As shown in FIG. 3 , the electronic device based on the Shenwei architecture includes: a Shenwei processor (processor) 301, a memory (memory) 302 and a bus 303;

Wherein, the Shenwei processor 301 and the memory 302 complete mutual communication through the bus 303;

The Shenwei processor 301 is used to call the program instructions in the memory 302 to execute the methods provided by the above method embodiments, for example, including: in the process of initializing the large page memory, obtaining the information required by the device to send and receive messages. memory usage; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage; Set the memory limit address, cut the large memory block of the huge page, to obtain the target memory block; filter in the memory to be allocated of the device according to the PKB identifier or the DMA identifier, to obtain the target memory block; wherein, The PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; the memory to be allocated includes several target memory blocks; the target memory block is allocated to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory.

This embodiment discloses a computer program product, the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program includes program instructions, and when the program instructions are executed by the computer, the computer The method provided by each of the above method embodiments can be executed, for example, including: in the process of initializing the large page memory, obtaining the memory usage required by the device to send and receive messages; the memory usage is the packet cache PKB memory usage and The sum of direct memory access DMA memory usage; after each huge page memory chunk merge is completed, according to the memory usage and the preset memory limit address, cut the huge page memory chunk to obtain the target A memory block; filter the memory to be allocated of the device according to the PKB identifier or the DMA identifier to obtain the target memory block; wherein the PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; the memory to be allocated includes several target memory blocks; allocating the target memory blocks to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory.

This embodiment provides a non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the methods provided in the above method embodiments, for example, including : In the process of large page memory initialization, obtain the memory usage required by the device to send and receive messages; the memory usage is the sum of the packet buffer PKB memory usage and the direct memory access DMA memory usage; After a large block of large page memory is merged, cut the large block of large page memory according to the memory usage and the preset memory limit address to obtain the target memory block; according to the PKB identification or DMA identification in the device Filtering in the memory to be allocated to obtain the target memory block; wherein, the PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; the memory to be allocated includes several targets A memory block; assigning the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory.

Those of ordinary skill in the art can understand that all or part of the steps for realizing the above-mentioned method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the It includes the steps of the above-mentioned method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

The above-described embodiments such as electronic equipment are only illustrative, and the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, It can be located in one place, or it can be distributed to multiple network elements. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. It can be understood and implemented by those skilled in the art without any creative effort.

Through the above description of the implementations, those skilled in the art can clearly understand that each implementation can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of software products, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic discs, optical discs, etc., including several instructions to make a computer device (which may be a personal computer, server, or network device, etc.) execute the methods described in various embodiments or some parts of the embodiments.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, not to limit them; although the embodiments of the present invention have been described in detail with reference to the foregoing embodiments, those skilled in the art The skilled person should understand that: it is still possible to modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the present invention The scope of the technical solutions of each embodiment.

Claims (10)

1. A method for processing memory in a device based on the Shenwei architecture, characterized in that it comprises: In the process of initializing the large page memory, the memory usage required by the device to send and receive messages is obtained; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage; After each large block of large page memory is merged, according to the memory usage and the preset memory limit address, the large memory block of the large page is cut to obtain the target memory block; Filtering the memory to be allocated of the device according to the PKB identifier or the DMA identifier to obtain the target memory block; wherein the PKB identifier or the DMA identifier is used to mark the PKB memory or the DMA memory; The memory to be allocated includes several target memory blocks; Allocating the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory. 2. The method according to claim 1, wherein, according to the memory usage and the preset memory limit address, cutting the large memory block of the huge page to obtain the target memory block comprises: Cut the huge page memory block at a time, so that the total amount of memory corresponding to at least one of the two parts after cutting is greater than or equal to the amount of memory used, and the start and end addresses of the total amount of memory are less than the preset memory limit address; The at least one part is used as the target memory block. 3. The method according to claim 1 or 2, wherein the filtering in the memory to be allocated of the device according to the PKB identifier or the DMA identifier to obtain the target memory block comprises: According to the PKB identifier or the DMA identifier, directly skip the target huge page memory block in the memory to be allocated until the target memory block is obtained; wherein, the start and end of the target huge page memory block The physical address is greater than the preset memory limit address. 4. The method according to claim 3, wherein the PKB identifier or the DMA identifier are respective name strings of the PKB memory or the DMA memory; correspondingly, the PKB identifier Or the DMA flag directly skips the target large page memory block in the memory to be allocated until the target memory block is obtained, including: By calling the find_suitable_element() function in the data plane development kit DPDK, the find_suitable_element() function will directly skip the target large page memory block in the memory to be allocated according to the name string until it is obtained The target memory block. 5. The method according to claim 1 or 2, wherein said obtaining the memory usage required by the device comprises: Perform the initialization of the huge page memory through the rte_eal_hugepage_init() function in the data plane development kit DPDK; Get the PKB and DMA memory usage by making the rte_eal_hugepage_init() function call the memory_get_pkb_dma_size() function. 6. The method according to claim 1 or 2, wherein the method further comprises: Obtain hardware information of the device; The specific value of the preset memory limit address is determined according to the index parameter of the hardware information. 7. The method according to claim 6, wherein the hardware information includes CPU information and/or PCI bus information; correspondingly, the index parameter corresponding to the CPU information includes CPU architecture and/or Or the main frequency; the index parameter corresponding to the PCI bus information includes the parameters of the peripheral device information connected to the PCI bus. 8. A memory processing device based on the Shenwei architecture, characterized in that it includes: The acquisition unit is used to obtain the memory usage required by the device to send and receive messages during the initialization process of the large page memory; the memory usage is the sum of the packet cache PKB memory usage and the direct memory access DMA memory usage ; A cutting unit, configured to cut the large memory block of the huge page according to the memory usage and the preset memory limit address after each large block merge of the large page memory is completed, so as to obtain the target memory block; A filtering unit, configured to filter in the memory to be allocated of the device according to the PKB identifier or the DMA identifier, so as to obtain the target memory block; wherein, the PKB identifier or the DMA identifier is used to mark the PKB memory or The DMA memory; the memory to be allocated includes several target memory blocks; An allocating unit, configured to allocate the target memory block to the PKB memory and the DMA memory for use by the PKB memory and the DMA memory. 9. An electronic device based on Shenwei architecture, characterized in that it includes: Shenwei processor, memory and bus, wherein, The Shenwei processor and the memory complete mutual communication through the bus; The memory stores program instructions executable by the processor, and the Shenwei processor invokes the program instructions to execute the method as claimed in any one of claims 1 to 7. 10. A non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the computer according to any one of claims 1 to 7. Methods.