JP2018196024A - Communication device and control method therefor, and information processing device - Google Patents

Abstract

To automatically generate different communication headers for a plurality of transmit packets by one session of setting when an information processing device transfers transmit information separately in a plurality of transmit packets via a network to an external apparatus and improve the transmission processing capability of the information processing device.SOLUTION: Provided is a network communication device, connected to an information processing device, for connecting the information processing device to a network, comprising: a determination unit for determining an N that represents the number of packets from the size of information to be transferred when a request for transfer to the network is received from the information processing device; a generation unit for generating N communication headers from a template communication header that conforms to the transfer request and dividing the information to be transferred into N payloads; an editing unit for doing different editing to the first, the second to N-1'th and the N'th communication headers among as many communication headers as the number of headers; and a transmission unit for transmitting N packets that are a combination of communication headers having undergone editing by the editing unit and payloads to the network.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a communication header generation technique for transmission packets in communication protocol processing.

  In recent years, there is an increasing need for browsing high-resolution moving images (8K / 4K) via a network using a multifunction terminal called a smart device. Naturally, this results in increased data traffic on the network. Network acceleration such as 10 Gbps communication by Ethernet (registered trademark) and 5G communication of mobile communication means that the load on the CPU for communication protocol processing such as TCP / UDP / IP becomes large. Therefore, measures have been proposed for realizing these high-speed data communications while reducing the load of communication protocol processing by the CPU.

  The communication protocol process is defined in seven layers as an OSI (Open Systems Interconnection) reference model divided into a hierarchical structure established by the International Organization for Standardization (ISO).

  There is TSO (TCP Segmentation Offload) as a processing method for speeding up transmission of communication protocol processing. TSO offloads a part of communication protocol processing (TCP segment processing) executed by the CPU to reduce the load on the CPU. This TCP segment process is a process for converting transmission data generated in the application layer defined by the OSI reference model into a communication packet used in a communication layer which is a lower layer.

  As an example of TCP segment processing, a transmission packet generation method described in Patent Document 1 is known. In this method, first, the payload dividing unit divides transmission data exceeding the transferable length generated in the application layer into a transferable length to generate a transmission payload. And a header production | generation part produces | generates the header provided to the divided | segmented transmission payload. Then, the header / payload combining unit combines the generated header and the divided transmission payload to complete the transmission packet.

JP 2006-81033 A

  The transmission packet generation method described in Patent Document 1 is configured to generate transmission packets from transmission data in order for batch input of transmission data exceeding the transferable length, and sequentially transmit the transmission packets from the network. The payload dividing method for transmission packets is a method in which the payload dividing unit divides transmission data generated in the application layer into transmission payloads at once.

  However, in Patent Document 1, the header generation unit generates one header at a time for each divided transmission payload. For this reason, this header generation method requires header information and header generation instruction information every time a header is generated, and thus there is a problem that it takes time to collect header information and generate instruction information.

In order to solve this problem, for example, the communication device of the present invention has the following configuration. That is,
A communication device for connecting an information processing device to a network,
Determining means for determining N representing the number of packets from the size of information to be transferred when a transfer request to the network is received from the information processing apparatus;
Generating means for generating N communication headers based on a model communication header corresponding to the transfer request;
Dividing means for dividing the information to be transferred into N payloads;
Transmitting means for transmitting N packets, which are a combination of the communication header generated by the generating means and the payload, to the network;
Have
When generating the N communication headers, the generation means generates the first, second to N−1th, and Nth of the N communication headers according to an instruction included in the transfer request. The communication header of the template is generated by editing at least one of the communication headers.

  According to the present invention, when an information processing apparatus divides transmission information into a plurality of transmission packets and transfers them to an external device via a network, different communication headers are automatically generated with a single setting of communication headers of the plurality of transmission packets. It becomes possible to do. As a result, the transmission information can be divided and a plurality of transmission packets can be generated at a high speed by a single setting, and the transmission processing capability of the information processing apparatus can be improved.

The functional block diagram of the communication apparatus in embodiment. The transmission operation | movement flowchart of the communication apparatus in embodiment. The figure which shows the example of the register setting information in embodiment. The figure which shows the communication header of the transmission packet in embodiment. The flowchart which shows the communication header production | generation process in embodiment.

  Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

[First Embodiment]
FIG. 1 is a diagram illustrating functional blocks of the communication device 100 according to the first embodiment. The communication device 100 includes a main CPU 102, a main memory 103, and a subsystem 104 that executes communication protocol processing as main functional blocks. The functional blocks 102 to 104 in the communication apparatus 100 are connected to each other via the system bus 101.

  For example, the communication device 100 is an information processing device such as a personal computer or a smart phone or a terminal device, and the subsystem 104 is easily understood as an interface unit for network communication such as a NIC (Network Interface Card). The communication form of the network may be wired or wireless. Note that the subsystem 104 may be realized by an ASIC or the like mounted on a substrate serving as the communication apparatus 100, or may be a card connected to a general-purpose bus such as PCI-Express. Furthermore, it may be a device connected to the USB.

  The main CPU 102 executes application software in the application layer defined by the OSI reference model, and also executes control of the entire communication device. The main CPU 102 loads an OS (operating system), data, various driver programs, and the like necessary for executing the application software from an external storage device (not shown) to the main memory 103 and executes them.

  The main memory 103 is a memory that can be shared and used by each functional block in the communication apparatus, and is mainly composed of a semiconductor memory represented by a DRAM (Dynamic Random Access Memory). The internal area is divided into an area for storing an OS and a program, an application buffer area used by application software, and the like.

  The sub system 104 includes a sub CPU 106, a local memory 107, a header processing unit 111, a DMA controller 108, and a communication interface 109 as main functional blocks. The reference numerals 106 to 109 are connected to each other via the local bus 105.

  The sub-system 104 is a sub-system that executes communication protocol processing in the communication device. The sub-system 104 specializes in communication protocol processing such as TCP / IP protocol and handles data transfer processing via an external device and a network. Then, hardware dedicated to communication protocol processing is installed, and high-speed communication protocol processing is executed instead of the main CPU 102. Note that the TCP / IP protocol is an example shown for explanation, and the type of communication protocol is not limited to this.

  The sub CPU 106 executes communication protocol processing. Specifically, the sub CPU 106 performs IPv4 (IP version 4), IPv6 (IP version 6), UDP, TCP communication protocol processing (transmission sequence control, congestion control, communication error control, etc.). In addition, the sub CPU 106 may be configured as one processor or may be configured as a multiprocessor with a plurality of processors. Further, it may be configured by hardware that accelerates a processor and some functions.

  The local memory 107 is a memory that can be used by the sub CPU 106 of the subsystem 104, and is mainly configured by a semiconductor memory represented by an SRAM (Static Random Access Memory). The main CPU 102 can also access the local memory 107 via the system bus 101 and the local bus 105.

  The DMA controller 108 executes DMA transfer according to an instruction from the sub CPU 106. The DMA controller 108 executes DMA transfer according to the contents described in the descriptor. Here, the descriptor is information in which information such as address information on the source side that performs data transfer, address information on the destination side, and transfer size of data transfer is described. After completing the writing of the descriptor in the local memory 107, the sub CPU 106 notifies the DMA controller 108 of a DMA transfer start instruction. For this reason, the sub CPU 106 notifies (sets) the address information of the local memory 107 storing the descriptor to the DMA controller 108.

  Upon receiving this notification, the DMA controller 108 executes DMA transfer between the main memory 103 and the local memory 107 and between the internal memory 113 and the local memory 107. Further, the DMA controller 108 also executes DMA transfer processing between the internal memory 113 and the communication interface, and between the local memory 107 and the communication interface.

  The communication interface 109 is an interface with a network 110 represented by a LAN (Local Area Network), and is responsible for communication control of a physical layer (PHY layer) and a data link layer (MAC layer) in communication. . Equipped with a buffer memory that temporarily stores transmission / reception information. The DMA controller 108 performs DMA transfer via this buffer memory, and input / output of transmission / reception information to / from the network is performed. Note that the communication interface 109 may incorporate the DMA controller 108 and use the DMA controller.

  The header processing unit 111 includes a register unit 112, an internal memory 113, and a core unit 114 as main functional blocks. And the header process part 111 bears the communication header production | generation process in a communication protocol process using these. Specifically, the header processing unit 111 executes the communication header generation process in the communication protocol process executed by the sub CPU 106 instead.

  The register unit 112 stores setting information for the core control unit 118 to control the header duplicating unit 115, the header editing unit 116, and the switch 117. Further, the register unit 112 also stores status information for confirming the states of the functional blocks 115 to 117.

  Then, the sub CPU 106 stores setting information in the register unit 112 in order to set header information (hereinafter referred to as “model header”) serving as a template and editing information for generating a communication header.

  The core unit 114 includes a header duplicating unit 115, a header editing unit 116, a switch 117, and a core control unit 118. The core unit 114 executes communication header generation processing according to the setting information of the register unit 112 and stores the generated communication header in the internal memory 113.

  The internal memory 113 is a memory that can be used by the core unit 114 of the header processing unit 111 and the sub CPU 106, and is mainly configured by a semiconductor memory represented by SRAM (Static Random Access Memory). The internal memory 113 stores the communication header generated by the core unit 114.

  The core control unit 118 controls the header duplication unit 115, the header editing unit 116, and the switch 117 in accordance with the setting information set in the register unit 112.

  The header duplicating unit 115 duplicates (stores) the template header information set in the register unit 112 according to an instruction from the core control unit 118 to the internal memory 113 through the switch 117 as many times as instructed.

  The header editing unit 116 edits each template header information copied to the internal memory 113 in accordance with an instruction from the core control unit 118. At this time, the header editing unit 116 overwrites only necessary portions of each template header information via the switch 117. The header editing unit 116 is also equipped with a function of counting the number of headers subjected to editing processing when editing each template header information copied to the internal memory 113.

  The switch 117 switches the data path connection between the header duplicating unit 115 or the header editing unit 116 and the internal memory 113 in accordance with an instruction from the core control unit 118.

  Next, in the communication apparatus 100 shown in FIG. 1, an operation flow in the communication apparatus 100 when connecting to an external device via the network 110 and transmitting a transmission packet to the external device is described with reference to the flowchart of FIG. explain. Here, the TCP / IP protocol is described as an example of the communication protocol used when the communication apparatus 100 transfers data to / from an external device. The TCP / IP protocol is taken as an example because the technical contents can be easily grasped by using a widely known representative protocol. It should be understood as an example only.

S201 (terminal):
This terminal indicates the start of this flowchart. It shows that the communication apparatus 100 is in communication with an external device via the network 110.

S202 (processing):
Application software executed by the main CPU 102 prepares transmission information for transmission to an external device in the application buffer area of the main memory 103. When the transmission information is accumulated up to a predetermined amount in the application buffer area, the main CPU 102 performs processing for notifying the sub CPU 106 that preparation of transmission information to be transmitted to the external device is completed. That is, the main CPU 102 issues a transfer request to the sub CPU 106.

  This notification includes information indicating the current communication state. For example, it is information indicating that a communication connection using the TCP protocol is started. Alternatively, it is information indicating that data transfer using the TCP protocol is started. Alternatively, it is information for starting communication disconnection processing using the TCP protocol.

S203 (condition judgment):
This condition determination indicates that the route selection is performed on the condition that the application software executed by the main CPU 102 is ready for transmission information to be transmitted to the external device. If the accumulation of transmission information for transmission to an external device in the application buffer area of the main memory 103 has been completed, the process proceeds to S204, and if not, the process returns to S202.

  This condition determination is determined by the main CPU 102 based on a response indicating whether the sub CPU 106 has received the notification issued in the process of S202.

S204 (processing):
This process indicates a process in which the sub CPU 106 moves transmission information for executing transmission to an external device according to the communication protocol from the main memory 103 to the local memory 107 in the subsystem 104.

  Here, the sub CPU 106 performs DMA transfer of transmission information from the main memory 103 to the local memory 107 based on the notification received from the main CPU 102 in the process of S202. For the DMA transfer setting, the sub CPU 106 generates a descriptor describing the contents of the DMA transfer and stores it in the local memory 107. Then, the sub CPU 106 sets the address information of the local memory 107 storing the descriptor in the DMA controller 108. As a result, the DMA controller 108 transfers transmission information from the main memory 103 to the local memory 107. Upon completion of this DMA transfer, preparation for transmission processing according to the communication protocol by the sub CPU 106 is completed.

S205 (processing):
This process indicates a process in which the sub CPU 106 divides the transmission information received in the process of S204 to determine the payload part of the transmission packet, and a process in which the header processing unit 111 generates a communication header part of the transmission packet. ing.

  Therefore, first, the sub CPU 106 determines the size per transmission packet and the number of transmission packets from the transmission information moved to the local memory 107. Next, the sub CPU 106 sets information necessary for the header processing unit 111 to generate the communication header part of the transmission packet in the register unit 112 and activates the header processing unit 111. The activated header processing unit 111 executes processing for generating the same number of communication headers as the number of transmission packets to be generated in the internal memory 113 in the header processing unit 111. Thereby, the determination of the payload portion of the transmission packet and the generation of the communication header portion are completed.

  Note that the sub CPU 106 changes the setting contents in the register unit 112 of the header processing unit 111 according to the current communication state included in the notification by the process of S202. Accordingly, the communication header generated by the header processing unit 111 in the internal memory 113 is generated as a communication header corresponding to the current communication state. For example, if the communication connection using the TCP protocol is started, a communication header necessary for starting communication is generated. Alternatively, if the data transfer using the TCP protocol is started, a communication header necessary for starting the data transfer is generated. Alternatively, if the communication disconnection process using the TCP protocol is started, a communication header necessary for disconnecting the communication is generated. The detailed description will be described later with reference to FIGS.

S206 (processing):
This process indicates a process in which the sub CPU 106 generates a transmission packet from the payload part determined in S205 and the generated communication header part, and transmits the transmission packet to the external device via the communication interface 109.

  First, the sub CPU 106 cuts out transmission information in the local memory 107 in units of payload, and performs DMA transfer setting in which it is combined with each communication header generated in the internal memory 113 and transferred as one transmission packet. Specifically, the sub CPU 106 transfers one communication header stored in the internal memory 113 and a corresponding payload subsequently stored in the local memory 107 to the communication interface 109. When one set of transfer of the header and payload is set as one set, the set processing for the number of packets to be transmitted is repeated. This transfer is executed by DMA transfer by the DMA controller 108. When the DMA transfer is activated by the sub CPU 106, the transmission packet (is transferred to the communication interface 109).

  Next, after the communication processing of the data link layer (MAC layer) and the physical layer (PHY layer) is performed in the communication interface 109, the generated transmission packets are sequentially transmitted to the external device via the network.

S207 (condition judgment):
This condition determination indicates that route selection is performed on the condition that a factor for terminating the communication state between the communication apparatus 100 and the external device has occurred. If a factor for terminating the state of communication with the external device occurs, the process proceeds to process symbol S208, and if not, the process proceeds to condition determination in S202.

  Here, when the application software executed by the main CPU 102 detects an instruction to end the connection between the communication apparatus and the external device by a user operation or the like, the above-described factors are caused. In addition, the above-described factor is also caused when the communication interface 109 detects that the communication line has been cut off by the user disconnecting the network cable from the communication device. The above-mentioned factor is also caused when a received packet containing a notification to disconnect the communication line is received from an external device that is a communication partner.

S208 (processing):
This process is a process for terminating the communication state between the communication apparatus 100 and the external device. Specifically, the communication line with the external device is disconnected in accordance with the communication protocol process carried out by the sub CPU 106. Then, the related information temporarily stored in the local memory 107 or the internal memory 113 of the header processing unit 111 is deleted. Thereafter, when a notification of the completion of disconnection of the communication line is transmitted from the sub CPU 106 to the main CPU 102, a process for displaying that the communication line has been disconnected by the application software is performed.

S209 (terminal):
This terminal indicates the end of this flowchart. This indicates that the communication device is not in communication with the external device via the network.

  Next, FIG. 3 shows an example of setting information set by the sub CPU 106 in the register unit 112 of the header processing unit 111 in the processing of S205 in FIG.

  A reference numeral 301 shown in the figure is a template header, and is header information that is a base when a plurality of communication headers are generated by being duplicated by the header processing unit 111. The internal configuration is divided into, for example, a transport layer TCP header defined in the OSI reference model and a network layer IP header. The TCP header and IP header are used when executing communication protocol processing in each layer.

  Note that the template header may include an Ether header as well as a TCP header and an IP header. The Ether header is a communication header used in the data link layer defined in the OSI reference model in which the destination MAC address and the source MAC address are described. Further, a UDP header may be used for the template header instead of the TCP header.

  Reference numerals 302 to 304 are editing information necessary for generating a communication header. The editing information includes copy number information 302 for specifying the number of times of copying the template header, protocol type information 303 for specifying the type of communication header to be generated, and which part of the communication header created by copying the template header. This is header correction information 304 that specifies how to correct the header.

  Here, the header correction information 304 has 2 bits. For example, “00” is a “0” value, “01” is a “1” value, and “10” is a “2” value. In the case of “11”, it is prohibited. When the sub CPU 106 sets a value “2” in the header correction information 304, the head communication header is edited. When the “1” value is set, the tail header is edited. If the “0” value is set, the intermediate header is edited. Note that how the communication header is corrected using the header correction information of the header correction information 304 will be described later with reference to FIG.

  Next, an example of a communication header generated in the internal memory 113 of the header processing unit 111 in the process of S205 described in FIG. 2 is shown in FIG.

  In step S205, the header duplicating unit 115 of the header processing unit 111 duplicates the template header to the internal memory 113, and a communication header that is a generated intermediate product is denoted by reference numeral 401 in the drawing. The communication header 401 is generated in the internal memory 113 by copying the template header 301 set in the register unit 112 as the copy source and by the number specified by the copy number information 302.

  Thereafter, the header editing unit 116 corrects each of the communication headers 401 that are intermediate products according to the protocol type information 303 and the header correction information 304, and creates completed communication headers 402 to 404. When the number of communication headers is N, the communication header 402 means the first communication header, the communication header 403 means the second to (N-1) th communication headers, and the communication header 404 means the Nth communication header. To do.

  After that, when transmitted to the external device in the process of S206, the completed communication headers 402 to 404 are sequentially transferred to the communication interface 109 by the DMA controller 108 in order to be combined with each payload. The completed transmission packet is transmitted from the communication interface 109 to the external device via the network 110 toward the transmission destination.

  At this time, a communication header of a transmission packet transmitted at the head is a reference numeral 402. A communication header of a transmission packet transmitted in the middle between the head and the tail is reference numeral 403. The communication header of the transmission packet transmitted at the end is reference numeral 404.

  The header editing unit 116 overwrites and rewrites the TCP header part and part of the IP header part of the communication header 401 generated by duplication to complete the communication header in accordance with the header correction information 304. In the case of the communication header 402 of the transmission packet transmitted at the head, the parts indicated by reference numerals 405 and 406 are overwritten and rewritten. And if it is the communication header 403 of the transmission packet transmitted in the middle, it overwrites and rewrites the part shown with the reference numerals 407 and 408. If it is the communication header 404 of the transmission packet transmitted at the end, the part indicated by reference numerals 409 and 410 is overwritten and rewritten.

  Note that the communication header 401 copied in the internal memory 113 may include not only a TCP header and an IP header but also an Ether header. The Ether header is a communication header used in the data link layer defined in the OSI reference model in which the destination MAC address and the source MAC address are described.

  Here, the information to be rewritten in the parts 406, 408, 410 is, for example, flag information of a communication protocol defined as TCP, and is information such as URG, ACK, PSH, RST, SYN, FIN. These are used for establishing and disconnecting communication paths.

  URG is information meaning that it includes data that is preferentially executed in an emergency. ACK is information indicating that the response confirmation number in the TCP header is valid. PSH is information that means that received data is immediately passed to an upper layer. RST is information indicating that the TCP connection is forcibly disconnected. SYN is information indicating a TCP connection establishment request. FIN is information indicating a TCP connection disconnection request.

  For example, if the communication connection using the TCP protocol is started, SYN required to start communication is “1”. Alternatively, if the data transfer using the TCP protocol is started, ACK necessary for starting the data transfer becomes “1”. Alternatively, if the communication disconnection process using the TCP protocol is started, the FIN required to disconnect the communication is “1”. The information on the parts 406, 408, and 410 includes information such as a sequence number and a checksum.

  Where the information of the parts 406, 408, 410 is located in the communication header (402 to 404) is specified by the protocol type information 303 and the header correction information 304. For example, the length of the Ether header and the length of the IP header are specified from information set in the protocol type information 303. Then, the offset value from the head address of the TCP header is specified from the information set in the header correction information 304. Then, it is possible to calculate the information positions of the portions 406, 408, and 410 of the communication headers 402 to 404 from the length of the Ether header, the length of the IP header, and the offset value from the top address of the TCP header.

  The information on the parts 405, 407, and 409 is, for example, flag information of a communication protocol defined as IP, and is information such as DF and MF. DF is information meaning that it should not be divided, and MF is information meaning whether it is the last packet. In addition, the information on the parts 405, 407, and 409 includes information such as packet length information, ID information, and checksum.

  Where the information of the parts 405, 407, and 409 is located in the communication headers 402 to 404 is specified by the protocol type information 303 and the header correction information 304. For example, an IPv4 header or an IPv6 header is specified from information set in the protocol type information 303. Then, the offset value from the head address of the IP header is specified from the information set in the header correction information 304. The positions of the portions 405, 407, and 409 of the communication headers 402 to 404 can be calculated from the fact that the IP header is an IPv4 header and the offset value from the head address of the IP header.

  Next, the flowchart of FIG. 5 shows a detailed operation flow of the processing of S205 by the header correction unit 116 in FIG. Hereinafter, with reference to the same figure, the process which produces | generates the flag part of the communication header by the header edit part 116 is demonstrated.

S204 (terminal):
This terminal indicates the start of this flowchart and corresponds to the process of S204 in FIG. When the process of S204 is completed, the process proceeds to S501.

S501 (processing):
This process shows a process of executing only the minimum unit count-up in the header editing unit (116). After counting up by the minimum unit, the process proceeds to the condition determination process of S502. For example, when shifting from the terminal of S201, the header editing unit 116 executes the first count up and sets the counter value to “1”. Thereafter, when the process shifts from the condition determination in S513, the counter value is set to “2” by counting up. Thereafter, every time the process shifts from the condition determination in S513, the counter is incremented, and the counter value is incremented to “3”, “4”.

  This process has a meaning that the header editing unit 116 counts the number of communication headers in which each communication header 401 in the internal memory 113 is overwritten and corrected according to the header correction information 304.

S502 (condition judgment):
This condition determination indicates that route selection is performed on the condition that the setting content of the header correction information 304 determined by the sub CPU 106 is “10” (= 2). If the setting content is “10” (= 2), the process proceeds to S503, and if not, the condition shifts to S506.

  Here, the setting content of the header correction information 304 is determined by the header editing unit 116.

S503 (condition judgment):
This condition determination symbol indicates that the header editing unit 116 performs route selection on the condition that the value of the internal counter is a value corresponding to the first communication header 402. For example, if the counter value is the minimum initial value ('1'), the process proceeds to S504, and if not, the process proceeds to S505. Here, the association between the counter value and the head communication header is set in advance by the sub CPU 106.

S504 (processing):
This process indicates that correction processing is performed by overwriting the corresponding portion of the communication header 401 in accordance with the header correction information 304 in order to generate the first communication header 402. For example, processing such as setting the ACK or SYN flag, which is an element of flag information of a communication protocol defined as TCP, to “1” is performed.

S505 (processing):
This process indicates that correction processing is performed by overwriting the corresponding portion of the communication header 401 in accordance with the header correction information 304 in order to generate other than the first communication header 402.

  For example, the processing is such that all the flag information of the communication protocol defined as TCP is set to “0”.

S506 (condition judgment):
This condition determination indicates that route selection is performed on the condition that the setting content of the header correction information 304 determined by the sub CPU 106 is “01” (= 1). If the set content is “01” (= 1), the process proceeds to S507, and if not, the condition shifts to S510. Here, the setting content of the header correction information 304 is determined by the header editing unit 116.

S507 (condition judgment):
This condition determination indicates that the header editing unit 116 performs route selection on the condition that the value of the internal counter is a value corresponding to the last communication header 404. For example, if the counter value is the set maximum value, the process proceeds to S508, and if not, the process proceeds to S509. Here, the association between the counter value and the last communication header is set in advance by the sub CPU 106.

S508 (processing):
This process is a process of overwriting the corresponding portion of the communication header 401 according to the header correction information 304 to correct the last communication header 404. For example, processing such as setting the FIN flag, which is flag information of a communication protocol defined as TCP, to “1” is performed. Similarly, processing such as setting the MF flag, which is flag information of a communication protocol defined as IP, to “0” is performed.

S509 (processing):
This process indicates that correction processing is performed by overwriting the corresponding portion of the communication header 401 according to the header correction information 304 in order to generate other than the last communication header 404. For example, the processing is such that all the flag information of the communication protocol defined as TCP is set to “0”. Similarly, processing such as setting an MF (More Fragment) flag, which is flag information of a communication protocol defined as IP, to “1” is performed.

S510 (condition judgment):
This condition determination indicates that the header editing unit 116 performs route selection on the condition that the internal counter value is a value corresponding to the intermediate communication header 403. For example, if the counter value is within a set range, the process proceeds to S511, and if not, the process proceeds to S512. Here, the association between the counter value and the intermediate communication header group is set in advance by the sub CPU 106.

S511 (processing):
This process indicates that correction processing is performed by overwriting the corresponding portion of the communication header 401 in accordance with the header correction information 304 in order to generate the intermediate communication header 403. For example, processing such as setting ACK and FIN flag, which are flag information of a communication protocol defined as TCP, to “0” is performed.

S512 (processing):
This process indicates that correction processing is performed by overwriting the corresponding portion of the communication header 401 in accordance with the header correction information 304 in order to generate other than the intermediate communication header 402. For example, the processing is such that all the flag information of the communication protocol defined as TCP is set to “0”.

S513 (condition judgment):
This condition determination indicates that route selection is performed on the condition that the value of the counter counted up in S501 is the set maximum value (final packet). If the counter value is the set maximum value, the process proceeds to S514. Otherwise, the process proceeds to S501. Here, it is determined whether or not the communication header generated based on the counter value has reached a predetermined number. The association between the counter value and the total number of communication headers is set in advance by the sub CPU 106.

S514 (processing):
This process indicates a process in which the header editing unit 116 resets the counter value to the initial value. For example, the initial value is set by setting the counter value to “0”.

S206 (terminal):
This terminal indicates the end of this flowchart. It shows that the processing shifts to S206 in FIG. 2 when the processing of S514 is completed.

  As described above, according to the present embodiment, when a communication device divides transmission information into a plurality of transmission packets via a network and transfers them to an external device, a plurality of communication headers that are different in one setting are automatically used. Can be generated. As a result, it is possible to divide the transmission information and generate a plurality of transmission packets at a high speed by one setting, and it is possible to improve the transmission processing capability of the communication apparatus as compared with the past.

DESCRIPTION OF SYMBOLS 100 ... Communication apparatus, 102 ... Main CPU, 103 ... Main memory, 104 ... Sub system, 106 ... Sub CPU, 107 ... Local memory, 108 ... DMA controller, 109 ... Communication interface, 111 ... Header processing part, 112 ... Register part 113 ... Internal memory, 114 ... Core unit, 115 ... Header replication unit, 116 ... Header editing unit, 117 ... Switch, 118 ... Core control unit

Claims (9)

A communication device for connecting an information processing device to a network,
Determining means for determining N representing the number of packets from the size of information to be transferred when a transfer request to the network is received from the information processing apparatus;
Generating means for generating N communication headers based on a model communication header corresponding to the transfer request;
Dividing means for dividing the information to be transferred into N payloads;
Transmitting means for transmitting N packets, which are a combination of the communication header generated by the generating means and the payload, to the network;
Have
When generating the N communication headers, the generation means generates the first, second to N−1th, and Nth of the N communication headers according to an instruction included in the transfer request. A communication apparatus, wherein at least one communication header of the communication header is generated by editing the communication header of the template.   The communication apparatus according to claim 1, further comprising setting means for setting contents of editing by the generating means.   The communication apparatus according to claim 1, wherein the instruction included in the transfer request is flag information of a TCP header of a communication header.   The communication apparatus according to claim 1, wherein the instruction included in the transfer request is flag information of an IP header of a communication header. The generation means, as a content of editing the flag information of the TCP header for a single execution instruction,
A process for setting a flag on any of the elements of the flag information of the first communication header and not setting a flag on any of the elements of the flag information of the non-first communication header,
A process that sets a flag on any of the elements of the flag information of the Nth communication header and does not set a flag on any of the elements of the flag information of the non-Nth communication header,
The communication apparatus according to claim 3, wherein any one of the processes in which a flag is not set on any of the elements of the flag information of all communication headers is executed. The generation means, as a content of editing IP header flag information for a single execution instruction,
A process of setting a flag of the MF (More Fragment) element of the flag information of the first to (N-1) th communication headers, and not setting a flag only of the MF element of the flag information of the Nth communication header;
The communication apparatus according to claim 4, wherein any one of the processes in which no flag is set on any of the elements of the flag information of all communication headers is executed.   The communication apparatus according to any one of claims 1 to 6, wherein the communication header includes an Ether header, an IPv4 header or an IPv6 header, a TCP header, or a UDP header. A method for controlling a communication device for connecting an information processing device to a network,
A determination step for determining N representing the number of packets from the size of information to be transferred when the determination means receives a transfer request from the information processing apparatus to the network;
A generating step for generating N communication headers based on a model communication header corresponding to the transfer request;
A dividing step of dividing the information to be transferred into N payloads;
A transmission step of transmitting N packets, which are a combination of the communication header generated by the generation step and the payload, to the network;
Have
In the generation step, when generating the N communication headers, the first, second to (N−1) th, and Nth of the N communication headers according to an instruction included in the transfer request. A communication apparatus control method, comprising: generating at least one communication header of the communication header by editing the communication header of the template.   An information processing apparatus comprising the communication apparatus according to claim 1.

Images