JPH0215747A - Functional distributed communication processing device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] [Industrial application field]

The present invention relates to a communication processing device that processes packets according to a hierarchical communication protocol.

[Conventional technology]

The conventional device is JP-A-61-264! '145, an example of which is shown in FIG. 16th
In the figure, 161 is a memory used in common by the layer 3 processing unit 162 and the layer 2 processing unit 163 to store data in the sending/receiving bucket 1, etc., and 162 is a memory for storing sending/receiving packets,
163 is a layer 3 processing unit that performs processing, and 163 is a layer 2 processing unit that performs layer 2 processing of transmitted and received frames;
4 is a bus connecting the memory 161, layer 3 processing unit 162, and layer 2 processing unit 163; 165 is a line control unit that controls layer 1 of transmitted and received frames;
are multiple communication lines. Although the above cited publicly known example also describes layers 4 and above, layers 3 and below, which are necessary for the following explanation, will be described here. Furthermore, in the publicly known example cited above, the device that performs layer 1 processing is also configured to be connected to the bus 164, but layer 1 specifies the physical conditions such as the shape of the connector and the voltage of the signal line. Therefore, the configuration shown in FIG. 16 should be adopted since no processing by a processor is required, and data that can be stored in the memory 161 cannot be extracted unless layer 2 processing such as zero deletion and flag detection is performed. . Although the above-cited publicly known example does not specify the operation at the time of packet transmission, at the time of transmission, the layer 3 processing unit 162 performs layer 3 processing on packet 1- stored in the memory 161, and then the layer 2 processing unit A packet transmission request is issued to 163. Layer 2 processing unit 162 that received the packet transmission request
performs layer 2 processing on the frame requested to be transmitted, and after completion, sends the frame to the line control unit 165. The line control unit 165 controls layer 1 of the frame and sends it to the communication line 166.

[Problem to be solved by the invention]

The above conventional technology has a problem in that when the upper layer processing unit (layer 3 processing unit) requests transmission of consecutive packets, it is not possible to continuously deliver packets to the lower layer. This will be explained using the example shown in FIG. In FIG. 17, the horizontal direction represents the passage of time. 171 is the layer 3 processing unit] 62; 172 is the layer 2 processing unit 1;
63, 173 represents frame transmission on the communication line 166, and 174 represents time. At time T, the layer 3 processing unit 162 processes packet 1.2.3.
The layer 2 processing unit 163 is requested to continuously transmit . Layer 2 processing unit 163 receives time T. First, layer 2 protocol processing for frame 1 is started. The layer 2 processing unit finishes the layer 2 protocol 1 processing of frame 1 at time T, and then starts the process of sending frame 1 to the communication line 166 via the line control unit 165 (line sending process). . Accordingly, at time T, frame 1 is sent from the communication line 166.
transmission begins. However, from the time the layer 2 processing unit starts line transmission processing to the time when the line is actually transmitted from the communication line 166, the layer 2 processing unit 163 and the line control unit 165
Since the delay that occurs is short, it is ignored here. The layer 2 processing unit 163 ends the line sending process when the sending of the frame 1 from the communication line 166 ends at time T l ], and then starts the layer 2 protocol process of the frame 2. Similarly, at time T1G, layer 2 protocol processing for frame 2 is finished, line transmission processing is started,
At time T2□, line sending processing for frame 2 ends, layer 2 protocol processing for frame 3 starts, at time T2□, layer 2 protocol processing for frame tz 3 ends, line sending processing starts, time T. The line sending process for frame 3 ends. In this case, from the communication line 166, frame 1 is transmitted from time T5 to T□□, frame 2 is transmitted from time T16 to T2□, and frame 2 is transmitted from time T27 to T. Frame 3 is transmitted from time T□ to Ti6, time T2. There is a free time between T2□ and T2□. In this way, layer 3 processing unit 1
62 requests that frames 1, 2, and 3 be sent out consecutively, but frames cannot be sent out consecutively from the communication line 166. An object of the present invention is to enable continuous delivery of packets to a lower layer when an upper layer requests continuous packet transmission. [Means for Solving the Problem] The above object is to provide a processor for processing a protocol (protocol processing processor), a processor for processing an interface with upper and lower layers (interface processing processor), and a processor for processing the interface with the upper and lower layers. A means (protocol processing request storage means) for sequentially storing protocol processing requests from the processing processor to the protocol processing processor is provided, and when the upper layer requests continuous transmission of multiple packets, the protocol of one packet is provided. This is achieved by the interface processing processor storing the protocol processing request for the next packet in the protocol processing request storage means before completing the processing and interface processing and completing the handover to the lower layer. Further, the protocol processing processor retrieves the protocol processing requests from the protocol processing request storage means according to the type of the protocol processing requests, regardless of the order in which the protocol processing requests are stored in the protocol processing request storage means. By enabling processing, it is possible to prevent only transmission requests from being processed with priority.

[Effect]

When the upper layer requests the interface processor to continuously transmit a plurality of packets (assuming packets 1.2.3), the interface processor stores the protocol processing request for packet 1.2.3 in the protocol processing request storage. Store in a means. The protocol processing processor takes out the protocol processing request of packet 1 from the protocol processing request storage means and starts protocol processing of packet 1. When the protocol processing processor finishes the protocol processing of packet 1, it notifies the interface processing processor of the completion of the protocol processing of packet 1, and retrieves the protocol processing request of packet 2 from the protocol processing request storage means and performs the protocol processing of packet 2. Start. The protocol processing processor similarly performs the protocol processing of packet 2.3 and notifies the completion thereof. The interface processing processor that received the packet 1 protocol processing completion notification processes packet 1.
The transmission process to the lower layer is started. By designing the protocol processing time to be shorter than the sending time to the lower layer, when the interface processor finishes sending the packet 1 to the lower layer, the protocol processing of the packet 2 has already been completed. Therefore, the interface processor can immediately start sending the packet 2 to the lower layer. After completing the process of sending packet 2 to the lower layer, the interface processor similarly immediately sends packet 3 to the lower layer.
The sending process to the lower layer can be started. As described above, according to the present invention, it is possible to continuously send packets to lower layers. Furthermore, for example, when the upper layer requests continuous transmission of 10 packets, the protocol processing request for transmission of 10 packets is stored in the protocol processing request storage means. Assuming that the device receives one packet immediately thereafter, the interface processor stores the protocol processing request of the received packet in the protocol processing request storage means. If the protocol processing processor retrieves and processes the protocol processing requests in the same order as the order stored in the protocol processing request storage means, the protocol processing of the received packets is entirely the same as the protocol processing of the 10 transmitted packets. This will be done after it is finished. This gives extreme priority to transmission processing, which causes problems such as overflow of the reception buffer. Therefore, in the present invention, the protocol processing processor retrieves the protocol processing request from the protocol processing request storage means according to the type of the protocol processing request, regardless of the order in which the protocol processing requests are stored in the protocol processing request storage means. processing. For example, the protocol processing processor accepts and executes a protocol processing request for one transmitted packet, and then accepts and executes a protocol processing request for a received packet, thereby managing processing priorities. becomes possible, and the above problems are solved. (Example) Hereinafter, an example of the present invention will be described using the drawings. The processing layer in this example is layer 2, and the protocol of layer 2 is LI DLC (Illight
evel Data Link Control). FIG. 1 is a diagram showing a mocha composition according to an embodiment of the present invention. 1
is a memory that stores transmitted and received packets, etc. and is commonly used by the layer 3 processing unit 2 and layer 2 processing unit 3, 2 is a layer 3 processing unit that performs layer 3 processing of transmitted and received packets, and 3 is a memory that stores transmitted and received frames. 4 is a layer 2 processing unit that performs processing other than real time among the layer 2 processing, and 4 is a bus that connects the memory 1, the layer 3 processing unit 2, and the layer 2 processing unit 3. In addition to this, the bus 4 may also be connected to an upper layer processing section of layer 4 or higher, a channel control section that performs interface processing with a computer, and the like. 5 is a line control unit that performs real-time processing such as flag addition/removal, zero insertion/deletion, FCS check, etc. and layer 1 control among the layer 2 processing of transmitted and received frames; 6, 7.8
is a communication line. 9 is a protocol processing unit that performs non-real-time protocol processing such as layer 2 sequence number control and state transition, and 10 is an interface processing unit that performs interface processing with the memory 1, layer 3 processing unit 2, and line control unit 5. It is. Protocol processing section 9 and interface processing section 1
0 are each made up of one or more processors and the hardware controlled by them. 11 is a main control unit that includes a processor and controls the processing of the interface processing unit 10. 12 is a reception management table for managing reception processing, and 13 is a transmission management table for managing transmission processing. 14.1
5.16 is a transmission waiting queue used to control transmission processing. The transmission queue is F r FO (First In First Out).
s t 0ut), and is provided for each line. A memory transfer control section 17 is hardware that controls data transfer between the memory 1 and the transmission/reception buffers 18 to 23 according to instructions from the main control section. 18.19.2
0 is a receive buffer that temporarily stores received packets,
21, 22, and 23 are transmission buffers that temporarily store transmission packets. Transmission and reception buffers are provided for each line. 24 is a PRQa in which the interface processing unit 10 stores a protocol processing request for a transmitted frame, and 25 is a PRQb in which a protocol processing request for a received frame is stored.
26 is a PRQc that stores protocol processing requests for the S frame 11 and the U frame. PRQa24, P
RQb25 and PRQc26 are FIFO. In the description of this embodiment, only normal transmission and reception processing of E-frames and RR frames is performed, so a protocol processing request for a received RR frame is stored in PRQc. A PEQ 27 stores the result of protocol processing performed by the protocol processing unit 9. FIG. 12 is a diagram showing the contents of the reception management table 12. In FIG. 12, each row corresponds to a line number. 12
o1 is a column that stores the value 1 when packet reception processing is currently being performed on the corresponding line, and the value O when 1 is not being performed, and 1202 is the value when the protocol processing in the protocol processing unit 9 has already been completed. 1, when you have not finished l1f
This is a column that stores lO, and 1203 is the reception bucket 1.
This column stores the value 1 when the transfer of - to the memory has already been completed, and the value O when it has not finished. FIG. 13 is a diagram showing the contents of the transmission management table 13. In FIG. 13, each row corresponds to a line number. 13
01 is a column that stores a value of 1 when packet transmission processing is currently being performed on the corresponding line, and a value of 0 when no packet transmission processing is currently being performed. The above reception management table 12 and transmission management table 1
The initial values of all 3 are O. A protocol processing request from the interface processing unit 10 to the protocol processing unit 9 is sent as a PRQ transaction in the form of PRQa24. PRQb25゜P RQ c 26
is stored in FIG. 14 is a diagram showing the contents of a PRQ transaction. 1401 is the interface processing unit 1
Type of request issued by 0 (Pro1 processing request for transmitted I frame, protocol processing request for received E frame, received R
This is an area for storing a code meaning (protocol processing request, etc. of R frame), and 1402 is an area for storing the
, is an area for storing the contents of the C field, and is 140
3 is an area for storing the corresponding line number, 140
4 is an area in memory 1 for storing the address of the packet. The processing result after the protocol processing in the protocol processing section 9 is completed is stored in the PEQ 27 in the form of a PEQ transaction. FIG. 15 is a diagram showing the contents of a PEQ transaction. 15o1 is an area for storing a code indicating the type of report, and 1502
are areas that store the contents of the A and C fields of the frame, 1503 is an area that stores the corresponding line number, 1504 is an area that stores the address of the packet in memory 1, and 1505 is the area that stores the contents of the received frame. This area stores a value of 1 when it is necessary to transmit an RR frame as a result of the protocol processing. The main control unit 11 sends the transmission request to the transmission waiting queue 14.15.
16 in the form of a queue transaction waiting to be sent. FIG. 20 is a diagram showing the contents of a transmission queue transaction. 2001 is an area for storing frame types waiting to be transmitted. 2002 is an area for storing the contents of the A and C fields of the frame, and 2003 is an area for storing the address of the packet in the memory 1. Next, specific processing in the case of frame transmission and reception will be explained. 2 to 10 represent the processing performed by the main control unit 11, and the first
FIG. 1 shows the processing performed by the protocol processing section 9. 2nd ~
Figure 11 is shown according to the PAD diagram technique. First, the case of receiving an I frame will be explained. If there is no process to be performed, the main control unit 11 monitors in an infinite loop whether there is a request to start processing from the main control unit (second
Figures 200, 201). When there is a received frame in the reception buffer, the hardware within the interface processing section 10 issues a processing start request to the main control section 11. Upon receiving the process start request, the main control unit 11 determines the content of the process start request in process 202 and proceeds to process 203 (see FIG. 3). The main control unit 11 identifies the line number that received the frame in process 301 from the contents of the process start request, and in process 30
2, column 1 of "Reception processing" in reception management table 12
201 is set to 1, in process 303 the header part of the received frame stored in the receive buffer is extracted, the type of received frame is determined, (2) frame reception is recognized, and in process 304 a PRQ transaction is created and placed in PRQb. In the PRQ transaction created in process 304, the "request type J 1401" contains a code that means "received I frame protocol processing request," and the "A. C field J 1402 contains the A and C fields in the header of the received frame. , "Line number-J1403 stores the number of the line that received the frame."Memory address J1404 is blank.Next, in process 305, the ■ field of the received I frame in the receive buffer is stored in memory 1. The main control unit 11 instructs the memory transfer control unit 17 to transfer the address.
It is delivered to the memory transfer control unit 17 along with an instruction to . In response to this instruction, the memory transfer control section 17 starts transferring the I field to the memory 1 independently of the main control section 11. After the process 305, the main control unit 11 returns to the loop of processes 200 and 201. The protocol processing unit 9 monitors PRQs a, b, and c in an infinite loop to see if there are any transactions, as shown in FIG. 11 1100 to 1103. PRQb by process 304
When a PRQ transaction is entered in , the processing of the protocol processing unit 9 proceeds from 1103 to 1120 . 1120
At step 1121, the PRQ transaction is taken out from PRQb, and its content is used to perform layer 2 protocol processing on the received frame.At step 1121, a PEQ transaction is created from the processing result and is placed in PEQ. "Report type J" of PEQ transaction created in 1121 1501
1503 is a code that means "end of protocol processing of received I frame", "line number" 1503 is the number of the line that received the received frame, and the result of this protocol processing is R.
If you decide to send an R frame, set 1 to rRR transmission required J 1505, rA, C field J 150
2, stores the A and C fields of the RR frame to be transmitted. "Memory address J 1504 is blank. Next, the protocol processing unit 9 checks whether there is a transaction among them in the order of PRQc, a, and b, and if so, proceeds to the corresponding process 1104.1112.1120. , if not, return to loop 1100-11o3 (1122
-1127), When there is a PEQ transaction in PEQ, the hardware in the interface processing unit 10 issues a process start request to the main control unit 11, and the main control unit 11 accordingly proceeds from process 200.201 to process 207 in FIG. (Go to Figure 7). The main control unit 11 extracts the PEQ l-run transaction from the PEQ, identifies the line number from the line number J 1503 in the PEQ transaction in process 701, and identifies the line number from the report type J 1501 in process 702. Approve “End” and proceed to process 704 (
(Go to Figure 9). In process 901, the "protocol processing completed" column 1202 of the reception management table 12 is set to 1, and in process 902, the PEQ
Determine rRR transmission requirement J 1505 in the transaction. If “rRR transmission required” 1505 is 1, process 903
Column 13 of "Transmission Processing" in Transmission Management Table 13
Determine 01. A value of 1 in the "transmission processing in progress" column 1301 means that a frame is currently being transmitted from that line, so it is necessary to wait before transmitting other frames. In this case, the process advances to step 904, where the transmission queue transaction is stored in the transmission queue corresponding to the transmission line number. The "sent waiting frame type J 20" of the sending waiting queue transaction created here
01 is a code that means waiting for RR frame transmission. The rA, C field J 2002 stores the contents of the rA, C field J 1502 of the PEQ transaction. In process 903, if the "transmission processing" column 1301 is 0, the process advances to process 9905, and rA of the PEQ transaction is stored in the transmission buffer corresponding to the number of the transmission line.
The contents of "C field" 1502 are stored (RR sending processing). When data enters the transmission buffer, the hardware of the interface processing unit 10 transfers the data to the line control unit 5.
Start transferring to. After performing processing 904 or 905, the main control unit 11 proceeds to processing 906 and determines the “memory transfer complete” column 1203 of the reception management table 12. If the "memory transfer completed" column 1203 is 1, it means that the memory transfer instructed in process 304 has been completed incorrectly. In this case, in process 907, the layer 3 processing unit 2 completes the layer 2 reception process. In step 908, all columns of the row corresponding to the line number in the reception management table 12 are initialized to O, and the process returns to the loop of steps 200 and 201. In process 906, column 1203 of “memory transfer complete” is set to 0.
In this case, the process directly returns to the loop of processes 200 and 201. When the memory transfer control unit 17 completes the memory transfer instructed in process 304 in FIG. 3, it issues a process start request to the main control unit 11. As a result, the main control unit 11 performs the process 200.2 in FIG.
The process proceeds from step 01 to process 205 (to FIG. 5). In process 501, the line number is identified, in process 502, the value 1 is stored in the “memory transfer complete” column 1203 of the reception management table 12, and in process 503, the value 1 is stored in the “memory transfer complete” column 1203 of the reception management table 12.
The column 1202 of "Protocol processing completed" is determined. "
If the "protocol processing completed" column 1202 is 1, the same processes 504 and 505 as processes 907 and 908 are performed, and the process returns to the loop of processes 200 and 201. If the "protocol processing completed" column 1202 is 0 in process 503, the process returns to the loop of processes 200 and 201. In the case of receiving an RR frame, after determining the received frame type in process 303 in FIG. 3, the process proceeds to process 306, where a PRQ transaction is created and placed in PRQ c. The PRQ created here is request type 140.
1 is the same as the one created in process 304 except that it stores the protocol processing request J of the received RR frame. After process 306, the process returns to the loop of processes 200 and 201. PRQc in figure 11 processing 1101
It is determined that there is a transaction in
to 1111 (same as processes 1120 to 1127), and returns to the loop of processes 1100 to 1103. Thereafter, the main control unit 11 advances from the loop of processing 200.201 in FIG. 2 to processing 202, 207, 701, 702.705 in FIG. return. Next, the case of I frame transmission will be explained. When there is an I frame transmission request, the layer 3 processing unit 2 notifies the main control unit 11 of the transmission request, the line number, and the address in the memory 1 of the data to be transmitted. Here, the layer 3 processing unit 2 may issue a plurality of consecutive E-frame transmission requests. The main control unit 11 processes 200 and 201 in FIG.
, 202 to process 204 (see FIG. 4). In process 401 of FIG. 4, the transmission line number is identified, and in processes 402 and 403, a PRQ transaction is created for each of all Eframes requested to be transmitted by the layer 3 processing unit and placed in PRQa. Request type 14 of the PRQ transaction created here
01 is “transmission frame protocol processing request J,”
[The number of the line to be sent is stored in line number j, and the address in memory 1 of the data to be sent is stored in ``memory address.''"The A and C fields J 1402 are blank. When a PRQ transaction is entered in PRQ a, the protocol processing unit 9 performs processing 1100.1102 in FIG.
．． Proceed to 1112, PRQ from PRQ a at 1112
The transaction is extracted, the content is used to perform layer 2 protocol processing on the sending Eframe, a PEQ transaction is created from the processing result in 1113, and the PEQ transaction is sent to the PE
Put it in Q. In the PEQ transaction created in step 1113, the report type J 1501 contains a code that means [End of protocol processing for sending e-frames], and the rA, C fields J 1502 send the A and C fields of the frame. The number J 1503 stores the number of the line for transmitting the frame, and the [memory address] 1504 stores the address in the memory 1 of the ■ field of the I frame to be transmitted. rRR transmission required J 1505 is blank. The protocol processing unit 9 next checks whether there is a transaction among PRQc, b, and a in this order, and if so, proceeds to the corresponding process 1104.1120.1112, otherwise returns to the loop 1100 to 1103 ( 1114
~1119). When a PEQ transaction enters PEQ, main control unit 1
1 is the process 200.201.20 in Figure 2, similar to the reception process.
2 to process 207 (to FIG. 7), process 701, 7
Proceed to 02. In process 702, it is recognized from the [Report Type J 1501 in the PEQ transaction that ``Protocol Processing End J of Transmission Eframe'' is recognized, and the process proceeds to Process 703 (to Fig. 8). Column 1301 of “Transmission processing in progress” is determined, and if it is 1, processing 802
Create a send queue transaction and store it in the send queue corresponding to the line to be sent. The "frame type waiting to send" 2201 of the waiting queue transaction to be created here indicates a code that means waiting for I frame transmission, and the rA, C field J 2202 contains the contents of the rA, C field J 1502 of the PEQ transaction. Memory address J 2203 stores the contents of memory address J 1504 of the PEQ transaction.
Return to loop 201. If the column 1301 of "Processing to send" is 0 in process 801, the value 1 is stored in the column 1301 of "Processing to send" in process 804, and the PEQ transaction is stored in the send buffer corresponding to the number of the line to be sent in process 805. The contents of the rA and C fields J 1502 are stored, and in process 806 the memory transfer control unit 17 is instructed to transfer one field of the T frame to be transmitted from the memory 1 to the transmission buffer, and the loop of processes 200 and 201 is executed. Return to Through process 806, the memory transfer control unit 15 starts memory transfer. When the data enters the transmission buffer, the hardware of the interface processing section 10 starts transferring the data to the line control section 5. When the line control unit 5 receives the transmission data from the transmission buffer, it performs real-time processing on layer 2 and control on layer 1, and sends the data to the communication line. When the transmission of one frame is finished, the line control section 5 notifies the interface processing section 10, and the hardware of the interface processing section 10 that has received the notification issues a processing start request to the main control section 11. The main control unit 11 processes 200.201.202.2 in FIG.
06 (to Fig. 6), the line number is identified in process 601, the row corresponding to the line number being processed in the transmission management table is initialized to 0 in process 602, and layer 3 is initialized in process 603.
Notifies processing unit 2 of completion of layer 2 transmission processing. That's 1
The frame transmission process ends, and then in process 604 it is determined whether there is a transaction in the transmission queue corresponding to the same line. If there is, in process 605, the transaction in the queue waiting for transmission is extracted and the ``sent waiting frame type J2201'' is determined.If the determination result is 1 frame, E-frame sending processing is performed in process 606.■ Frame sending in process 606 The processing is the same as processing 804, 805, and 806 in Fig. 8. If the determination result of processing 605 is an RR frame, processing 607 is performed.
RR frame transmission processing is performed. The RR frame sending process in process 607 is the same as process 905 in FIG. FIG. 18 is a diagram showing the time course of the transmission process. In Figure 18, the horizontal direction shows the passage of time. 181 is the processing in the layer 3 processing unit 2, 182 is the processing in the protocol processing unit 9, 183 is the processing in the main control unit 11 in the interface processing unit 10, excluding the idling processes 200 and 201,
184 is the processing of the interface processing unit 10 excluding the processing of the main control unit 11 (hereinafter referred to as the processing of the interface processing unit 10); 185 is the packet transmission on the communication devices a6 to a8; 186 represents the time. Time T. Then, the layer 3 processing unit 2 requests the main control unit 11 to continuously transmit frames A of 1.2°3. Main control unit 11
is time T in processing 402 and 403. In frame 1.2.3
The PRQ I-ranzaktion, which is a protocol processing request, is stored in the PRQ a 24. The protocol processing unit 9 extracts the PRQI-Run sequence of frame 1 from the beginning of PRQa24 at time T, and
Layer 2 protocol processing for frame 1 starts, and time T
6, and the processing result is stored in PEQ27. The protocol processing unit 9 executes the next PR in PRQa at time T6.
Take out Q transaction and layer 2 of frame 2
Protocol 1 - Begins protocol processing. The main control unit 11 at time T6
Then, the processing is started because the PEQ transaction of the processing result of frame 1 is contained in PEQ, and the frame sending processing 804, 805, and 806 is performed, and then the processing 200
．． Return to loop 201. At time T, the interface processing section 10 starts transmitting frame 1 to the communication line via the line control section 5. Time T1
At □, the protocol processing unit 9 finishes the layer 2 protocol processing for frame 2 and stores the PEQ transaction in PEQ, and then the main control unit 11 performs frame sending processing 804.805.806 at time T□, and then The process returns to the loop of processes 200 and 201. The line control unit 5
When transmission of frame 1 is finished at step 13, the interface processing unit 10 is notified, and the main control unit 11 starts processing based on this notification, and processes 601.602.604.605.60
6 and returns to the loop of sono post-processing 200 and 201. Thereafter, the interface processing section 10 starts transmitting frame 2 to the communication line via the line control section 5. Similarly, at time T, the protocol processing section 9 ends the protocol processing of frame 3, and at time T2, the line control section 5 indicates that the sending of frame 2 has ended and the interface processing section 10
The interface processing unit 10 starts sending frame 3 to the communication line via the line control unit 5 at time T21, and at time T2□, the line control unit 5 notifies the interface processing unit 10 that the sending of frame 3 has ended. do. When the layer 3 processing unit 2 requests the continuous transmission of frames, in the conventional technology, an idle time occurs between frames 11 to be sent as shown at 173 in FIG. As shown, frames can be transmitted continuously with almost no idle time between transmitted frames. Next, PRQ is PRQa24, PRQ
b25. There are not three types of PRQc27, but the 19th.
There is only one PRQ191 shown in the figure, and PRQ191
is F I F O (First In First Ou)
Consider the case t). Assuming that the layer 3 processing unit 2 requests continuous transmission of 10 Eframes at a certain time, the main control unit 11 immediately creates 10 PRQ transactions 192 to 195 and stores them in the PRQ 191. Immediately after receiving one I frame, the main control unit creates a PRQ transaction 196 and stores it in PRQ 191. In this case, the protocol processing of the PRQ transaction 196 in the protocol processing unit 9 consists of 10 PRQ transactions 1
This will not be performed until after the protocol processing steps 92 to 195 have been completed. According to this, the transmission processing is given extreme priority, and errors such as overflow of the reception buffer occur. In this embodiment, as shown in FIG.
, PRQb, and PRQc are provided, and each FIFO stores PRQ transactions for a protocol processing request for I frame transmission, (2) a protocol processing request for frame reception, and a protocol processing request for S and U frame reception. In addition, in this embodiment, as shown in FIG.
After RQ transaction processing, PRQc, PRQb
, PRQa, and after the PRQ transaction processing in PRQb, P
RQ c , PRQ a . By checking the presence or absence of PRQ transactions in the order of PRQb, priority is given to the processing of PRQ transactions in PRQc, and the processing priorities of (1) frame transmission and reception are managed to be the same. In this embodiment, only one PEQ is provided, but PEQ
It is also possible to provide multiple. In this case, for example, PEQs are provided for normal and emergency processing results, and the main control unit 11 processes transactions in the emergency PEQ with priority, thereby making it possible to perform efficient processing. . In this embodiment, as shown in processes 402 and 403, if there are multiple I frame transmission requests from the layer 3 processing unit, the main control unit 11 sends all PRQ)-ranzagions of the multiple I frames to PRQ a. The protocol processing request may be made at any timing before the completion of handing over one frame to a lower layer, such as when the protocol processing for one frame is completed. In this case, although it is possible that there may be a gap between transmission frames, the gap will be shorter than in the prior art. This allows more flexible processing, such as prioritizing important processing over protocol processing requests. In this embodiment, the protocol processing request storage means includes three FIFOs (PRQa, PRQb, PRQ) corresponding to the request type.
Although the configuration is described as c), it is also possible to divide the request types more finely and configure it with a larger number of FIFOs. This allows for more detailed priority management. It is also possible to divide the request types into larger categories and configure the FIFO with a smaller number of FIFOs. According to this, it becomes possible to configure the system with a smaller amount of hardware. Also FIFO
It is also possible to configure one queue instead of a single queue, and allow the protocol processing unit to freely read out the types of transactions stored in the queue. According to this, detailed priority management is possible with a small amount of hardware. [Effects of the Invention] According to the present invention, when there is a request from the upper layer processing unit to send successive buckets i-, it is possible to continuously send the packets to the lower layer processing unit. There is.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram showing the configuration of an embodiment of the present invention, and FIGS. 2 to 10 are processing flows (P
11 is a processing flow (PAD) diagram showing the operation of the protocol processing unit 9, FIG. 12 is an explanatory diagram showing the contents of the reception management table 12, and FIG. FIG. 14 is an explanatory diagram showing the contents of a PRQ transaction. FIG. 15 is an explanatory diagram showing the contents of a PEQ transaction. FIG. 16 is a block diagram showing an example of the configuration of a conventional technique. 18 is an explanatory diagram showing the time course of processing in the conventional technology, FIG. 18 is an explanatory diagram showing the time course of processing in the embodiment of the present invention, and FIG. 19 is an example of the content of PRQ when there is only one PRQ. FIG. 20 is an explanatory diagram showing the contents of a transmission queue transaction. Explanation of symbols 1...Memory, 2...Layer 3 processing unit, 3...Layer 2
Processing unit, 4... Bus, 5... Line control unit, 6.7.8.
- Communication line, 9...Protocol processing unit, 1o...Interface processing unit, 11...Main control unit, 12...Reception management table, 13...Transmission management table, 14.15.1
6... Transmission waiting queue, 17... Memory transfer control unit, 1
8.19.20...Receive buffer, 21.22.23...
・Transmission buffer, 24.25.26...PRQ, 27.
・PEQ. λe Shift No. 37, 5' area, 7th ρ, 74th, PF 6) To the run 2nd 3rd in) Ivy crossbow value 6 notoru] 1st Z mouth/3 图躸庸庸开傛27゛ lure J and/L? ρ/

Claims (1)

[Claims] 1. In a communication processing device that processes each communication layer using a plurality of processors, one or more processors (protocol processing processor) that performs protocol processing of a certain layer, and an interface with upper and lower layers. Processing 1
Two or more processors (interface processing processors)
, and means (protocol processing request storage means) for storing a protocol processing request from the interface processing processor to the protocol processing processor. 2. In the functionally distributed communication processing device according to claim 1, when there is a request to transmit consecutive packets from an upper layer, the protocol processing and interface processing of one packet are finished, and the delivery to the lower layer is performed. 2. A functionally distributed communication processing device, wherein the interface processing processor stores a protocol processing request for the next packet in the protocol processing request storage means before completion. 3. The protocol processing processor, regardless of the order in which the protocol processing requests are stored in the protocol processing request storage means by the interface processing processor,
2. The distributed function communication processing device according to claim 1, wherein the request can be retrieved from the protocol processing request storage means depending on the type of the request. 4. The functionally distributed communication processing device according to claim 1, wherein a plurality of the protocol processing request storage means are provided depending on the type of the protocol processing request.