JPS59168537A - Control system of communication controller - Google Patents

Abstract

PURPOSE:To shorten the initial loading time of a program by deciding a positive answer if no negative answer is received in case the program is loaded into a processor from a CPU. CONSTITUTION:A CPU transmits a packet PKT(0) to all processors PU via a data channel device and a bus control part and then discontinues transmission of the next packet PKT(1) for a fixed time interval (t). If no signal is received from any PU for the interval (t), it is decided that the PKT(0) is received correctly by all PUs. Then the next PKT(1) is transmitted. The PU that received a packet in a wrong way transmits a negative answer NAK to the CPU. The CPU transmits the PKT again to all PUs when it receives the answer NAK within the interval (t) after transmission of the PKT.

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention relates to a communication control device for a data communication system.
and a data channel device DC that connects the CPU and CCP.
This shows a data communication system consisting of H. CCP is a plurality of processing units PUi that control communication lines.
(i-0, 1,..., n-1), a data bus BUS that connects PUi, and an interface control unit BC that controls the connection between this BUS and DCH, and further includes a processing unit PUi Correspondingly, it has a line control unit LCi and a storage device Mi. FIGS. 2 and 3 each show a prior art method for loading a program from a CPU to a CCP.

Hereinafter, the problems of the prior art will be explained using FIGS. 1.2.3.

The program to be loaded from the CPU to the PUi is B.
It is necessary to divide the data into appropriate lengths and load them in consideration of the probability of error occurrence on the US.

If one unit of this divided program is called a packet PKT, the CPU that sends the PKT and the P that receives the PKT.
PKT error detection control and retransmission control when an error occurs is required between Ui and Ui. Also, to reduce loading time, one PKT sent by the CPU can be
A bulk loading scheme received at the U is generally adopted.

First, FIG. 2 will be explained.

In Figure 2, the loaded program is C
The signal is divided into m PKTs in the PU, assigned an error detection code, and transmitted from the CPU. As shown in FIG. 2, when the first PKT (0) is transmitted from the CPU, it is simultaneously received by all PUs via the DCH, BC, and BUS. At this time, each PU checks the error detection code assigned to PKT (0) and stores whether or not an error has occurred. Next, in order to inform all PUs whether or not PKT (0) has been correctly received, the CPU first transmits a timing signal POL o to PUo. When the PUo receives this POL, it transmits an acknowledgment ACKo to the CPU if there is no error based on the previously stored record of whether an error has occurred. This ACK, CP
When U receives it, it sends POLl to the next (home PU), and when it receives the same (Aα) from PU for this POL, the CPU sends POL to the next PU.This operation is performed sequentially. From PUn-1 to ACKn-
Receiving a 1 means that PKT (0) was correctly received by all PUs. Then, the CPU executes the next PKT (
1) and POLO to PUo as described above, the PU may not be able to correctly receive PKT(1) and may send a negative response NAK. When the CPU receives this NAKo, it retransmits PKT (1) to all PUs. In this way, PKTs are transmitted sequentially, and upon receiving ACKn-1 from PUn-1, which is the final response to the final PKT (m-1),
All programs have finished loading.

In the method shown in Fig. 2, all PUs are finally
In order to adjust the packet reception status, there is a problem that the loading time of the program becomes longer.

Next, FIG. 3 will be explained.

In this method, programs are loaded simultaneously to all PUs in the same way as shown in Figure 2, but PKT
All P'KTs from (0) to PKT (m-+)
is transmitted sequentially, and then POL is finally transmitted to all PUs to adjust the reception state of PKT.

In this method, loading of the program is completed when ACKs are received from all PUs. However, for example, if PUl is from PKT (0) to PKT C1)...
If the PK of the CPU 7J- is received in error, a NAK is sent to the POL1 of the CPU 7J- and others. Upon receiving this NAK, the CPU sends P for the last p'[Jn-1.
After sending OLn-1 and receiving a response from PUn-1, all programs will be reloaded only for PU+ from now on.

In this method, the reloading for the error-generated PU becomes extremely slow (and the retransmission program) monopolizes the data channel and BUS #'' in CO2, which interferes with the communication operations of other PUs.

As described above, the conventional method has problems in that it takes a long time to load the program and when an error occurs during loading. This is a serious problem especially when the number of PUs is large in a CCP for a large-scale system.

[Purpose of the invention]

An object of the present invention is to provide a communication control system that eliminates the drawbacks of the prior art described above, shortens the initial loading time of a program from a central control unit to a communication control unit, and realizes prompt reloading when a program loading error occurs. The objective is to provide a control method for the device.

[Summary of the invention]

In the present invention, a positive response to program loading is confirmed by not receiving a negative response from any processing device for a certain period of time, and reloading is performed when a negative response is received within the certain period of time.

[Embodiments of the invention]

The present invention will be explained below with reference to FIG.

CPU, DCH, BC' via t, PKT (o)
When PKT(1) is transmitted to all teno PUs, the intermittent transmission of PKT(1) at fixed time intervals t is suspended. If no signal is received from the PU during this t, PKT(o)
is interpreted as being correctly received by all PUs.

Then, the next PKT (1) is transmitted. Similarly, PKT
(n-1) is performed sequentially (but, for example, P
If the PU sends KT (2) and receives it by mistake, the PU sends a NAK indicating this to the CPU.
Send to. When the CPU receives this NAK within time t after transmitting PKT(2), it transmits PK'r(2) again to all PUs. If NAKs are not received from all PUs after that, the next PKT (3) is transmitted.

As described above, the present invention confirms a positive response to a PKT by not receiving ~■ from any PU for a certain period of time.

As shown in FIG. 1, since the BUS is shared by all PUs, each PU cannot respond at the same time.

This is because the conventional method stipulates a certain response format in order to identify the PU that sent the response. The specified PUs sequentially transmit responses. A polling method is required. On the other hand, in the method according to the present invention, an affirmative response is indicated by not transmitting any signal from all PUs to the CPU for a certain period of time after the PKT is transmitted, and a negative response is indicated by a PU that detects an error in the PKT.
This can be realized by transmitting an amorphous signal to the CPU within a certain period of time for display.

Note that if the fixed time t for displaying a response after sending a PKT is long, there is no effect of reducing the loading time, but it is sufficient to send a signal of several bytes to display a negative response, while the length of one PKT is at least several hundred bytes, so the length of t is sufficiently short compared to the length of PKT (it can be done).

〔Effect of the invention〕

Recently, data communication systems have become larger and larger, and efforts have been made to improve the processing power of communication control devices. However, when implementing this using a multiprocessor method, increasing the number of processors will cause problems in the initial stages of the program. However, according to the present invention, the loading time is dependent only on the amount of programs, regardless of the number of processors, so the system size is This has the effect of being easily scaled up.

[Brief explanation of drawings]

Fig. 1 shows a data communication system consisting of a central processing unit and a communication control unit, and Figs. 2 and 3 show the initial loading of a program from the central processing unit to the processor in the communication control unit according to the conventional method. System Flow Diagram FIG. 4 is a system flow diagram showing one embodiment of the control method according to the present invention. CPU... Central processing unit, CCP... Communication control unit, MM -... Main memory, DC? H...Data channel device, BC...Bus control unit, BUS...Data node, PUi...Processing device, Mi...
・Storage device, LCi...Line control unit, PKT...
Packet, POL...timing signal, X...acknowledgement, NAK...negative response, t...response monitoring time. Representative Patent Attorney Akira Takahashi ＾ Volume 2 / Funabori 2 Figure 3 Figure 4

Claims (1)

[Claims] In a data communication system consisting of a communication control device configured with a multiprocessor system that connects multiple storage program control processing devices via the same bus, and a central processing unit that controls this communication control device via a data channel. , when performing initial program loading from the central processing unit to each processing unit in the communication control unit, if the program is successfully loaded to all processing units, each processing unit does not send a response to the central processing unit. , when a program is erroneously loaded, a negative response is sent from the erroneously loaded processing unit within a certain period of time, prompting the central processing unit to reload the erroneous program. control method for communication control equipment.