JP2906958B2 - User program loading method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a user program loading method, and more particularly to a user of a multiprocessor system in which each processor has an independent local memory, each processor is connected via a network, and the contents of the memory can be copied to each other. It relates to the program loading method.

[0002]

2. Description of the Related Art Japanese Patent Application Laid-Open No. 1-228670 is an example of this type of conventional user program loading system. FIG. 2 shows a block diagram of the example.

In FIG. 2, processors 21 to 23 are connected to a common data bus 24, and data read to the processors is switched by control signals 27 and 28 from a host computer 20.

The execution start address holding block 25 is a means for holding the execution start address of the user program to be read into each of the processors 21 to 23. When reading the execution start address into the execution start address holding block 25,
The control signal 27 from the host computer 20 is transmitted to the processor 21
23, the execution start address is fetched from the common data bus 24 into the execution start address holding block 25, and the processor is turned on via the bus selection block 26 by turning on the control signal 28. The execution start address of the user program is acquired from the execution start address holding block 25. And each processor 21 to 23
Transfers the user program from the host computer 20 to all the processors 21 to 23 after acquiring the execution start address as described above.

The host computers 20 are provided to the processors 21 to 23.
When the user program is transferred from the
Is turned off, each of the processors 21 to 23 reads data from the common data bus 24 via the bus selection block 26. Therefore, when the user program is started, the execution start address holding block 2
By reading the user program execution start address again from 5, the user program can be read at a different address for each of the processors 21 to 23.

[0006]

In the above-described conventional user program loading method, the user program is individually transferred from the host computer to each processor through the data bus. Therefore, as the number of processors connected to the data bus increases, the number of processors connected to the data bus increases. There is a problem that the transfer speed decreases and the load time of the user program increases. Further, since it is assumed that the application programs loaded from the host computer to the respective processors are exactly the same, there is a disadvantage that the processors cannot be operated under different conditions by changing the arguments of the same program.

[0007]

According to the present invention, there is provided a controller in which a plurality of processors each having an independent local memory CPU and an input / output device are connected via a network, and perform file input / output processing with each processor. In a program loading method in a multiprocessor system, the application program to be written to the local memory is classified into a program argument and an environment variable, which are independent from the instruction code and initial data, which are a common part in each processor, The common area transmitting means for transferring the common part to only the input / output device of one processor, and the processor having received the common part writes data into the specified local memory and inputs data to the other processors. output A memory copying means between processors to copy the contents of the local memory between processors via the network to the location, the processor which has received the common part by a memory copying means between said processor,
A common area writing means for writing data extracted from the input / output device to a specified local memory; and an independent part of each processor in the application program from the controller to the input / output device of each processor via the network. The independent area transmitting means for individually transferring the data and the processors which take out the data transmitted by the independent area transmitting means from the input / output device each include an independent area writing means for writing the data to the local memory of the processor.

[0008]

Next, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention, which is a multiprocessor system having two processors b and c and a host controller a as a controller for performing file input / output between each processor. .

The application programs to be loaded into the processors b and c exist on the host controller a separately in a common part and an independent part in each processor, and the common part is indicated by d and the independent parts e and f. . Independent portions e and f of each processor are required for the number of processors, and are separately provided on the host controller a for loading the processors b and c, respectively.

When such an application program existing on the host controller a is loaded into the processors b and c, first, the part d common to the processors b and c is transferred only to the processor b.
From the common area transmitting means 1. Common part d
Receives the start address and the size of the received data in the reception data buffer g and actually writes the data to the local memory j.

Then, referring to the reception data buffer g,
Every time the total of the received data size exceeds a certain value, the common area d is copied to the processor c via the inter-processor network n. Is written to the common area d.

The independent parts e and f of each processor are transmitted from the host controller a to the independent area transmitting means 4.
Are transmitted to the processors b and c, respectively, and the processors b and c that have received the independent areas e and f received from the host controller a are respectively supplied to the local memories j and k by the independent area writing means 5. Write f.

Specifically, a portion d common to each processor of the application program stores a machine language instruction code, initialization data, and non-initialization data of the application program. Data is an external variable or a static variable referred to by an application program. The common part d in each processor is included in an object file obtained by compiling the application program.

In addition, independent parts e and f in each processor
Stores program arguments specified when the application program is actually executed and environment variables set when the application program is executed. Independent portions e and f in each processor are not included in the object file as a result of compiling the application program, but are obtained from the stack area of the host controller a.

Therefore, a common part d for each processor
Reads the object file resulting from compiling the user program and transfers it to the processor b. Since the address to be loaded into each processor is already defined in the object file, the common area transmitting means 1 determines which address of the local memory j is to be loaded.

Independent portions e and f of each processor are information that must be written in the stack area when the user program is executed by each processor, and the start address of the stack can be arbitrarily set by each processor. Therefore, which area of the local memories j and k is to be loaded is determined by the independent area writing means 5 of each processor.

[0018]

According to the method of the present invention, the application program is separated into a common part and an independent part in each processor, and the common part in each processor is loaded from the host controller to one processor. , Can be loaded to all processors, so that the amount of communication data between the host and each processor can be reduced.

In other words, by using inter-processor communication having a high communication capability, the amount of data between a host having a low communication processing capability and a processor can be reduced. It is also possible to execute the program under different conditions, and it is possible to reduce the time required to load each processor.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is a block diagram showing an example of the related art.

[Explanation of symbols]

DESCRIPTION OF REFERENCE NUMERALS 1 common area transmitting means 2 memory copying means between processors 3 common area writing means 4 independent area transmitting means 5 independent area writing means a host controller b, c processor d part common to each processor in application program e, a each part in application program Processor-independent part g Received data buffer h, i CPU j, k Local memory 1, m Input / output device n Network between processors 21, 22, 23 Processor 24 Data bus 25 Execution start address information holding block 26 Bus selection block.

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) G06F 9 / 06,9 / 445 G06F 13 / 00,15 / 177 H04Q 3/545

Claims (2)

(57) [Claims] An independent local memory CPU
And a plurality of processors each having an input / output device connected via a network, in a program loading method in a multiprocessor system having a controller for performing file input / output processing with each of the processors. Instruction code and initial data, which are common parts of the program
A common area transmitting means for classifying the common part into arguments and environment variables and transferring the common part to only the input / output device of one processor; While writing data to the other processor, an inter-processor memory copy unit that copies the contents of the local memory between the processors via the network to an input / output device of another processor, and the common part is received by the inter-processor memory copy unit A processor configured to write data extracted from the input / output device to a specified local memory; and a network from the controller to an input / output device of each processor for a part independent of each processor in the application program. Through Independent region transmitting means for individually transferring, and an independent area writing means for writing the data transmitted by the independent area transmitting means from the input / output device to a local memory of the processor. User program loading method in a loosely coupled multiprocessor system. 2. The inter-processor memory copying means,
2. The user program loading method according to claim 1, wherein the function is activated each time the total of the received data sizes exceeds a certain value.