JPH0535496A - Inter-process shared data reference system - Google Patents

Abstract

PURPOSE:To share data to be referred to from a compiled code object between plural processes. CONSTITUTION:Data 6 designated to be inter-process shared data by a compile indicating means 14 in a source program 3, are allocated to a data area 4 of a inter-process shared space 1 by a loader 8-1 of an interpreter 7-1, at the time of the load of a compiled code object 13-2 prepared by a compiler 12-1, the address is set in an address storing area 5, and an in-address storage area offset solving processing 11-1 is operated to a mechanical word command for referring to the data 6 in the compiled code object 13-2, based on the address. When the allocation of the data 6 to the data area 4, and the address setting are already operated at the time of the load of a compiled code object 13-3 or the like of another process 2-2, only the in-address storage area offset solving processing 11-1 is operated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention refers to a compiled code object in an information processing device that executes a plurality of language processors as one process such that an interpreter executes and manages the compiled code object. The present invention relates to an inter-process shared data reference method in which data can be shared by multiple processes.

[0002]

2. Description of the Related Art As one method of a language processor, as shown in FIG. 4, for example, a compiler 42 for translating a source program to generate a compiled code object 44 in order to improve the execution efficiency, and a source program other than the compiler 42 described above. There is a language processor having an interpreter 41 for managing the storage area and execution of the compiled code object 44.

The compiled code object 44 in such a language processor has a header area 45 for storing some kind of header information (including a name for identifying the compiled code object, such as a function name and size). , The data stored in the machine language instruction sequence area 46 storing the sequence of the machine language instruction sequence and the heap area 43 generated and managed by the interpreter 41 is referred from the machine language instruction in the machine language instruction sequence area 46. And an address storage area 47 for storing an address used for that purpose.

When referring to the data on the heap area 43 from within the compiled code object 44,
Since the address of the data in the heap area 43 cannot be determined when compiling the source program, the interpreter 41 uses the compiled code object 4
4 is loaded into the storage area space 40, the address of the data in the heap area 43 is fixed, and the fixed address is stored in a predetermined location of the address storage area 47.

Therefore, the compiler 42 has an address storage area 47 for storing the address of the heap area 43 in the compiled code object 44 at the time of compilation when the interpreter 41 stores the address. Further, it is determined in advance in which place of the address storage area 47 the address of the data to be referred to by the machine language instruction is stored, and a machine language instruction having a structure for referring to the data via the place is generated. ..

For example, in the source program corresponding to the compiled code object 44, the heap area 43
When there is a machine language instruction that refers to the above data 49, the compiler 42 decides to store the address of the data 49 in the heap area 43 in, for example, the third (third word) location 48 of the address storage area 47. At the same time, an instruction that refers to the data 49 via the location 48 is generated as the machine language instruction, and the interpreter 41 loads the address of the data 49 on the heap area 43 to the location 48 when the compiled code object 44 is loaded. It is something to store. As a result, the base register R used when interpreting and executing the interpreter 41 points to the beginning of the compiled code object 44, and the header area 4
Assuming that the size of the 5 + machine language instruction string area 46 is α bytes, the address stored in the area 48 which is the relative position from the base register R (α byte + 3rd word) is
The pointer data points to the data 49.

[0007]

By the way, this kind of language processor is executed as one process in parallel with other processes in the information processing apparatus. Depending on the information processing apparatus, such a language processor may be used. There are some that execute multiple in parallel.

In such an information processing apparatus, in the conventional data reference method of allocating all the data referred from the compiled code object to the heap area generated and managed by the interpreter, each interpreter exists as an individual process. Therefore, since the heap area is also taken up in each process, there is a disadvantage in that the data referenced from the compiled code object cannot be shared by the language processing processors in each process.

The present invention has been made in view of such circumstances, and an object of the present invention is to enable data referred from a compiled code object to be shared by a plurality of processes.

[0010]

In order to achieve the above object, the inter-process shared data reference method of the present invention has an inter-process shared space in which a plurality of processes can be executed in parallel and can be shared between processes. Information for executing a plurality of language processors as one process, including a compiler that outputs a compiled code object including a machine language instruction that refers to data by translating a source program, and an interpreter that interprets and executes the compiled code object In the processing device, data shared between processes is allocated to the inter-process shared space, the allocated address is set in the address storage area provided in the inter-process shared space, and the address storage area is set based on the set address. The above-mentioned compie that has undergone internal offset resolution processing And to perform a reference shared data between the process by machine instructions during de code object.

[0011]

In the inter-process shared data reference method of the present invention, the data shared by the processes is allocated to the inter-process shared space, and the allocated address is set in the address storage area provided in the same space. Data in the inter-process shared space is referred to by a machine language instruction in a compiled code object that has undergone offset resolution processing in the address storage area based on the set address.

[0012]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention.
In the figure, an interpreter 7-1 and a compiler 12-
1 constitutes one language processor, and an interpreter 7
-2 and the compiler 12-2 constitute another language processor.

The language processor including the interpreter 7-1 and the language processor including the interpreter 7-2 are concurrently executed as processes 2-1 and 2-2 in the information processing apparatus to which the present invention is applied.

The process 2-1 includes an interpreter 7-1.
There is a heap area 15-1 which is generated and managed by, and the compiled code object 13-2 and the like are stored therein. Also, the interpreter 7-1 includes a loader 8-1, and the loader 8-1 performs the address storage area registration processing 1
0-1 and address storage area offset resolution processing 1
The shared data generation means 9-1 for performing 1-1 is included.

Similarly, the process 2-2 has a heap area 15-2 generated and managed by the interpreter 7-2, and the compiled code object 13-3 is contained therein.
Etc. are stored. Also, the interpreter 7-2 is a loader 8
-2, and this loader 8-2 has a shared data generation means 9-2 for performing the address storage area registration processing 10-2 and the address storage area offset resolution processing 11-2.

The inter-process shared space 1 is the process 2-1.
And the process 2-2 is a shared storage space. The inter-process shared space 1 has a data area 4, and the data 6 shared by the processes 2-1 and 2-2 in the data area 4 and the processes 2-1 and 2-2 refer to the data 6. The address storage area 5 used when performing is stored.

When a certain compiled code object is shared by the processes 2-1 and 2-2, the compiled code object is allocated to the inter-process shared space 1 as indicated by reference numeral 13-1. .. As a document describing a technique relating to sharing such a compiled code object, for example, Japanese Patent Application No. 63-3
There is an attached specification and a drawing of 10827.

Next, the operation of the present embodiment configured as described above will be described.

Now, the compiler 12-1 of the process 2-1
If the data "abc" to be used as the inter-process shared data exists in the source program 3 to be compiled in step 1, the fact is specified in the source program 3 by the compiler instruction means 14 including the compile directive'common.

The compiler 12-1 compiles the source program 3 to generate a compiled code object. At this time, the compiler instruction determines whether or not the data referred to by the compiled code object is inter-program shared data. Identification is performed by the means 14, and the identification result is included in the compilation result.

The compiled code object generated by the compiler 12-1 is loaded into the heap area 15-1 of the own process 2-1 or the inter-process shared space 1 by the loader 8-1 of the interpreter 7-1. In this case, it is assumed that the heap area 15-1 has been loaded as the compiled code object 13-2.

At the time of this loading, the loader 8-1 executes the processing as shown in FIG.

First, the compiled code object 1
When the data referred to from 3-2 is generated, it is checked whether the data is interprocess shared data (A-
1). If it is not shared data between processes, the data is generated in the heap area 15-1 on its own process space (A-2).

In the case of interprocess shared data, it is checked whether or not the data has already been generated in the data area 4 of the interprocess shared space 1 (A-3).

If such data is not generated, the interprocess shared data is set to the interprocess shared space 1
To the data area 4 (A-4), the next step A
At -5, the address storage area registration processing 10-1 by the shared data generation means 9-1 is performed.

That is, as described above, when the data "abc" is designated as the inter-process shared data and it is not yet allocated to the data area 4, the data as shown in FIG. 3 at step A-4 in FIG. Data 6 is allocated to area 4, and address storage area registration processing 1 in step A-5
At 0-1, the address of the allocated data 6 is stored in the entry of the address storage area 5 of the data area 4. In the example of FIG. 3, it is stored in the second entry (4th byte from the beginning) of the address storage area 5.

After performing the address storage area registration processing 10-1 in step A-5, or in step A-3, it is determined that the interprocess shared data is already allocated to the interprocess shared space 1. In this case, in step A-6, the address storage area offset solving process 11-1 by the shared data generating means 9-1 is performed.

In the offset solution processing 11-1 in the address storage area, the following processing is performed.

For example, on the machine language instruction sequence of the compiled code object 13-2, the instruction "m" indicating that the start address of the data "abc" is transferred to the register Rx.
When there is a machine language instruction that refers to the inter-process shared data “abc” like the machine language instruction corresponding to “ove“ abc ”Rx”, first, the address of the inter-process shared data “abc” that is referred to is the address storage area 5. Find out in which entry of is stored. In the case of FIG.
It is stored in the second entry of the address storage area 5, that is, in the fourth byte from the beginning.

Then, based on this examination result, the contents of the machine language instruction existing in the compiled code object 13-2 are changed. In this case, since the address of the inter-process shared data “abc” is stored in the fourth byte from the beginning of the address storage area 5, the base register R stores the beginning of the address storage area 4 when accessing the inter-process shared space 1. By the way, the machine instructions above are
As indicated by reference numeral 30 in FIG. 3, “move 4 (R), R
x ". As a result, it becomes possible to refer to the inter-process shared data “abc” from the compiled code object 13-2.

As described above, after the source program 3 is compiled by the compiler 12-1 of the process 2-1 to generate the compiled code object 13-2, an instruction for referencing the interprocess shared data "abc" is issued. When another source program including the above is compiled by the compiler 12-2 of the process 2-2 to generate the compiled code object 13-3, the inter-process shared data “abc” is already allocated to the data area 4, Since the address is stored in the address storage area 5, the loader 8-2 of the process 2-2 is set in FIG.
Steps A-4 and A-5 of step 9-2 are omitted, and in step A-6, offset solution processing 9-2 in the address storage area is performed.
Will only do. As a result, the compiler 12-2
Compiled code object 13-3 generated by
Is loaded into the heap area 15-2 of the process 2-2, the machine language instruction that refers to the inter-process shared data “abc” is modified as indicated by reference numeral 31 in FIG.

When the compiled code object 13-1 is loaded in the inter-process shared space 1 as the compiled code object 13-1 in order to share it between processes as shown in FIGS. 1 and 3, it is loaded first. The above process performs the process shown in FIG. 2, and the process using the already loaded compiled code object does not need to perform the process shown in FIG.

[0034]

As described above, according to the present invention,
Since the data allocated on the inter-process shared space can be referenced by the machine language instruction in the compiled code object, the data referenced from the compiled code object can be shared by a plurality of processes.

Therefore, when a large amount of data that is not modified or changed at the time of execution, such as a message character string or dictionary-like data, is used as inter-process shared data, the storage space is effectively used. can do.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an embodiment of the present invention.

FIG. 2 is a flowchart showing an example of processing by a loader.

FIG. 3 is an explanatory diagram of address storage area registration processing and address storage area offset resolution processing.

FIG. 4 is a configuration diagram of a conventional method.

[Explanation of symbols]

1 ... Inter-process shared space 2-1, 2-2 ... Process 3 ... Source program 4 ... Data area 5 ... Address storage area 6 ... Data shared between processes 7-1, 7-2 ... Interpreter 8-1, 8-2 ... Loader 9-1, 9-2 ... Shared data generating means 10-1, 10-2 ... Address storage area registration processing 11-1, 11-2 ... Address storage area offset resolution processing 12-1, 12 -2 ... Compilers 13-1 to 13-3 ... Compiled code object 14 ... Compile instructing means 15-1, 15-2 ... Heap area

Claims (1)

Claim: What is claimed is: 1. A compile including a machine language instruction that allows parallel execution of a plurality of processes, has a shared space between processes that can be shared between processes, and refers to data by translating a source program. In an information processing apparatus that executes a plurality of language processors including a compiler that outputs a coded code object and an interpreter that interprets and executes the compiled coded object as one process, data shared between processes is allocated to the interprocess shared space. , The assigned address is set in the address storage area provided in the inter-process shared space, and the machine language instruction in the compiled code object is used to perform the offset resolution processing in the address storage area based on the set address. Inter-process sharing day Interprocess shared data reference method which is characterized in that the reference.