JPH047643A - Data discriminating system - Google Patents

Abstract

PURPOSE:To improve the processing efficiency of a whole information processing system by starting the allocation of data from a specified address at the time when the data is allocated to a storage area, and referring to the data by correcting a tagged pointer to that data by a value corresponding to the sort of the data. CONSTITUTION:At the time when the data is allocated to the storage area 3, a tagged pointer means 2 starts the allocation from the address to be an integral multiple of the unit of the size of the data of the power of '2', and simultaneously, a tag to show the sort of that data is given to the low order several bits to be 0 of the pointer to that data. Then, at the time when the data allocated to the storage area 3 is referred to, an address correcting and referring means 4 refers to it by correcting the tagged pointer to that data by the value corresponding to the sort of the data, and the case the address for referring to the data of the storage area 3 is not the integral multiple of its unit, address exception detecting means 5 processes generated exception. Thus, the processing is simplified, and the efficiency of the whole information processing system is improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a data identification method in an information processing system using an electronic computer.

[Conventional technology]

When multiple types of data are stored in the storage area,
Processing that reads data from a storage area according to predetermined standards, performs predetermined processing only if it is a specific type of data, and treats it as an error if it is other types of data is often used. , When performing such processing, it is necessary to identify the type of data.

Conventionally, there have been the following methods for identifying data types (data identification methods).

■ By limiting the storage area space for data allocation, the upper few bits of the data pointer are set to 0,
A method in which a tag indicating the type of data is attached to the data, the tag is extracted from the pointer when the data is referenced, and the type of data is identified from the value.

■By setting the address of the memory 5i31i to which data is allocated to an integer multiple of a power of 2, the lower several bits of the pointer to the data are set to 0, and a tag indicating the type of data is attached thereto. A method of extracting tags and identifying the type of data from their values.

[Problem to be solved by the invention]

As mentioned above, in conventional data identification methods, the type of data is identified by the tag attached to the pointer to the data when the data is referenced. A comparison operation is required, and the efficiency of the entire information processing system is reduced due to the processing required for this operation.

The present invention has been proposed in view of the above points, and its purpose is to provide a data identification method that can improve the processing efficiency of the entire information processing system.

[Means to solve the problem]

In order to achieve the above object, the present invention provides an information processing system that dynamically allocates and processes multiple types of data in units of power of 2 to storage areas. tagged pointer means that allocates from an address that is an integral multiple of the unit and adds a tag indicating the type of data to the lower several bits of the pointer to the data, and refers to the data allocated to the storage area; address correction reference means that corrects and references a value according to the type of data for the tagged pointer to the data, and an exception when the address that references the data in the storage area is not an integral multiple of the unit. The address exception detection means for generating an address exception is also provided.

[Effect]

In the data identification method of the present invention, when allocating data to a storage area, the tagged pointer means allocates from an address that is an integral multiple of the data size unit that is a power of 2, and also points to the data. A tag indicating the type of data is attached to the lower several bits that are O in The address exception detecting means generates an exception when the address that refers to the data in the storage area is not an integral multiple of the unit.

(Example) Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of an information processing system to which the data identification method of the present invention is applied.

In FIG. 1, the information processing system of this embodiment includes an information processing system control section 1, a tagged pointer means 2 including a data allocation section 21 and a tagging section 22, and a plurality of types of data AI, A2. A storage area 3 in which B1.01 etc. are stored, a data transfer section 41 and an address correction section 42
and an address exception detection means 5 including an address exception detection mechanism 51 and an exception processing section 52. Note that the functions of each part will be explained through the following operations.

Hereinafter, the operation of the above embodiment will be explained for each case.

(1) When data allocation is requested When data allocation is requested from the information processing system control unit l, data allocation of the tagged pointer means 2 is performed by the data allocation unit 2.
1 is a storage type [3], data is allocated from an address that is an integral multiple of the data size unit that is a power of 2, and the tagging unit 22 sets the pointer to the data to 0.
A tag indicating the type of data is attached to the lower several bits.

Here, for example, if the unit of data size (unit of data transfer) is 4 bytes, the address of the storage area 3 to which data is allocated will be a multiple of 4. Furthermore, since the lower two bits for identifying addresses of 4 bytes or less are 0, these two bits can be used to identify four types of data.

For example, for data type A, the tag is set to "10".
”, and the tag for data type B is set as “01”.

FIG. 2(a) shows the tagged pointer PI for data of type A, and the address on the storage area 3 formally indicated by the entire tagged pointer P1 is the 4-byte address where the data is stored. This will point to the beginning of the third byte from the beginning. In addition, Figure 2 and others) show the tagged pointer P2 for the data of seed 11B, and the address in the storage area 3 formally indicated by the entire tagged pointer P2 is the 4-byte address where the data is stored. It points to the beginning of the second byte from the beginning.

(2) At the time of data reference When a request for data reference is generated from the information processing system control unit 1, the address correction unit 4 of the address correction reference means 4
2 corrects the pointer (tagged pointer) to the data designated by the information processing system control unit 1. Specifically, the reference address is obtained by subtracting the tag value corresponding to the type of data from the value of the tagged pointer to the referenced data. That is, when processing only type A data, the lower bits are aligned and 10'' is subtracted from the tagged pointer.

Next, data is read from the storage area 3 using the corrected address, and if the reading is performed normally, the data transfer unit 41 transfers the data to the information processing system control unit 1. Further, if the reference address is not an integral multiple of the data transfer unit, that is, in the above example, if it is not a multiple of 4, the address exception detection mechanism 51 of the address exception detection means 5 operates and the exception handling unit 52 handles the exception. occurs,
No reading is done.

Therefore, in the above example, the address formally indicated by the entire tagged pointer is located after the start address of the data, which is a multiple of 4, by the value of the tag indicating the data type. If the data is of the desired type, the reference address obtained by subtracting it by address correction will be the start address of the data, that is, a multiple of 4, and no exception will occur. If the data is of a type other than the desired one,
The reference address obtained by subtraction through address correction is different from the start address of the data and is not a multiple of 4, so an exception will occur. This results in
Only the desired type of data can be referenced.

〔Effect of the invention〕

As explained above, in the data identification method of the present invention, when referring to data, only the pointer to the data (tagged pointer) is corrected according to the type of data and the actual reference is made. If it is data of this type, it can be read as is, and if it is data of other types, an exception will occur and that fact will be recognized. This has the effect of simplifying the processing and improving the efficiency of the entire information processing system compared to the case where comparisons are made repeatedly.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of an information processing system to which the data identification method of the present invention is applied, and FIG. 2 is an explanatory diagram of a tagged pointer. In the figure, 1... Information processing system control unit 2...
Tagged pointer means 21...Data allocation section 22...Tagging section 3...Storage area 4...Address correction reference means 41...Data transfer section 42... Address correction unit 5...Address exception detection means 51...Address exception detection mechanism 52...Exception processing units Al, A2. Bl, CI...Data PI. P2...Tagged pointer

Claims (2)

[Claims] (1) In an information processing system that dynamically allocates and processes multiple types of data in units of power of 2 to storage areas, when allocating data to storage areas, addresses that are integral multiples of the units are tagged pointer means that assigns a tag indicating the type of data to the lower several bits of the pointer to the data and assigns a tag indicating the type of data; address correction reference means for correcting and referencing a value according to the type of data for a pointer; and address exception detection means for generating an exception when an address for referencing data in a storage area is not an integral multiple of the unit. A data identification method characterized by comprising: (2) The address correction reference means actually performs reference from an address obtained by subtracting a tag value according to the type of data from a value of a tagged pointer to the data to be referenced. data identification method.