JP2851739B2 - Data exchange control method between applications - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data exchange control method for exchanging data between applications operating on an apparatus using a computer such as an image terminal, a word processor, and a workstation, and more particularly, to securing a data exchange area. About the method.

[0002]

2. Description of the Related Art Conventionally, data has been exchanged between applications operating on devices utilizing a computer such as an image terminal, a word processor, and a workstation. This exchange is performed as follows. Referring to FIG. 13, it is assumed that application 1 operates on first window 78 provided on the operation screen and application 2 operates on second window 80. As shown in FIG. 13, in this example, it is assumed that the document processing program is operating as both applications 1 and 2.

In some cases, the application 2 needs to copy a partial area 82 of the document displayed on the application 1 into the document being created. In this case, a so-called “cut and paste” function is used. "Cut and paste" is performed as follows.

First, a cut-and-paste area 82 is designated on a document displayed on the application 1 in a window 78. Subsequently, as shown in FIG. 14, by operating the save function by designating the save key 84 displayed on the window 78, for example, the shared key prepared in the storage area of the computer used as this terminal is displayed. Memory 9
In a buffer called cut buffer 92 in 0,
The specified character string is temporarily saved. Subsequently, as shown in FIG. 15, in the application 2 operating on the window 80, the cut / paste position is specified, and the cut / paste key 8
By causing the function 8 to function, the character string stored in the cut buffer 92 is inserted as a cut-and-paste sentence 86 at a designated position in the document displayed on the application 2.

In a recent server-client model, for example, an ICCCM (Inter-client Communication Conventio) for a certain multi-window system is used.
n Manual: According to the provisions of the MIT X Consortium Standard), this cut buffer can be
Or use the server's cut buffer.

Generally, in data exchange performed between these applications, the capacity of data to be exchanged greatly depends on the application, and the required cut buffer capacity may not be accurately predicted. Therefore, as shown in FIG. 16, when the required size of the cut buffer 92 exceeds the standard size prepared in advance in the main memory 94, it is generally used in a large computer or a workstation. As in the case of the virtual storage system, the extended area is often provided in the secondary storage device 96 such as a hard disk. The extension of the cut buffer using the secondary storage device 96 in this manner is the same as in the case of a system having a multiprocessor configuration or the use of the X server cut buffer in the server / client model.

[0007]

When a secondary storage device is used as an extended area of a cut buffer as in the case of the virtual storage system, the area that can be used is very large. However, it can be exchanged by using this extended area. But,
A hard disk or floppy disk used as a secondary storage device is replaced with a DRAM (Dynamic Random Access Memory) used as a temporary storage device (main memory).
y), the reading speed and the writing speed are very slow as compared with SRAM (Static Random Access Memory).

For example, a rough estimate of the current standard access speed for SRAM, DRAM, hard disk, and floppy disk is shown in Table 1 in units of 4 KB.

[0009]

[Table 1]

As apparent from Table 1, a hard disk or a floppy disk used as a secondary storage device requires much more than 100 times the time required for reading and writing compared to an SRAM or a DRAM. Therefore, when a secondary storage device is used as an extended area of the cut buffer, the low access speed is a bottleneck in improving the speed of data exchange between applications.

One way to avoid this problem is to use a disk cache. However, since the disk cache requires special hardware for the secondary storage device, there is a problem that the cost of the device increases.

[0012] Therefore, an object of the present invention is to provide a data exchange control capable of increasing the data exchange speed without increasing the cost of the apparatus when exchanging data between applications operating on a device using a computer. Is to provide a way.

[0013]

According to the present invention, there is provided a method for controlling data exchange between applications, comprising: first and second storage means which can be accessed at a predetermined average access speed; and first and second storage means. And a third storage unit accessible only at an average access speed lower than the access speed of the first storage unit, wherein the first storage unit stores a first data exchange area of a predetermined capacity in advance. Securing the data, detecting whether the data to be exchanged requires a larger capacity than the capacity of the first data exchange area, and requiring a capacity larger than the capacity of the first data exchange area. In response to the detection that the available area has been detected, the available areas of the second and third storage means are searched in this order, and the first available area found is stored. And a step of securing the extended area of the first data exchange area.

[0014]

In the data exchange control method according to the present invention,
If it is detected that the data exchange area secured in advance is insufficient, instead of immediately providing an extended area in the third storage unit having a low access speed, the second storage unit having a high average access speed is first stored in the second storage unit. The third storage means is checked, and the first usable area found is secured as an extended area. Since the second storage means can be accessed at a higher speed than the third storage means, the time required for accessing the extended area is more likely to be reduced.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The concept of a cut buffer used in an inter-application data exchange control method of the present invention will be described first with reference to FIGS. Referring to FIG. 1, when data is exchanged between applications A and B operating on the main memory, one of the applications requests allocation of a cut buffer to a software group called a toolkit layer on the system side. I do. In the toolkit layer, an area called a heap area, which is used to temporarily secure a storage area when an application is executed, is prepared in advance. Then, a standard cut buffer is secured in the heap area in advance. In this case, if the standard cut buffer has a sufficient size for data exchange, data exchange can be performed by reading out the data once stored in the standard cut buffer by the application B.

However, in general, the heap area of the system cannot be so large, and the capacity of the standard cut buffer becomes small accordingly. Depending on the application, a capacity exceeding the standard cut buffer may be required as the cut buffer. In this case, referring to FIG. 2, a temporary file is not immediately created in the secondary storage device unlike the related art, and, for example, an idle memory under the control of a sub-processor in the system is searched to find an available area. If there is, a temporary buffer is secured in the idle memory of this subprocessor and used as a cut buffer. Further, when the capacity of this temporary buffer is also insufficient, a temporary file is prepared in the secondary storage device for the first time, and the standard cut buffer, the temporary buffer, and the temporary file secured in the secondary storage device are used as the cut buffer. .

As described above, for example, in a multiprocessor system, when the capacity of the standard cut buffer secured in the main memory is insufficient, the cut buffer is expanded by first using the idle memory of the sub-processor. The hardware itself is as it is,
There is an effect that the speed of data exchange using the cut buffer can be improved by using the memory efficiently. In particular, as shown in the following embodiments, an image buffer prepared in a graphic processor has a large capacity but is often idle when there is no printing or image input. Therefore, it is expected that the idle memory such as the image buffer can be used more effectively than before, and the speed of data exchange can be increased without increasing the cost.

In such data exchange, if the format of the data to be exchanged is made the same as that of a general file, data exchange using a memory and file reading can be unified. Therefore, the data exchange method using a file and the data exchange using a memory between applications can be realized by the same method. In this case, the idle memory of the sub-processor can be used not only for data exchange between applications but also as a temporary buffer for file exchange.

The same concept can be extended not only between two applications operating on the same device but also to memory use outside the device by using a network. That is, data exchange can be realized using the network by using the idle memory existing in all devices connected to the network.

If there are many idle memories available in the system, the use of secondary storage as temporary files is less likely. Therefore, such a data exchange system can be used particularly effectively in a multiprocessor system or a network to which a large number of devices are connected.

FIG. 3 shows another concept of the data exchange control method according to the present invention. Referring to FIG. 3, when there is a request to secure a cut buffer in FIG. 3, a standard cut buffer in the main memory shown in (b) is first used as a cut buffer. The access to the cut buffer is performed according to the head pointer prepared for the cut buffer 1. The pointer shown in FIG. 3A points to the address 14K of the main memory, and the head of the standard cut buffer can be accessed by referring to this address of the main memory. As shown in FIG. 3B, this standard cut buffer can be expanded a plurality of times. In FIG. 3, the extension area has a variable length. But,
The present invention can be applied not only to extending the standard cut buffer with a variable length as described above, but also to extending the standard cut buffer with a fixed length as in the embodiment described later.

When it is no longer possible to secure a cut buffer in the main memory, as shown in FIG. 3C, a sub-processor other than the main processor connected to the system (for example, a processor of a graphic module) Allocate a temporary buffer using the idle memory of. Then, when the capacity is insufficient even by using the temporary buffer of the graphic module, a temporary file is secured in the secondary storage shown in FIG. As shown in FIG. 3, a standard cut buffer in the main memory,
The temporary buffer of the graphic module and the temporary file of the secondary storage are used by being logically connected to each other using a pointer.

In the above description of the concept of the present invention and embodiments described later, a case where a memory connected to a sub-processor is used will be described as an example. However, the present invention is not limited to this, and can be applied to, for example, a shared memory between a processor and another device, or a high-speed file composed of a semiconductor memory called a so-called IC card or RAM disk. In addition, although data exchange is taken as an example, the same treatment can be realized when exchanging the processor itself as data.

An embodiment of the present invention will be described below in detail with reference to FIGS. In the following, an embodiment will be described in which the present invention is implemented for a document processing workstation (hereinafter, WS) having a hierarchical multiprocessor configuration. However, the present invention is not limited to this. For example, as described above, a system having a shared memory between a processor and another device, or a system in which a high-speed file is connected other than the main memory is used. Can also be applied.

Referring to FIG. 4, this document processing dedicated WS1
Reference numeral 0 denotes a main module 12 for performing document editing, file processing, communication processing, and the like; a graphic module 14 for performing processing such as a window system on display, font management, outline font drawing, and display control; An input control module 16 for controlling user input processing using a mouse, a hand scanner, or the like.

A shared memory 24 accessible from both sides is provided between the main module 12 and the graphic module 14, and a shared memory 54 accessible from both sides is provided between the graphic module 14 and the input control module 16, respectively. Is provided.

The main module 12 includes a main CPU (Central Processing Unit) 18 for controlling document editing, file processing, communication processing, etc., an internal bus 20 connected to the main CPU 18, and a main CPU connected to the internal bus 20. A local memory 22 of the module, a bus 28 in accordance with a standard used in a general personal computer or the like for ensuring expandability of the system, and a bus conversion between the bus 20 and the bus 28 And a bus conversion circuit 26. A hard disk 30 and a bus 28
The AI dictionary 32 and the I / O interface 34 are connected. The I / O interface 34 includes a 3.5-inch disk drive 38 and a SCSI (small computer system).
interface) 36 is connected. SCSI36
Is connected to an external hard disk. Also, I / O
A printer is connected to the interface 34, and an RS232C input / output interface is further provided.
The aforementioned shared memory 24 is connected to the internal bus 20. The main module 12 includes an event-driven type pseudo multitask monitor for facilitating communication processing.

The graphic module 14 controls a large-screen high-resolution CRT (cathode ray tube), which will be described later, and operates a window system at high speed, an internal bus 42 connected to the graphic CPU 40, and an internal bus 42. C connected to the bus 42 and storing frequently used fonts including outline fonts
GROM (Character Generator Read-Only Memory) 4
4, a vector drawing LSI (large scale integrated circuit) 46 connected to the bus 42, a local memory 48, and a frame memory 50. In the frame memory 50, an entire page of A4 size is WYSIWYG (What You See Is
What You Get: 17 inch high resolution CRT (1400 × 19)
00 dots) 52 are connected. The aforementioned shared memories 24 and 54 are also connected to the internal bus 42. In this way, the graphic C is
Since the PU 40 and the vector drawing LSI 46 are provided, the graphic module 14 can process and display a large amount of data at high speed, and has a configuration suitable for processing bitmap data.

The input control module 16 includes an input control CPU 56 for controlling input control, an internal bus 58 connected to the input control CPU 56, and an I / O connected to the bus 58.
O interface 60. The above-described mouse, keyboard, hand scanner, and the like are connected to the I / O interface 60. The shared memory 54 is also connected to the internal bus 58. This input control module 16
Adopts a download method so as to inherit the conventional input method and to easily support a new device.

The architecture in which a plurality of modules are stacked hierarchically from the input side and a shared memory is arranged between the respective layers addresses the following problem.

Conventionally, a workstation employing a single processor and a single bus structure has been used.
However, in this structure, processes from requests requiring a high response (eg, input processing using a mouse) to processes requiring a large load (eg, editor editing processes) are created and operated on the same base. Therefore, there is a problem that it is difficult to predict the operation and responsiveness of the workstation depending on the processing content and performance between the applications. To handle so-called multimedia, a high-speed window system is essential.

The multi-module configuration shown in FIG. 4 has been developed to solve such a problem.
In other words, each process is separated by a hierarchical structure,
By multiplexing buses in which a bus is arranged in each processing module, it is possible to handle from processing of large-capacity data such as images to processing of high response such as mouse control.

In the following description, a method according to the present invention when data is exchanged between applications executed by the main CPU 18 of the main module 12 using a cut buffer will be described with reference to FIGS. explain.

In the following description, a method for securing the cut buffer capacity requested by the application as it is will be described to simplify the description. However, the present invention is not limited to such a method. For example, it is possible to secure an extended area in units of pages or segments as in the virtual memory system, or to use a track buffer as in the disk cache. The present invention can be applied to a ring (when data is accessed for a part of data in a track, the data of the entire track including the data is temporarily stored in a buffer once). it can.

First, referring to FIG. 5, a storage medium used below will be described. Standard cut buffer 62
Are secured in the local memory 22 of the main module 12. The standard cut buffer 62 is provided with a cut buffer management table 64 and cut buffers 1, 2, and 3 of fixed length (each 3 KB).

As the primary extension area of the standard buffer 62, a buffer area for printing prepared in the graphic module 14 is used. In the document processing dedicated WS shown in FIG. 4, the print buffer is arranged in the shared memory 24. The buffer for this printing is a large area of 4 MB or more in order to perform A3 size printing of 400 dots per inch. In addition, the printing buffer is not used except during printing, and is usually idle. Therefore, the shared memory 2 used as a print buffer in this way
If 4 is used as the first extended area of the standard cut buffer, the effect of the present invention is very well exhibited.

When the extended area secured in the shared memory 24 is insufficient, or when the extended area cannot be secured in the shared memory 24, the
0 (see FIG. 4), a temporary file is secured and used as an extension area.

Referring to FIG. 6, cut buffer management table 64 has a cut buffer information table 66 for storing information on cut buffers to be used.
And a standard cut buffer table 68 for storing information on the cut buffers 1, 2, and 3 prepared as standard cut buffers among the cut buffers.
And a graphic module temporary buffer table 70 for storing information about a temporary buffer secured in the graphic module, and a secondary storage for storing information about a temporary file secured in a secondary storage device such as a hard disk. And a temporary file table 72.

The cut buffer information table 66 includes a cut buffer ID (identifier) for specifying a secured cut buffer, a name of each cut buffer,
The type of each cut buffer, a reference flag for indicating whether or not each cut buffer was referenced from the application within a predetermined time, and the ID of the first cut buffer among the standard cut buffers assigned to each cut buffer
And a use flag for indicating which of the standard cut buffers 1, 2, 3, the secondary storage temporary file, and the graphic module temporary buffer are used as the cut buffer. For example, as the top ID,
If "1" is stored, it indicates that the head of the corresponding cut buffer starts from the standard cut buffer 1.

The standard cut buffer table 68 includes an ID (identifier) of each standard cut buffer, an identifier of a file using the standard cut buffer (used file ID), a capacity of each standard cut buffer, When the capacity of the cut buffer is insufficient, a chain list for indicating the next extended area (standard cut buffer, graphic module temporary buffer, or secondary storage temporary file) is stored. Each chain list stores an ID for specifying each extended area. For example, if "G" is stored in the field of the chain list, it indicates that the area has been expanded from the standard cut buffer to the temporary buffer of the graphic module.

The graphic module temporary buffer table 70 has an ID for specifying the temporary buffer.
(Identifier), the ID of the file using the temporary buffer (used file ID), the capacity of the secured temporary buffer, and the extension destination when the temporary buffer is further expanded if the capacity is insufficient. Chain list information indicating the ID and the shared memory 24 of the temporary buffer.
(See FIG. 4), the start address, the time at which this temporary buffer was secured, and the backward pointer 74 indicating the extended area used prior to this temporary buffer are stored. For example, if “F” is stored as the chain list information, it indicates that the extended area is further extended to a temporary file in the secondary storage.

In the secondary storage temporary file table 72,
Similarly to the graphic module temporary buffer table 70, the ID, the used file ID, the capacity, the chain list, the temporary file name, the time at which the temporary file was secured, and the backward pointer 76 are stored.
When “0” is stored as the chain list information, it indicates that there is no area further extended from this area.

Referring now to FIGS. 7 to 12, the operation of document processing dedicated WS 10 when the method of controlling data exchange between applications according to the present invention is implemented in the document processing dedicated WS shown in FIGS. Will be described with reference to a flowchart.

FIG. 7 and FIG.
5 shows a flowchart for acquiring a cut buffer having a predetermined capacity (xKB). FIGS. 9 and 10 show flowcharts when releasing the cut buffer once obtained. FIG. 11 and FIG. 12 are flowcharts when the main module 12 secures the yKB area in the shared memory 24 from the graphic module 14 when the extended area of the cut buffer is secured in the shared memory 24. .

First, referring to FIG. 7, when a request for allocating an xKB cut buffer occurs, in step S (hereinafter simply referred to as S) 001, a request for allocating an xKB area is made to the standard cut buffer management. It is.

Subsequently, in S002, it is determined whether or not all of the xKB area can be allocated by the standard cut buffer as a result of the processing in S001. If the assignment is possible, the process proceeds directly to S009,
If not all can be assigned, the process proceeds to S003.

In step S003, a capacity yKB for requesting allocation to the graphic module is calculated by the following equation.

Y = x-Standard Cut Buffer Allocated Capacity Subsequently, in S004, a subroutine requests the memory management of the graphic module to allocate a partition of the capacity yKB determined in S003.

In S005, as a result of the processing in S004, it is determined whether or not all of the remaining yKB area of the xKB requested to be allocated can be allocated by the memory in the graphic module. If the assignment is possible, the process proceeds directly to S009, but if not all of the rest is assignable, the process proceeds to S006.
Proceed to.

In S006, the following formula is used to calculate the temporary file capacity zKB when requesting the secondary storage management to create a temporary file.

Z = y-Graphic Module Allocation Capacity In S007, a process of creating a temporary file of capacity zKB in the secondary storage (hard disk 30) is performed.

Subsequently, in S008, it is determined whether a temporary file having a capacity of zKB can be created in the secondary storage as a result of the processing in S007. If allocation is possible, the process proceeds to S009. In other cases, however, it means that the xKB area could not be secured after all, and this process returns to an error.

In S009, processing is performed to update the management tables (see FIG. 6) for the standard cut buffer, the graphic module temporary buffer, and the temporary file in the secondary storage, respectively. After updating the management table in S009, the process returns to the main routine with the cut buffer number set to P.

FIG. 8 is a flowchart of a subroutine performed in S004 of FIG. 7, for example. The processing of S001 can be performed in the same manner. Referring to FIG. 8, when a request for allocating a yKB partition is issued to the memory management of the graphic module, an initial value 1 is assigned to a variable f in station S101. The variable f is a variable indicating a position in the free area table prepared for the memory management of the graphic module.

In the station S102, it is determined whether or not the position in the free space table represented by the variable f is the end of the free space table. If the end of the free area table has been reached, it means that a necessary area could not be secured by this processing, and this subroutine returns an error.

If it is not the end of the free area table, the processing is S
Proceeding to 103, a determination is made as to whether the fth free area is currently free. If it is not empty, the process proceeds to S106.
Proceed to 104.

At S104, a process is performed in which the address of the section to be newly assigned is set to the address of the f-th free area.

Subsequently, in S105, a determination is made as to which is larger, between the capacity of the f-th free area and the yKB requested to be allocated. If yKB is larger, the process proceeds to S106; if they are equal, the process proceeds to S108; if the capacity of the f-th free area is larger than yKB, the process proceeds to S107.

In S106, since the allocation of all the yKB has not been completed yet when the f-th free area has been checked, the variable f is incremented by 1, and the flow returns to S102 again to repeat the following processing.

If the capacity of the f-th free area is equal to the allocation request yKB, the process proceeds to S108. S
At 108, a process of setting the state of the f-th free area in the free area table to blank is performed, and the process proceeds to S109.

The processing result of the judgment in S105 is S10
7, when yKB is subtracted from the capacity of the f-th free area to obtain a new capacity of the f-th free area;
The process of increasing the address of the third free area by yKB is performed. As a result, the first half of the fKB free area is used as the extended area of the cut buffer, and the free area is reduced to the remaining area. The post-processing of S107 proceeds to S109.

In S109, a section P in the temporary buffer table of the graphic module in which the state is blank is obtained. Subsequently, in S110, a process of setting the capacity, the starting address, and the like in the obtained section column is performed. Subsequently, in S111, the section P
Is performed, and the subroutine returns to the main routine.

Although the processing of FIG. 8 has been described using the processing of S004 as an example, the same processing is performed in S001.

FIG. 9 is a flowchart of a process for releasing the secured cut buffer area. When there is a release request, first in S401, the cut buffer management table 64
(See FIG. 6).

Subsequently, in S402, a process for requesting the standard cut buffer management to release the area allocated to the cut buffer is performed. As a result, unnecessary ones of the standard cut buffers used as the cut buffers are released.

Subsequently, in S403, it is determined whether or not there is another chain other than the standard cut buffer. If there is no other chain, the process proceeds directly to step S408, but if there is another chain, it is necessary to release those areas, so the process proceeds to step S4.
Go to 04.

At S404, a determination is made as to whether there is a temporary buffer allocated to the graphics module. If no assignment has been made, the process proceeds directly to S406, but if there is an assignment, the process proceeds to S4.
Go to 05. In S405, a process of requesting the graphic module memory management to release the temporary buffer allocation is performed. As a result of this request, the temporary buffer secured in the graphic module is released by processing according to a subroutine described later. S40
The post-processing of 5 proceeds to S406.

In S406, it is determined whether or not there is an area allocated as a temporary file in the secondary storage. If there is no assignment, the processing is directly S4
The process proceeds to 08, but if there is an assignment, the process proceeds to S407. In S407, the temporary file is released. S
The post-processing of 407 proceeds to S408.

In S408, processing for updating the management table is performed according to the processing performed above, and control returns to the program that called this processing.

FIG. 10 is a flowchart of a process for releasing the temporary buffer in response to the temporary buffer release request performed in step S405 of FIG. 9, for example. In the following description, it is assumed that a request to release the section Q has been made. First, in S501, the capacity of the section Q is stored as the capacity, and the address of the section Q is stored as the address.

Subsequently, in S502, a process of setting the state of the section Q to blank is performed. Subsequently, in S503, it is determined how many places the partition Q is adjacent to the empty area. If the section Q is not adjacent to the free area, the process proceeds to step S504. If the section Q is adjacent to the free area only at one of the top and the last, the process proceeds to step S507. If both are adjacent to the free area, the process proceeds to S508.

In step S504, a process is performed to find a blank column (fth) in the free area table prepared for the memory management of the graphic module.

Subsequently, in S505, f of the free area table
A process for setting the capacity of the new free area and its address is performed in the second column.

Subsequently, in S506, a process of setting the state of "free" as the state of the new free area is performed, and this subroutine ends.

[0075] When the process advances from S503 to S507, in S507, processing for merging the free space f ₁ the new free area and adjacent to the free space is performed. That is, the amount of free space f ₁ adjacent to the new free area, the process of adding the capacity of the new free space (compartment Q) is performed, the process is terminated.

If the processing has proceeded from S503 to S508, in S508 these empty areas are
A process for merging with one of the two free areas f ₁ and f ₂ , for example, f ₁ , is performed. That is, f
A _first capacitor, the process for adding the capacity of the newly generated f th compartment Q and the capacitance of f ₂ is carried out.

Subsequently, in S509, a process of setting the state of the other free area f ₂ to blank is performed. S
After 509, this subroutine ends. S508, S
By the processing of 509, the two free areas f ₁ and f ₂ that existed across the section Q are integrated into _one free area f ₁ together with the section Q. Similar processing is performed in S402.

The process shown in FIGS. 7 to 10 eliminates the need for the process of securing a cut buffer for data exchange between applications in the order of the local memory 22, the shared memory 24, and the hard disk 303 from the main module side. Release processing is performed at the point in time.

On the other hand, the graphic module 14 may newly request the allocation of a predetermined area in the shared memory 24. The processing in this case is performed as follows.

Referring to FIG. 11, when a yKB storage area allocation request is issued from the graphic module, S
At 201, a request for allocation of a yKB partition is made to the memory management of the graphic module.

In S202, it is determined whether or not the yKB section can be allocated by the graphic module. If the assignment is possible, the section is returned as the section number R. If the graphic module is partially used, for example, by using the cut buffer by the main module, and the yKB partition cannot be allocated, the process proceeds to S203.

At S203, it is determined whether or not shared memory 24 is used as a cut buffer. If the cut buffer is not used as a result of the determination, it is impossible to create a new usable area, so the process proceeds to S210, where an error status is set in the status register and error recovery is performed. On the other hand, when the cut buffer is used, it is possible to create a yKB area by rolling out one of the used cut buffers to a secondary storage or the like. The process is performed in S204 and subsequent steps.

In S204, a process of selecting a less frequently used cut buffer from among the cut buffers stored in the shared memory is performed. In this case, the condition is that the total capacity zKB of the selected cut buffer is larger than yKB for which allocation has been requested.
In S204, a cut buffer is selected by an algorithm described later with reference to FIG.

Subsequently, in S205, it is determined whether or not the cut buffer to be rolled out can be selected as a result of the processing in S204. If the selection is not possible, the cut buffer cannot be rolled out to create a new section, so the process proceeds to S210, where an error status is set in the status register and the error is returned. On the other hand, if it is possible to select a cut buffer satisfying the condition, the process proceeds to S206.

In S206, a temporary file is created in a secondary storage such as a hard disk, and a process of rolling out the cut buffer selected as a result of the processing in S204 to this temporary file is performed. The process proceeds to S207.

In S207, it is determined whether or not the selected cut buffer can be rolled out as a result of the processing performed in S206. For example, if there is not enough free space in the secondary storage and it is not possible to create a temporary file, rollout cannot be performed, so creating a new usable partition in shared memory No, the process proceeds to S210, and sets an error status in the status register and returns to the error. If rollout is possible, the process proceeds to S208.

At S208, the management table
A process for updating information about the rolled-out cut buffer is performed.

Further, in S209, processing is performed to request the memory management of the graphic module to release the cut buffer area rolled out in S206. This processing itself can be performed by the same processing as the processing described with reference to FIGS. S
As a result of the processing in 209, the selected cut buffer is released as a free area in the shared memory. The post-processing of S209 returns to S201 again, and the processing after S201 is performed.

In the newly executed processing after S201, since a sufficient free area has already been created in the shared memory, the yKB area can be allocated for the graphic module.

FIG. 12 is a flowchart of the cut buffer selection process performed in S204 of FIG. FIG.
The algorithm shown in FIG. 2 is a simple LRU (Least Re
According to the cently-used method, the cut buffer which has been used for the longest time is selected by this algorithm.

Referring to FIG. 12, first in S301,
A process of setting yKB for which the allocation request has been made as the search capacity is performed.

Subsequently, in S302, a pointer k for sequentially referring to each cut buffer in the management table.
Is advanced from the current value to the next cut buffer. In this case, when the pointer has advanced to the last cut buffer in the management table, the pointer k is advanced to the first cut buffer.

In S303, it is determined whether or not the reference flag (refer to the cut buffer information table in FIG. 6) of the cut buffer indicated by the current pointer k is "1". This reference flag has been cleared to zero when the cut buffer selection processing was performed last time. Then, “1” is set when there is a reference to this cut buffer from the application program. Therefore, the result of the determination in S303 is YE
If it is S, it means that the corresponding k-th cut buffer has been used since at least the previous cut buffer selection processing, and is not a rollout target.
In this case, the process proceeds to S306. In S306, a determination is made as to whether the pointer k has made a round in the cut buffer information table of the management table. If the round has already been completed, it means that all the cut buffers have been used between the previous cut buffer selection processing and the current cut buffer selection processing. That is, there is no error, and an error is returned. If the pointer has not completed one cycle, the process proceeds to S307.

In S307, a process for clearing the reference flag field of the k-th cut buffer to zero is performed. The process proceeds to S302 again.

On the other hand, if the result of the determination in S303 is that the reference flag of the cut buffer indicated by the pointer k is zero, the process proceeds to S304. In S304,
A process of adding the cut buffer capacity indicated by the pointer k to the roll buffer cut buffer capacity zKB is performed. Note that this capacity zKB has been cleared to zero during the cut buffer selection processing.

Subsequently, in S305, a comparison is made between the zKB obtained as a result of the calculation in S304 and the yKB requested to be allocated. As a result of comparison, zKB is still yK
If it is smaller than B, it means that all of the requested capacity has not been secured yet, so the process returns to S302, and the process is repeated from S302. On the other hand, if zKB is greater than or equal to yKB for which allocation has been requested, the area of yKB for which allocation has been requested can be released by rolling out and releasing all the selected cut buffers. Then, the process returns to the original routine.

By doing so, there is no contradiction between the processing for securing a temporary buffer for the cut buffer using the shared memory and the processing for securing a necessary area in the shared memory when necessary on the graphic module side. Will be done.

The method according to the present invention has been described in detail based on one embodiment. However, the present invention is not limited to the above embodiment. For example, in the above embodiment, a temporary buffer was created in a shared memory with the graphic module before a temporary file was secured in the secondary storage (hard disk). However, any storage medium such as a memory connected to the system that can be accessed at a higher speed than a hard disk can be used as an area for securing the temporary buffer.
When the system is connected to a network, data exchange between applications operating on the network can be performed using this method.

[0099]

As described above, according to the present invention, when securing the extended area of the data exchange area, the access speed is reduced before the third storage means such as a secondary storage medium having a low access speed. The extension area is secured on the accessible second storage means, and if that is not sufficient, the extension area is secured on the third storage means. When exchanging large data, the necessity of accessing the third storage means having a long access time is reduced, and the speed of data exchange can be increased. Further, there is an advantage that it can be realized by using the conventional hardware as it is, and the cost of the apparatus is not unnecessarily increased.

As a result, without increasing the cost of the device,
It is possible to provide a method for controlling data exchange between applications that can increase the speed of data exchange.

[Brief description of the drawings]

FIG. 1 is a schematic diagram of a software structure for explaining the present invention.

FIG. 2 is a schematic diagram of a software structure for explaining the present invention.

FIG. 3 is a schematic diagram showing a state of a link in an extended area when a cut buffer is extended according to the present invention.

FIG. 4 is a block diagram of a WS dedicated to document processing for implementing the present invention.

FIG. 5 is a schematic diagram showing a state of securing an extended area of a cut buffer according to the present invention.

FIG. 6 is a schematic diagram of a cut buffer management table.

FIG. 7 is a flowchart when a cut buffer is allocated according to the present invention.

FIG. 8 is a flowchart for expanding a region of a predetermined capacity in a shared memory according to the present invention.

FIG. 9 is a flowchart for releasing a cut buffer secured according to the present invention.

FIG. 10 is a flowchart when a predetermined section Q is released in the graphic module.

FIG. 11 is a flowchart illustrating a memory allocating method when a graphic module requests allocation of a partition having a predetermined capacity.

FIG. 12 is a flowchart of a process for selecting a cut buffer to be rolled out in response to an allocation request from a graphic module.

FIG. 13 is a schematic diagram showing a state of data exchange between applications.

FIG. 14 is a schematic diagram showing a method of exchanging data between applications.

FIG. 15 is a schematic diagram showing a method of exchanging data between applications.

FIG. 16 is a schematic diagram showing a conventional method for expanding a cut buffer.

[Explanation of symbols]

 Reference Signs List 10 document processing workstation 12 main module 14 graphic module 16 input control module 18 main CPU 22 local memory 24 shared memory 30 hard disk 40 graphic CPU 48 local memory 54 shared memory 56 input control CPU 62 standard cut buffer 64 cut buffer management table 66 cut Buffer information table 68 Standard cut buffer table 70 Graphic module temporary buffer table 72 Secondary storage temporary file table

Claims (1)

(57) [Claims] 1. A data exchange control method between applications for securing a data exchange area for temporarily storing exchanged data when data exchange is performed between a plurality of applications running on a computer, the computer comprising: Are first and second storage means each accessible at a predetermined average access speed, and third storage means accessible only at an average access speed lower than the access speed of each of the first and second storage means. The method of controlling data exchange between applications, comprising: securing a first data exchange area of a predetermined capacity in the first storage means in advance; and storing data to be exchanged in the first data exchange area. Detecting whether a capacity larger than the capacity is required; and a capacity of the first data exchange area. In response to the detection that a larger capacity is required, the available areas of the second and third storage means are searched in this order, and the first available area is found. As an extension area of the first data exchange area.