JPH09319691A - Load reducing system for inter-memory copy erasure - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the burden of a CPU and to improve transfer efficiency by eliminating the copy of transfer data between memories in the case of transferring data between devices by reading the transfer data into a buffer device based on the position and size information of transfer data at the transfer source device. SOLUTION: The position and size information of data as a transfer object at a transfer source device 1 is possessed by a means 3. That possessed position and size information of data is stored in a device 4 while being related with a data buffer device 5 for storing data in the case of writing out the data to a transfer destination device 6. Based on the stored position and size information of data transfer source, the transfer data are read from the transfer source device 1 into the data buffer device 5 by a data reading means 2. Thus, when transferring data between devices, it is not necessary to copy the transfer data between memories, the load of the CPU can be reduced and transfer efficiency can be improved.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to data transfer performance between devices, and more particularly to a system for transferring data in a secondary storage device of an information processing device such as a computer to another peripheral device.

[0002]

2. Description of the Related Art Conventionally, as a data transfer method between peripheral devices, data is read into a dedicated buffer for reading data on a transfer source device and necessary transfer data therein is copied to a data area in an application program between memories. In addition, a method is used in which data is copied between memories in a buffer dedicated to the transfer destination device and then written to the transfer destination device.

C when copying this data between memories
A method for reducing the load on the PU is disclosed in, for example, Japanese Patent Laid-Open No.
It is proposed in Japanese Patent Laid-Open No. 195646 and Japanese Patent Laid-Open No. 12227/1993. These methods are designed to reduce the CPU load when copying data between memories.
Another data processing device is added, and this data processing device copies data between memories to reduce the load on the CPU.

[0004]

However, the above-mentioned conventional data transfer method has the following problems.

(1) The first problem is that when data is transferred between peripheral devices, a delay occurs after the data of the transfer source device is read and before the data is written to the transfer destination device. Is to do.

This is because, in the conventional data transfer system, when data is transferred between peripheral devices, data copy occurs between the memories several times, so that the CPU operates and the memory is read. / Because it takes time to write. For example, when transferring data between disk devices, the data is copied from the buffer of the transfer source disk device to the memory, and then the data of the memory is copied to the buffer of the transfer destination disk device. In the case of data transfer between the disk device and the network device, the data is copied from the buffer of the disk device of the transfer source to the memory, and the data is copied from the memory to the communication buffer. If division or the like is required, further copying will be required.

(2) The second problem is that the CPU load increases when transferring data between peripheral devices.

This is because, when data is transferred between peripheral devices, data copying between memories occurs several times, and at that time, the CPU reads / writes the memory.

(3) A third problem is that the load on the data bus between the CPU and the memory becomes high when data is transferred between the peripheral devices.

This means that when data is transferred between peripheral devices, data copying between memories occurs several times. At this time, the data bus between the CPU and the memory must be occupied while the memory is read / written. by.

Therefore, the present invention has been made to solve the above problems, and an object thereof is to eliminate copying of transfer data between memories when transferring data between devices. To reduce the load on the CPU and improve the transfer efficiency.

[0012]

In order to achieve the above-mentioned object, a load reduction method by inter-memory copy deletion of the present invention provides a means for acquiring the position and size of transfer data on a transfer source device and a transfer destination device. A buffer device for storing the data when transferring the data, a storage unit for storing the position and size information of the transfer data in the transfer source device in association with the buffer device for storing the transfer data, and the transfer source And a means for reading the transfer data into the buffer device on the basis of the position and size information in the devices, thereby eliminating the inter-memory copy when transferring data between the devices.

Further, in the present invention, when the buffer device for accumulating the transfer data is divided, the position and size on the transfer source device of the data to be read into the divided buffer device are recalculated, It is characterized in that it is stored in association with a buffer device.

[0014]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a configuration of an embodiment of the present invention.

Referring to FIG. 1, according to the embodiment of the present invention, a means 3 for acquiring the position and size information of data in a transfer source device 1 of data to be transferred, the position and size information thereof, and a transfer destination device. Data buffer device 5 that stores data when writing the data to device 6, and stores the data in association with data buffer device 5. Transfer data from transfer source device 1 to data buffer device 5 based on the stored position and size information of the data transfer source. And data reading means 2 for reading.

In the embodiment of the present invention, the position and size of the transfer data on the transfer source device 1 are acquired, and are directly read into the data buffer 4 which stores the data to be sent to the transfer destination device 6, and the transfer destination device 1 The data is sent to 6. Therefore, it is not necessary to read the data into the data buffer 5 for the transfer source device 1 and copy the data between the memories to the data buffer device via the data area on the application.

When the transfer destination device 6 is a network device, division of the data buffer may occur. In that case, the position and size are reacquired and associated with the new divided buffer. Just enough. Therefore, even if the data buffer is divided, there is no need to copy data between memories.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In order to explain the above-described embodiments of the present invention in more detail, embodiments of the present invention will be described below with reference to the drawings.

FIG. 2 is a diagram showing the configuration of an embodiment of the present invention. Referring to FIG. 2, in the present embodiment, a hard disk device 1 in which transfer data is stored, a data reading means 2 for reading data from the hard disk device 1,
Position of data on hard disk device 1, means 3 for acquiring size information, message buffer device 5 for storing transfer data, position corresponding to message buffer device for storing transfer data, size storage device 4, network for sending transfer data An apparatus 6 and an operating system 7 that controls the entire computer system are provided.

Next, the operation of this embodiment will be described. Here, the operation in the case of transmitting the data on the hard disk device 1 to the network device 6 will be described.

First, when a request for transferring the data stored in the hard disk device 1 to the network device 6 is made, the operating system 7 stores a message for accumulating the data when transferring the data to be transferred to the network device 6. A buffer device 5 and a position / size information storage device 4 corresponding to the message buffer 5 are prepared.

Next, the operating system 7 acquires the position and size information on the hard disk device 1 in which the target data is stored via the position and size information acquisition means 3, and stores it in the position and size information storage device 4. Store.

After this, the operating system 7
Through the data reading means 2, the target data is read into the message buffer device 5 based on the position and size information.

Finally, the data stored in the message buffer device 5 is sent to the network device 6.

As described above, the requested data can be transferred from the hard disk device to the network device without copying the data between the memories.

When sending data to the network, 1
The operation when it is necessary to divide the message buffer due to the limitation of the amount of data that can be transferred at one time will be described with reference to FIG.

Message buffer device A (5 in FIG. 3)
When dividing 1) into halves, first, a message buffer device B that stores the divided data before reading the transfer data into the message buffer device A (5-1 in FIG. 3) is stored.
1 and B2 (5-2 and 5-3 in FIG. 3) and corresponding position and size information storage devices B1 and B2
(4-2 and 4-3 in FIG. 3) are prepared.

Next, the position and size information of the data read into the divided message buffer devices B1 and B2 through the position and size information acquisition means 3 is recalculated and stored in the position and size information storage devices B1 and B2, respectively.

After this, the message buffer devices B1 and B
The target data is read in 2 and sent to each network device. As described above, even if it is necessary to divide the message buffer, data can be transferred without copying between memories.

[0030]

As described above, according to the present invention,
As a first effect, since the transfer data is directly read into the data buffer for accumulating the data to be transmitted to the transfer destination device and is directly transmitted to the transfer destination, the data is transferred when the peripheral device transfers the data. After the data of the original device is read, the data can be written to the transfer destination device without delay, which has the effect of improving the data transfer performance.

Further, according to the present invention, as a second effect, since it is not necessary to copy data between memories when transferring data between peripheral devices, almost no CPU is used, which is why This has the effect of significantly reducing the CPU load when transferring data between peripheral devices.

Further, according to the present invention, as a third effect, since it is not necessary to copy data between memories when transferring data between peripheral devices, a CPU for reading / writing the memory- There is an advantage that the data bus between the memories is not occupied and the load on the data bus between the CPU and the memory can be reduced.

[Brief description of drawings]

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

FIG. 2 is a diagram for explaining an embodiment of the present invention.

FIG. 3 is a diagram for explaining one embodiment of the present invention,
It is a figure showing the example at the time of message buffer division | segmentation occurring.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Transfer source device 2 Position and size information acquisition means 3 Data reading means 4 Position and size information storage device 5 Data buffer device 6 Transfer destination device 7 Operating system

Claims (3)

[Claims] 1. A means for acquiring the position and size of transfer data on a transfer source device, a buffer device for storing the data when transferring the data to the transfer destination device, and a transfer device for storing the transfer data in the transfer source device. A storage unit that stores the position and size information in association with the buffer device that stores the transfer data; and a unit that reads the transfer data into the buffer device based on the position and size information in the transfer source device. A load reduction method that is characterized by having copy deletion between memories. 2. A data transfer method for transmitting data stored in a randomly accessible secondary storage device such as a hard disk device to a network or transferring to another secondary storage device in an information processing device, Means for acquiring the position and size of the transfer data existing on the secondary storage device as the transfer source, a buffer device for storing the data to be sent to the transfer destination network or the secondary storage device, and the previously acquired Means for storing the position and size information of the transfer data at the transfer source in association with the buffer device for accumulating the transfer data, and the transfer data based on the position and size information of the transfer data at the transfer source And a means for reading into the buffer device of, and a load reduction method by deleting the copy between memories. 3. When the buffer device for accumulating the transfer data is divided, the position and size of the data to be read in the divided buffer device on the transfer source device are recalculated and associated with the buffer device. 3. The load reduction method by inter-memory copy deletion according to claim 1, wherein the inter-memory copy is stored and the inter-memory copy is reduced.