JP2002202912A - Recording device, recording control method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the frequency of generation of an entrainment saving processing. SOLUTION: Assuming that the size of all data required for writing from the outside is q, the number of recording media is m, and the size of a block is p, the quotient zm+w+y (z; integer, 0<=z, w; integer, 0<=w<m, y; 0<=y<1) is acquired by diving the size of all the data q by the size of the block p. The control is executed so that data in the blocks to the number of zm are written in parallel to the recording media to the number of m, and that thereafter data in the blocks to the number of (q-zm) are written to the recording media to the number of w+1 (where, w in the case of y=0). Resultantly, the frequency of generation of the entrainment saving processing is reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a recording device for recording data, and more particularly to a recording device for recording data on a recording medium such as a flash memory.

[0002]

2. Description of the Related Art In a recording apparatus of a portable device which handles music data and video data, a recording medium such as a flash memory is used because data can be rewritten, portability is high, and backup by a battery or the like is not required. Is generally used.

However, according to the current flash memory, there is a problem that a waiting time is required when writing data. The reason is that the time required to write the data transferred to the buffer from the buffer to the flash memory is shorter than the time required to transfer the data to be written to the buffer in the recording device. Is significantly longer.

[0004] Japanese Patent Laid-Open Publication No. 2000-132982 discloses a recording apparatus having a plurality of flash memories. According to such a recording apparatus, data can be written to a plurality of flash memories in parallel, so that the above-described problem of waiting time does not occur.

[0005]

Generally, a unit of data writing to a flash memory is called a sector, and a unit of data erasure (described later) on the flash memory is called a block. That is, as shown in FIG. 9A, the data in the flash memory is managed in units of blocks composed of 32 sectors (hereinafter, referred to as blocks).
It is assumed that the data length of the sector is 512 bytes and the data length of the block is 16 kilobytes.)

Further, the flash memory can only rewrite data in one direction, that is, when the data value is 1
It cannot be rewritten from 0 to 0 (or 0 to 1). Therefore, in order to perform data write, all data values in the write destination area are set to 1
(Hereinafter, this conversion process is referred to as “erasing process.” A region where the erasing process has been completed is referred to as an “erased region”, and a region where the erasing process has not been completed is referred to as “erased region”. Unerased area ”).

From the above circumstances, when rewriting data 0 to 15 in block A shown in FIG. 9B, first, as shown in FIG.
6 to 31 are temporarily read from the flash memory F1 to the buffer B. Then, the data 16 to 31 thus read out to the buffer B are rewritten in the erased area E of another flash memory F2 as shown in FIG. .

The above-described series of processing (that is, processing for preventing the data 16 to 31 that do not need to be rewritten from being caught in the rewriting processing) is called wrap-in evacuation processing, and lowers the writing speed to the flash memory. This is a major factor. Note that the rewritten data 0 to 15 shown in FIG. 9D is passed from the outside to the recording apparatus, but a detailed description of this point is omitted here.

[0009] Here, as described above,
According to the recording apparatus disclosed in Japanese Patent No. 132982, data can be written to a plurality of flash memories in parallel, so that the above-mentioned problem of waiting time does not occur. However, when data is written to a plurality of flash memories in parallel,
There is a problem that the frequency of occurrence of the entanglement evacuation process at the time of data erasure increases.

That is, when writing certain music data M1 to the flash memory, if the data size is not an integral multiple of (the size of one block of the flash memory) × (the number of flash memories), as shown in FIG. An empty area is generated in each of the last blocks (described later) B1 to B4 of the flash memories F1 to F4. Thereafter, when another music data M2 is written to the flash memories F1 to F4, the music data M2 is written from the empty area of the last blocks B1 to B4. Will be within. Therefore, if it becomes necessary to erase all of the music data M1 in this state, the music data M2 written in the last blocks B1 to B4 must be saved.

The last block is a block in which data to be written is to be written last. That is, as shown in FIG. 6, when data is written in parallel to m flash memories, no free space is generated in the blocks indicated by hatching to the left of the dashed line, but the blocks are hatched to the right of the dashed line. The indicated block may have an empty area. As described above, blocks in which a free area is likely to be generated are blocks to which data to be written is to be written last, so these blocks will be particularly referred to as “final blocks”.

The present invention has been proposed based on the above-mentioned conventional circumstances, and it is an object of the present invention to reduce the frequency of occurrence of entanglement and evacuation processing in a recording apparatus which writes data to a plurality of flash memories in parallel. Aim.

[0013]

The present invention employs the following means to achieve the above object. That is, according to the present invention, as shown in FIG. 1, data is written in a recording area where erasing processing has been completed, and data is erased collectively in a predetermined block unit composed of a plurality of recording areas. It is assumed that the recording device 1 writes data requested to be externally written in parallel to a plurality of recording media 21, 22, 23, and 24 that perform the above.

Here, if the size of all data requested to be written from outside is q, the number of recording media is m, and the size of the block is p, the address management control unit 4
2 is zm + w by dividing the size q of all data requested to be written from outside by the size p of the block.
+ Y (z: integer of 0 ≦ z, w: integer of 0 ≦ w <m, y:
Obtain a quotient of 0 ≦ y <1). Then, zm blocks of data are written in parallel to m recording media, and then (q-zm) blocks of data are written w + 1 (where,
When y = 0, control is performed such that writing is performed on w) recording media.

For example, the address management control unit 42
As shown in (a), control is performed so that data for the above (q-zm) blocks is written in parallel to w + 1 recording media. By doing so, the number of blocks in which a free area occurs becomes smaller as compared with the case where the above-described conventional technology is applied. As a result, it goes without saying that the frequency of occurrence of the entrainment evacuation process is reduced.

Of course, as shown in FIG. 8B, when writing the data for the above (q-zm) blocks, the process of writing the data for the y blocks to one recording medium is performed by w + 1 recording media. May be controlled so that the process of writing (1-y) blocks of data to one recording medium is performed in parallel on w recording media. . If you do this,
Since there is at most one block with an empty area,
The frequency of occurrence of the entanglement evacuation process is further reduced.

Alternatively, as shown in FIG. 8C, control is performed such that data of the above (q-zm) blocks is written in the order of each recording medium in a state allocated to w + 1 recording media. You may. This control method is particularly effective when “data that does not require rewriting” exists in a block to which a recording area to which a write request is sent belongs. That is, in such a case, the address management control unit 42 controls the total sum of the data that does not need to be rewritten and the data requested to be written as the data of the (q-zm) blocks. I have.

[0018]

FIG. 1 is a block diagram of a recording apparatus to which the present invention is applied, and its configuration will be described below. In the following description, it is assumed that the data length of the sector is 512 bytes and the data length of the block is 16 kilobytes.

First, when an input / output control unit 41 receives a processing request such as data writing / reading / erasing from the outside, the input / output control unit 41 issues an instruction to start the processing to the address management control unit 42 and performs the following data input / output control. I do.

That is, the input / output control unit 41, which has received the data write request, sequentially writes data input from the outside into the plurality of buffers 31, 32, 33, and 34 in units of 512 bytes. On the other hand, upon receiving the data read request, the input / output control unit 41 sets the plurality of buffers
Data is sequentially read out in units of 512 bytes from 3.34 and output to the outside.

Further, the address management control unit 42 responds to a processing start command from the input / output control unit 41 to store the plurality of flash memories 21, 22, 23, 24 and the plurality of buffers 31, 32, 33, 34. Between the flash memories 21, 22, 23, and 24, and between the plurality of flash memories 2.
The management of the data written in 1, 22, 23, and 24 is performed (details will be described later).

Further, in response to a connection switching signal from the address management control unit 42, the selector 43 stores data between the plurality of flash memories 21, 22, 23, 24 and the plurality of buffers 31, 32, 33, 34. Switch the bus connection.

Here, the data write control method performed by the address management control unit 42 is roughly classified into the following two patterns. That is, one is a control method when a data write request is received for a logical address area to which data has not yet been written (hereinafter, referred to as a “first data write control method”). This is a control method when a data write request is received for a logical address area to which data has already been written (hereinafter, referred to as a “second data write control method”).

Hereinafter, these data write control methods will be described in detail. [First data write control method] First, when a data write request is received for a logical address area where no data is written, the four flash memories 21, 22, 23, and 24 are sent to the four flash memories 21, 22, 23, and 24 in the same manner as in the above-described conventional case. And write data in parallel. However, in the present invention, FIG. 2A to FIG.
As shown in (c), the flash memories 21, 22, 2
The writing control method for the last block differs between 3 and 24.

For example, if the size of the data to be written is 64 kilobytes, even if this data is written in parallel to the four flash memories 21, 22, 23, and 24, the final blocks B21, B22, B23,
No empty area is generated in any of B24. This is because the above-mentioned data of 64 kilobytes corresponds to four blocks, and matches the sum of the last blocks B21, B22, B23 and B24.

Therefore, in this case, the address management control unit 4
2 are the four final blocks B as shown in FIG.
Data 0-12 for 21, B22, B23, B24
7 are written in parallel (hereinafter, this writing mode is referred to as “four-block writing”).

If the size of the data to be written is 32 kilobytes, this data is written in parallel to the four flash memories 21, 22, 23, and 24, and the last block B21, B22, B23, B2
4 respectively, a free area of 1/2 block is generated. This is because the 32 kilobytes of data correspond to two blocks, which is equal to half the sum of the last blocks B21, B22, B23, and B24.

Therefore, in this case, the address management control unit 4
2 are two final blocks B as shown in FIG.
Control is performed such that data 0 to 63 are written in parallel only to 21 · B22. That is, the last blocks B21, B22, B2
The degree of parallelism for writing to 3.multidot.B24 is limited from 4 to 2 (hereinafter, this writing form is referred to as "two-block writing").

Here, when the size of the data to be written is 24 kilobytes, even if the degree of parallelism is limited as in the case of the above 32 kilobytes, it is not possible to prevent a free area from being generated in the last block. This is
This is because kilobyte data corresponds to 1.5 blocks and does not match an integral multiple of the block.

Therefore, in this case, the address management control unit 4
2 controls the data 0 to 47 so that only one final block having an empty area is generated as shown in FIG. 2C and parallel writing is performed as much as possible.

That is, one block of data 0 to 31
Are written in parallel to the two final blocks B21 and B22, and the remaining 0.5 blocks of fraction data 32 to
47 is singly written into one final block B21 (hereinafter, this writing form is referred to as “1.
5 block writing ”). Note that “single writing” means that parallel writing is performed with a parallel degree of 1.

As described above, the address management control unit 42
Is designed to determine the control method according to the size of the data to be written. Hereinafter, the determination method will be described.

First, when the size of all data requested to be written is q, the number of flash memories is m, and the size of a block is p, the size q of all data is divided by the size p of the block to obtain zm + w + y (Z: integer of 0 ≦ z, w: integer of w ≦ w <m, y: 0 ≦ y <1)

Here, the above zm means the number of blocks having a degree of parallelism of m, and the above w + y means the number of blocks not reaching the degree of parallelism of m. That is, with reference to FIG. 6, the number of blocks indicated by hatching to the left of the dashed line corresponds to zm, and the number of blocks indicated by hatching to the right of the dashed line corresponds to w + y (where w is Corresponds to block B1, and y corresponds to block B2).

Therefore, the address management control unit 42
Control is performed so that data for blocks is written in parallel to m flash memories, and then data for q-zm blocks is written to w + 1 (but w when y = 0) flash memories. I do. However, when writing the data for (q-zm) blocks, the process of writing the data for y blocks to one flash memory is executed in parallel for w + 1 flash memories, and then (1- y) The process of writing data for a block into one flash memory is executed in parallel on w flash memories.

Hereinafter, as a specific example of the discrimination method, in a situation where one block is 16 kilobytes, writing 152 kilobytes of data to four flash memories (p = 16, m = 4, q = 152) ) Will be described.

First, q ÷ p = zm + w + y is 152 ÷ 1.
Since 6 = 2 × 4 + 1 + 0.5, it is understood that z is 2, w is 1, and y is 0.5.

Therefore, two blocks of data are written in parallel to the four flash memories, and then 1.
Data for 5 blocks is written to two flash memories. However, when writing 1.5 blocks of data, the process of writing 0.5 blocks of data to one flash memory is executed in parallel for two flash memories, and then (1- 0.5)
The process of writing data for a block to one flash memory is executed in parallel for one flash memory. In other words, referring to FIG. 6, writing 0.5 blocks of data in parallel to blocks B1 and B2, and then writing 0.5 blocks of data only to block B1 singly. Become.

Although the size p of all data requested to be written is described as 152 kilobytes and the size q of the block is 16 kilobytes, the units of the sizes p and q are not limited to bytes. Absent.
For example, if the size p of all data requested to be written is 30
The same effect as described above can be obtained even when the size of the block is 4 sectors and the size q of the block is 32 sectors.

As described above, according to the present invention, at most one block has an empty area. As a result, the frequency of occurrence of the entanglement evacuation process is reduced. As a result, it is possible to improve the performance of writing to the flash memory as a whole.

However, according to the present invention, the degree of parallelism may be limited, and in this respect, the writing speed is reduced. However, the degree of parallelism is limited only to the last block, and since parallel writing is performed as much as possible for this last block, even if the degree of parallelism is limited as described above, The effect on the writing speed is extremely small.

In the above description, the operation of performing the parallel writing in the order of the flash memories 21 → 22 → 23 → 24 → 21 →... Has been described, but the present invention is not limited to this. That is, since the hardware performance of each flash memory varies, the time required for each write process also varies.
Therefore, instead of performing parallel writing as described above,
It is preferable to control so that the next data is written in order from the flash memory in which the writing process has been completed earlier.

As shown in FIGS. 2B and 2C,
If the size of the data to be written is 32 Kbytes or 24 Kbytes, the period during which the degree of parallelism is limited to 2, that is, the two flash memories 21.2
There is a period during which the parallel write process is performed only for the second. Therefore, the remaining two flash memories 23
If there is an unerased area in 24, it is preferable to perform the erasing processing of this area in parallel with the writing processing to the flash memories 21 and 22.

Further, in the above description, when writing (q-zm) blocks of data, the process of writing y blocks of data into one flash memory is performed by w +
Although the process of executing data in one flash memory in parallel and then writing (1-y) blocks of data in one flash memory is performed in parallel on w flash memories. However, the present invention is not limited to this. That is, as shown in FIG. 8A, data of the above (q-zm) blocks may be written in parallel to w + 1 flash memories.

In this case, however, the number of blocks having free areas may be two or more. However, it goes without saying that the number is smaller than in the case where the above-mentioned conventional technology is applied.

In order to read or rewrite the data written in parallel as described above, the address management control unit 42 manages the data writing state in the flash memories 21, 22, 23, and 24 in an integrated manner. Must be kept. That is, the address management control unit 42 generates a block management table (see FIG. 3) having the following fields for each logical address block requested to be accessed from outside.

First, the parallel degree field is a field indicating how many flash memories have data written in parallel. That is, "4" is set in the parallelism field of the logical address block in which four blocks have been written as described above, as shown in FIG. On the other hand, as described above, two-block writing and 1.5
As shown in FIGS. 3B and 3C, "2" is set in the parallel degree field of the logical address block in which the block is written.
Set. Note that “0” is set in the parallelism field of the logical address block that is an unwritten area.

Next, the non-parallel sector number field is a field indicating the number of sectors for which parallel writing has not been performed among the 32 sectors included in the logical address block. That is, the non-parallel sector number field of the logical address block in which the four blocks are written and the two blocks are written as described above is shown in FIG.
“0” is set as shown in FIG. On the other hand, as shown in FIG. 3C, "16" is set in the non-parallel sector number field of the logical address block in which 1.5 blocks have been written as described above.

Next, the flash memory designation field is a field indicating the flash memory in which the logical address block exists. For example, “0” is set in the flash memory designation field of the logical address block existing in the flash memory 21, and “1” is set in the flash memory designation field of the logical address block existing in the flash memory 22. I do. Also, "2" is set in the flash memory designation field of the logical address block existing in the flash memory 23, and "3" is set in the flash memory designation field of the logical address block existing in the flash memory 24. I do.

Next, the physical address field is a field indicating a physical address corresponding to the logical address block. Of course, the physical address set in this field is the physical address in the flash memory indicated in the flash memory designation field.

Next, the sector column management information field is:
A field indicating the order of parallel writing in sector units. For example, flash memory 21 → 22 → 23 → 2
When parallel writing is performed in the order of 4 → 21 →.
As shown in FIG. 3A, "0" is used as the sector column management information corresponding to the logical address a, "1" is used as the sector column management information corresponding to the logical address a + 1, and the sector column corresponding to the logical address a + 2 is used. "2" as management information,
"3" is arranged as sector column management information corresponding to the logical address a + 3. [Second Data Writing Control Method] By the way, the data to be written to the flash memory is usually a series of image data or music data (one image file or one music file). Therefore, when erasing data written in the flash memory, there is a high possibility that a series of image data and music data (that is, data written continuously) will be erased collectively. Therefore, in the first data writing control, parallel writing is performed as much as possible from the viewpoint of not lowering the writing speed.

However, in some cases, the performance of writing to the flash memory can be improved as a whole by performing a single write rather than performing a parallel write as much as possible as in the first data write control method. .

For example, in the editing work of music data written in a flash memory, there is a case where it is desired to rewrite a part of continuously written data. In this case, the data to be rewritten is written in one block as shown as a hatched area in FIG. 7B, rather than written over a plurality of blocks as shown as a hatched area in FIG. It is better to be written.
This is because in the scene shown in FIG. 7A, it is necessary to perform the entrainment and evacuation processing for four blocks,
This is because in the scene shown in FIG. 7B, the entrainment and retreat processing may be performed for one block.

From the above, if continuous data is written in parallel to a plurality of flash memories by the first data write control method, and if it becomes necessary to rewrite a part of the continuous data, It is preferable to employ the following control method.

First, the address management control unit 42, which has received the data write request, refers to the contents of the block management table to determine whether the data write request is
Either a request for rewriting part of continuous data (hereinafter sometimes referred to as a “data rewrite request”) or a request for writing new continuous data (hereinafter sometimes referred to as a “new data write request”) Determine if there is. That is, if it is a data write request to a logical address area in which data has already been written,
The data write request is determined to be a data rewrite request, while if the data write request is a data write request to a logical address area in which data has not yet been written, the data write request is determined to be a new data write request. Has become.

Then, the address management control unit 42 that has determined that the data write request is a new data write request adopts the first data write control method. On the other hand, the address management control unit 42 that has determined that the data write request is a data rewrite request further determines whether or not there is data that does not need to be rewritten in the block to which the data to be rewritten belongs.

For example, as shown in FIG.
In a situation where 1.5 blocks have been written to the flash memories 21 and 22 and there is a request to rewrite the data 32 to 47, in the block to which the data 32 to 47 belongs, data 0 that does not need to be rewritten ~
31 (excluding odd numbers). Accordingly, in this case, the address management control unit 42 determines that "there is data that does not need to be rewritten", and performs the following control.

First, the address management control unit 42 stores the data 0 to 31 (32 sector data) that do not need to be rewritten and the data 32 to 47 (1
The data of 1.5 blocks, which is the total sum of the data and 6 sector data), is regarded as the data of the above (q-zm) blocks. At this time, needless to say, w is 1 and y is 0.5.

The address management control unit 42 allocates the (q-zm) blocks of data to the w + 1 flash memories 23 and 24 and assigns the order (here, 23 (In the order of → 24). That is, the data 0 to 31 written in parallel to the two blocks are once read out to the buffer 33,
3 is rewritten singly into the erased area, and then the data 32-
47 is controlled so as to be written solely into the erased area of the flash memory 24. The data 32 to 47 are externally passed to the address management control unit 42 via the input / output control unit 41 and the buffer 34.

As described above, according to the second data write control, blocks on the flash memories 23 and 24 are arranged with continuous data as much as possible. By doing so, when it becomes necessary to rewrite the data 0 to 31 or the data 32 to 47 again, the entanglement evacuation processing does not occur.

In the above description, only the operation in the case where the address management control unit 42 determines that "data that does not need to be rewritten" exists has been described. That is, although the operation in the case where the address management control unit 42 determines that “there is no data that does not need to be rewritten” is not described, the address management control unit 42 in such a case determines A data write control method is adopted.

The reason is that “there is no data that does not need rewriting” means that all of the data is rewritten. For example, the fact that there is a data write request for two blocks in an area of 1.5 blocks means that the data for the previous 1.5 blocks is unnecessary. Therefore, it is appropriate to consider such a data write request to be a request for writing new continuous data.

Finally, FIG. 5 shows how the contents of the block management table change according to the above-described second data write control example. Note that the field configuration and the setting information are the same as those in the first data write control described above, and a description thereof will not be repeated.

[Brief description of the drawings]

FIG. 1 is a block diagram of a recording apparatus to which the present invention is applied.

FIG. 2 is an explanatory diagram of a first data write control method.

FIG. 3 is a state diagram of a block management table during first data write control;

FIG. 4 is an explanatory diagram of a second data write control method.

FIG. 5 is a state diagram of a block management table at the time of second data write control;

FIG. 6 is an explanatory diagram of a determination method of a data write control method.

FIG. 7 is an explanatory diagram of data rewriting.

FIG. 8 is a diagram illustrating a writing mode of data for (q-zm) blocks.

FIG. 9 is an explanatory diagram of a block / sector / entanglement evacuation process;

FIG. 10 is an explanatory diagram of a block in which separate data exists.

[Explanation of symbols]

 Reference Signs List 1 recording device 21 flash memory 22 flash memory 23 flash memory 24 flash memory 31 buffer 32 buffer 33 buffer 34 buffer 41 input / output control unit 42 address management control unit 43 selector

Claims (14)

[Claims] 1. A method according to claim 1, wherein data is written to a recording area in which erasure processing has been completed, and data is erased collectively in a predetermined block unit including a plurality of recording areas. In a recording device for writing data externally requested to write in parallel, the size of all data externally requested to be written is q, the number of recording media is m, and the size of the block is p
And dividing the size q of all data requested to be written from the outside by the size p of the block, zm + w + y
(Z: integer of 0 ≦ z, w: integer of w ≦ w <m, y: 0 ≦
When a quotient y <1) is obtained, zm blocks of data are written in parallel to m recording media, and then (q
-Zm) A recording apparatus comprising an address management control unit for controlling writing of data for blocks to w + 1 (where w = 0 when y = 0) recording media. 2. The address management control unit according to claim 1, wherein
2. The recording apparatus according to claim 1, wherein control is performed such that data of zm) blocks is written in parallel to w + 1 recording media. 3. The address management control unit according to claim 2, wherein
zm) When writing data for blocks, the process of writing data for y blocks to one recording medium is performed by w
Are executed in parallel on +1 recording media, and then (1
2. The recording apparatus according to claim 1, wherein control is performed such that the processing for writing data for one block onto one recording medium is performed in parallel on w recording media. 4. The address management control unit according to claim 1, wherein:
2. The recording apparatus according to claim 1, wherein control is performed such that data of zm) blocks is written in the order of each recording medium in a state allocated to w + 1 recording media. 5. The address management control unit, when there is data that does not need to be rewritten in a block to which a recording area to which a write request belongs belongs, the data that does not need to be rewritten and the data that is requested to be written 5. The recording apparatus according to claim 4, wherein the control is performed by regarding the sum total of (q-zm) blocks as data. 6. The recording apparatus according to claim 1, wherein the address management control unit controls to write next data in order from a recording medium for which writing processing has been completed earlier. 7. The method according to claim 1, wherein the address management control unit performs an erasing process on the block in parallel with the writing process, when there is a block on which the erasing process has not been completed on a recording medium that is not a writing destination. The recording device according to any one of the preceding claims. 8. The address management control unit indicates recording medium designation information for designating a recording medium and a physical address of a recording medium in which data is written, corresponding to a logical address requested to be accessed from outside. Physical address information, parallelism information indicating how many recording media were written in parallel, non-parallel recording area number information indicating the number of recording areas that were not written in parallel, and the order of parallel writing expressed in recording area units 2. The recording apparatus according to claim 1, further comprising a management table capable of setting at least one piece of recording area sequence management information. 9. A method for writing data to a recording area on which erasing processing has been completed and for a plurality of recording media for collectively erasing data in a predetermined block unit including a plurality of recording areas. In a recording control method of writing data externally requested to write in parallel, the size of all data externally requested to be written is q, the number of recording media is m, and the size of the block is p
And dividing the size q of all data requested to be written from the outside by the size p of the block, zm + w + y
(Z: integer of 0 ≦ z, w: integer of w ≦ w <m, y: 0 ≦
When a quotient y <1) is obtained, zm blocks of data are written in parallel to m recording media, and then (q
-Zm) A recording control method characterized by controlling data to be written to w + 1 (where w = 0 when y = 0) blocks of recording media. 10. A method for writing data to a recording area in which erasing processing has been completed and for a plurality of recording media for collectively erasing data in a predetermined block unit including a plurality of recording areas. The size of all data externally requested to be written is q, the number of recording media is m, and the size of the block is p.
And dividing the size q of all data requested to be written from the outside by the size p of the block, zm + w + y
(Z: integer of 0 ≦ z, w: integer of w ≦ w <m, y: 0 ≦
When a quotient y <1) is obtained, zm blocks of data are written in parallel to m recording media, and then (q
-Zm) a program for executing an address management control process for controlling writing of data for blocks to w + 1 (where w = 0 when y = 0) recording media. 11. The program according to claim 10, wherein control is performed such that data for the (q-zm) blocks is written in parallel to w + 1 recording media. 12. When writing data of (q-zm) blocks, a process of writing data of y blocks on one recording medium is executed in parallel on w + 1 recording media, and thereafter, 11. The program according to claim 10, wherein (1-y) the program for controlling the process of writing the data of the block onto one recording medium is executed in parallel on w recording media. 13. The program according to claim 10, wherein the control is performed such that the data of the (q-zm) blocks is written in the order of each recording medium in a state where the data is allocated to w + 1 recording media. 14. When data that does not need to be rewritten exists in a block to which a recording area to which a write request belongs belongs, the sum of the data that does not need to be rewritten and the data requested to be written is calculated by the above (q− z
m) The control is performed by regarding the data as blocks.
The program described in.