JP2014059889A - Memory system, nonvolatile memory device, method for controlling nonvolatile memory device, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To increase the speed of a read operation and increase the number of free blocks that can be used for writing.SOLUTION: A memory system includes: a nonvolatile memory device 3 that has a storage area for storing first data; and an information processing device 2. The information processing device 2 reads data from a first logical address area of the nonvolatile memory device, and includes a first control circuit that if the read data is identical to the first data, provides a deletion notification for associating the first logical address area with the storage area, to the nonvolatile memory device 3. The nonvolatile memory device includes a nonvolatile storage medium 13 and a second control circuit 11. The second control circuit 11 rewrites a management table 12A so as to associate a logic address specified by the deletion notification with the storage area, and transmits the first data to the information processing device 2 if the logic address of a read instruction received from the information processing device 2 specifies the storage area.

Description

  Embodiments described herein relate generally to a memory system, a nonvolatile storage device, and a control method thereof.

An external storage device (eg, HDD (Hard D
isk Drive), SSD (Solid State Drive)) file data is deleted
There is a method in which only data management information is deleted, and data is not actually deleted from the external storage device. Since deletion processing is not performed in the external storage device every time data is deleted on the OS, the performance of the operation of the external storage device is improved by the amount that the deletion processing is not performed. This method is particularly effective for a slow-access HDD.

On the other hand, since the data that should have been deleted on the OS is recognized as valid data in the external storage device (this is referred to as invalid data), a free area where no data exists on the external storage device Will always be less than the free space it recognizes. In particular, when the free area of the external storage device is exhausted, there may be a serious problem in an SSD in which a logical address and a physical address do not necessarily correspond one-to-one.

JP 2010-157133 A JP 2010-161199 A JP 2011-29586 A

Embodiments provide a memory system, a non-volatile memory device, and a control method thereof that can speed up a read operation and increase the number of free blocks that can be used for writing.

A memory system according to an embodiment is a memory system including a non-volatile storage device having a storage area for storing first data, and an information processing device that communicates with the non-volatile storage device. Reads data from the first logical address area of the non-volatile storage device, and when the read data corresponding to the first logical address area is the same as the first data, the non-volatile storage device And a first control circuit for performing a deletion notification for associating the first logical address area with the storage area. The non-volatile storage device rewrites the management table so as to associate a non-volatile storage medium, a logical address and a physical address, and a logical address specified by the deletion notification with the storage area, and the information And a second control circuit that transmits the first data to the information processing apparatus when a logical address included in a read command received from the processing apparatus designates the storage area.

1 is a block diagram of a memory system according to a first embodiment. The figure explaining a part of management table. 1 is a block diagram of a host device according to a first embodiment. The figure explaining a part of data stored in NAND type flash memory. Schematic diagram of a desktop computer. 1 is a schematic diagram of a portable computer. 5 is a flowchart showing an SSD write operation. The flowchart which shows the NAND arrangement | sequence process of SSD. The flowchart which shows the deletion notification process of SSD. 7 is a flowchart showing an SSD read operation. The flowchart which shows the optimization process of a host apparatus. The flowchart which shows the optimization flag setting process of a host apparatus. Schematic which shows an example of GUI of an optimization program. The graph which shows the relationship between the ratio of the data filled up with "0", and average reading speed. The graph which shows the relationship between the ratio of the data filled up with "0", and the amount of free blocks. 9 is a flowchart showing optimization processing of the host device according to the second embodiment. 10 is a flowchart showing optimization flag setting processing of the host device according to the third embodiment. 10 is a flowchart showing a process for generating optimized startup information for an SSD according to a fourth embodiment. The flowchart which shows the optimization flag setting process of a host apparatus. 10 is a flowchart showing optimization processing of the host device according to the fifth embodiment. The block diagram of the memory system which concerns on 6th Embodiment. The figure explaining a part of management table. The flowchart which shows the optimization process of SSD. The block diagram of a memory system in case a comparison source data area is allocated on DRAM. The figure explaining the mapping table of a physical address. The figure explaining the mapping table of a physical address. The block diagram of a memory system in case a comparison source data area is allocated on a SSD controller. The figure explaining the mapping table of a physical address. The figure explaining the mapping table of a physical address. The block diagram of a memory system in case a comparison source data area is allocated on a NAND type flash memory. The figure explaining the mapping table of a physical address. The figure explaining the mapping table of a physical address. FIG. 10 is a block diagram of a host device according to a seventh embodiment. The flowchart which shows the NAND arrangement | sequence process of SSD. The flowchart which shows the deletion notification process of SSD. 7 is a flowchart showing an SSD read operation. The flowchart which shows the optimization process of a host apparatus. 5 is a flowchart illustrating optimization processing of the host device according to the first embodiment. 10 is a flowchart illustrating comparison source data rewriting processing of the host device according to the second embodiment. 10 is a flowchart showing SSD comparison source data rewriting processing according to the second embodiment; 20 is a flowchart showing SSD optimization processing according to the eighth embodiment. FIG. 3 is a diagram illustrating an example of logical addresses and data according to the first embodiment. The figure which shows existence probability distribution of 0-fill sector. FIG. 10 is a diagram illustrating an example of logical addresses and data according to the second embodiment. The figure which shows the existence probability distribution of 0-fill sector and 1-fill sector. FIG. 10 is a diagram illustrating an example of logical addresses and data according to the third embodiment. The figure which shows the existence probability distribution of three types of data patterns. FIG. 20 is a diagram illustrating an example of logical addresses and data according to the sixth embodiment. The figure explaining the sector released by an optimization process. FIG. 10 is a diagram for explaining a method for determining a function f according to a seventh embodiment.

When the free space of the SSD is depleted, that is, when the physical block (free block) of the NAND flash memory that is recognized as having no valid data in the SSD controller is depleted, write to the SSD Then, the invalid data area on the physical block (active block) of the NAND flash memory recognized as having valid data in the SSD controller is searched, and valid data is organized (NAND organized). A load is applied to the process of securing a new free block, and the writing performance is significantly deteriorated.

In addition, when writing is performed in a situation where there is a lot of invalid data in the NAND flash memory, that is, when the free area on the SSD is exhausted, the writing concentrates on a specific area, or NAND occurs due to frequent NAND arrangement. Since extra erasure of the flash memory may occur, the reliability of the SSD may be significantly deteriorated.

Hereinafter, embodiments will be described with reference to the drawings. The following embodiments exemplify apparatuses and methods for embodying the technical idea of the present invention, and the technical idea of the present invention is specified by the shape, structure, arrangement, etc. of components. Is not to be done. In the following description, elements having the same function and configuration are denoted by the same reference numerals, and redundant description will be given only when necessary.

[First Embodiment]
[1. Memory system configuration]
FIG. 1 is a block diagram of a memory system 1 according to the first embodiment. The memory system 1 includes a host device (information processing device) 2 and a nonvolatile storage device 3 as an external storage device. In the present embodiment, the nonvolatile storage device 3 is an SSD (Solid State Drive).
) As an example.

The SSD 3 includes an interface controller 10 and an SSD controller (control circuit) 1
1. A DRAM 12 as a main memory, a NAND flash memory 13 as a nonvolatile storage medium, and a bus connecting them.

The interface controller 10 executes interface processing with the host device 2. The interface controller 10 is, for example, SATA (Serial Advanced Techno
connected to the host device 2 with a communication interface conforming to the logy Attachment) standard, and S
An interface process conforming to the ATA standard is executed. Regarding interface standards, in addition to SATA, SAS (Serial Attached SCSI) and USB (universal serial bus)
) Etc. may be used.

The SSD controller 11 controls various operations of the entire SSD 3. The SSD controller 11 creates various tables in the DRAM 12 by reading the program stored in the NAND flash memory 13 to the DRAM 12 and executing predetermined processing.
Further, the SSD controller 11 receives a write command, a read command, an erase command, etc. from the host device 2 and executes predetermined processing on the NAND flash memory 13 in response to these commands.

The NAND flash memory 13 includes (n + 1) NAND memory chips Chip0.
~ Chipn. n is an integer of 0 or more. The NAND memory chips Chip0 to Chipn may be configured to be capable of parallel operation. Each NAN
The D memory chip includes a memory cell array composed of a plurality of flash memory cells and peripheral circuits (for example, a row decoder, a column decoder, a page buffer, a data cache, etc.)
It has. The memory cell array of the NAND flash memory 13 is composed of a plurality of blocks, and each block is composed of a plurality of pages. A block is a minimum unit of data erasing, and a page is a minimum unit of data writing and reading.

A memory cell transistor constituting the memory cell array of the NAND flash memory 13 is a MOSFET (Metal Oxide Se) having a stacked gate structure formed on a semiconductor substrate.
miconductor Field Effect Transistor). The stacked gate structure includes a charge storage layer (floating gate electrode) formed on a semiconductor substrate with a gate insulating film interposed therebetween, and a control gate electrode formed on the floating gate electrode with an inter-gate insulating film interposed therebetween. It is out. The memory cell transistor changes the threshold voltage according to the number of electrons stored in the floating gate electrode, and stores data according to the difference in the threshold voltage. The memory cell transistor may be configured to store 1 bit, or may be configured to store multiple values (data of 2 bits or more). The threshold voltage can be adjusted by trapping electrons at the nitriding interface as a charge storage layer, such as a MONOS (Metal-Oxide-Nitride-Oxide-Silicon) type, as well as a structure having a floating gate electrode. It may be a simple structure. M
Similarly, the ONOS type memory cell transistor may be configured to store one bit or may be configured to store multiple values. Further, the nonvolatile storage medium may be a semiconductor storage medium in which memory cells are arranged three-dimensionally as described in Japanese Patent Application Laid-Open Nos. 2010-161199 and 2011-29586.

The DRAM 12 functions as a cache for data transfer between the host device 2 and the NAND flash memory 13 and a working memory for the SSD controller 11. DR
What is stored in the work area memory of the AM 12 includes a management table and a function f read from the NAND flash memory 13 at the time of startup or the like. The DRAM 12 includes a storage area 12A for storing the management table and a storage area 12B for storing the function f. Details of the management table and the function f will be described later. As main memory, DRA
A memory other than M may be used. As the main memory, random access memory (RA
It is desirable to use M). As main memory, in addition to DRAM, SRAM, Fe
RAM (Ferroelectric Random Access Memory), MRAM (Magnetoresistive Random)
(Access Memory), PCRAM (Phase Change Random Access Memory), or the like may be used. The main memory may be mounted on the SSD substrate as an independent chip, or may be mounted in the SSD controller as an embedded memory.

FIG. 2 is a diagram for explaining a part of the management table stored in the DRAM 12. When the host device 2 sends a command to the SSD 3, the command is sent to the SSD controller 11 via the interface controller 10. The SSD controller 11 processes the received command. At that time, the SSD controller 11 refers to the management table on the DRAM 12 to appropriately convert the logical address received from the host device 2 into a physical address. The logical address is an address managed by the host device 2, and the physical address is SSD3.
This is the real address managed by.

As shown in FIG. 2, the management table includes a free block table, an active block table, and an active page table. The physical block ID and the physical page address are included in the physical address. The active block table manages the physical block ID and the logical address in association with each other, and the physical block (active block) registered in the active block table holds valid data at least for the SSD controller 11. It will be. The free block table manages physical blocks that are not linked to logical addresses. Since physical blocks (free blocks) registered in the free block table are not linked to logical addresses, they do not hold valid data.

The active page table manages the physical page address and the logical address in association with each other, and the physical page (active page) registered in the active page table
Holds data valid for at least the SSD controller 11. A physical page registered in the active page table is associated with a physical block ID to which the physical page belongs. The logical address managed in the management table is an address having a minimum data size that can be managed by the SSD 3 or a data size larger than that.

Next, the configuration of the host device 2 will be described. FIG. 3 is a block diagram of the host device 2. A CPU (control circuit) 20 is a central processing unit in the host apparatus 2, and various calculations and controls in the host apparatus 2 are performed by the CPU 20. The CPU 20 controls the optical drive 31 such as the SSD 3 or DVD-ROM via the south bridge 21.
The CPU 20 controls the DRAM 23 as the main memory via the north bridge 22.

A user controls the host device 2 through an input device such as a keyboard 29 and a mouse 30, and signals from the keyboard 29 and mouse 30 are processed by the CPU 20 via a USB (Universal Serial Bus) controller 28 and a south bridge 21. . The CPU 20 sends image data, text data, and the like to the display (display device) 25 via the north bridge 22 and the display controller 24. The user can visually recognize image data, text data, and the like from the host device 2 via the display 25.

CPU 20 is a BIOS (Basic Input / Output System) -B stored in ROM 26.
Run IOS. The BIOS is a program for hardware control. Other,
The CPU 20 controls a LAN (Local Area Network) controller 27 via the south bridge 21.

The DRAM 23 functions as a working memory for the CPU 20 and operates as an OS (Operating System).
, A storage area 23B for storing an optimization program, a storage area 23C for storing an optimization flag, and a storage area 23D for storing a function f. As is generally known, the OS manages input / output devices of the host device 2, manages disks and memories, and controls the software to make the hardware of the host device 2 available. A program for managing the entire host device 2. Details of the optimization program and the optimization flag will be described later. The OS, the optimization program, and the function f are the host device 2
4 is stored in the NAND flash memory 13 of the SSD 3 as shown in FIG. 4, but is loaded from the NAND flash memory 13 to the DRAM 23 when the host device 2 is started or when the program is started. Is done. NAND flash memory 13
Includes a storage area 13A for storing an OS, a storage area 13B for storing an optimization program, a storage area 13C for storing a function f, and a storage area 13D for storing a management table. The NAND flash memory 13 includes a storage area (not shown) for storing user data.

Next, a configuration example of the memory system 1 will be described. The memory system 1 can be realized as, for example, a desktop computer or a notebook portable computer.
FIG. 5 is a schematic diagram of a desktop computer as the memory system 1.

The desktop computer 1 includes an information processing apparatus main body 40, a display 25, a keyboard 29, a mouse 30, and the like. The information processing apparatus main body 40 includes a motherboard 41 on which main hardware is mounted, an SSD 3, a power supply device 42, and the like. SS
D3 is a CPU that is physically connected to the motherboard 41 via a SATA cable and is also mounted on the motherboard via a south bridge mounted on the motherboard 41.
20 is electrically connected. The power supply device 42 generates various power sources used in the desktop computer 1 and supplies power to the motherboard 41, the SSD 3 and the like via a power cable.

FIG. 6 is a schematic diagram of a portable computer as the memory system 1. The portable computer 1 includes an information processing apparatus main body 50 and a display unit 51. The display unit 51 incorporates a display device 25 configured by, for example, an LCD (Liquid Crystal Display).

The display unit 51 is attached to the information processing apparatus main body 50 so as to freely rotate between an open position where the upper surface of the main body 50 is exposed and a closed position covering the upper surface of the main body 2. The main body 2 has a thin box-shaped casing, and a power switch 52, a keyboard 29, a touch pad 53, and the like are arranged on the upper surface thereof. The main body 50 also includes an SSD 3, a motherboard, a power supply device, and the like, like the desktop computer.

In addition, the memory system 1 may be an imaging device such as a still camera or a video camera, or may be a game machine or a car navigation system.

[2. Operation]
<2-1. Write operation>
FIG. 7 is a flowchart showing the write operation of the SSD 3. The SSD 3 receives a write command from the host device 2 (step S10). The write command includes a write command, a logical address, and data.

Subsequently, the SSD controller 11 reads the free block table from the DRAM 12 and acquires the physical block ID from the free block table (step S11). If no free block exists, the NAND rearrangement process described later is performed, and the physical block ID
Is acquired (steps S13 and S12).

Subsequently, the SSD controller 11 performs a program process on the free block corresponding to the acquired physical block ID (step S14). Next, SSD controller 1
1 deletes the programmed physical block ID from the free block table (step S15). Subsequently, the SSD controller 11 updates the active block table and the active page table (step S16). That is, the logical address and physical block ID corresponding to the programmed block are added to the active block table, and the logical address, physical block ID, and physical page address are added to the active page table.

Note that the physical block to be written may be erased before the program processing in step S14.

<2-2. NAND organization processing>
Next, the NAND organizing process of the SSD 3 will be described. FIG. 8 is a flowchart showing NAND organization processing of SSD3. For a certain physical block, among all pages included in the physical block, an active page (valid page) included in the active page table is associated with a logical address. On the other hand, not all pages included in a physical block are valid pages, and pages that do not correspond to valid pages (invalid pages) are not associated with logical addresses. The physical block has data corresponding to invalid pages, and free blocks can be secured by classifying these invalid pages and valid pages.

First, the SSD controller 11 sets the column number i = 0 and the free space accumulated amount s = 0 (
Step S20). Subsequently, the SSD controller 11 reads the first row of the active block table from the DRAM 12 and acquires the physical block ID (step S21).
The physical block ID acquired in step S21 is referred to as a current physical block ID. continue,
The SSD controller 11 reads the active page table from the DRAM 12 and acquires all physical page addresses included in the current physical block ID (step S22).

Subsequently, the SSD controller 11 calculates the size of the invalid page in the current physical block, and substitutes this invalid page size for the variable v (step S23). Subsequently, the SSD controller 11 determines whether or not “v> 0” (step S24). If “v> 0”, the current physical block ID is added to the arrangement target block list (step S2).
6). On the other hand, if it is determined in step S24 that there is no invalid page, the SSD controller 11 increments the current physical block ID and moves the work target to the next physical block ID (step S25).

Subsequently, the SSD controller 11 adds the variable v to the free space accumulated amount s (step S27). Subsequently, the SSD controller 11 determines whether or not the accumulated free space amount s has reached the physical block size, that is, whether or not the total size of invalid pages acquired so far has reached the physical block size (step). S28). If the accumulated free space amount s has not reached the physical block size, 1 is added to the column number i (step S25), and then the processing from step S21 is repeated.

On the other hand, when the free space accumulated amount s reaches the physical block size, the SSD controller 1
1 reads data of all active pages from all physical blocks in the arrangement target block list, and stores the read data in the DRAM 12 (step S29). Then S
The SD controller 11 performs an erasure process on all physical blocks in the organization target block list (step S30). Subsequently, the SSD controller 11 updates the active block table, the free block table, and the active page table (Step S
31). That is, the physical block that has been erased is deleted from the active block table and added to the free block table, and the active page from which data has been read is deleted from the active page table.

Subsequently, the SSD controller 11 programs the data of all active pages stored in the DRAM 12 into free blocks (step S32). Subsequently, the SSD controller 11 updates the active block table, the free block table, and the active page table (step S31). That is, the programmed physical block is deleted from the free block table and added to the active block table, and the programmed physical page is added to the active page table.

As described above, since the NAND organization involves many read operations, erase operations, and program operations of the NAND flash memory 13, a long time is required. Therefore, if writing is performed in a situation where free blocks are depleted, NAND rearrangement occurs, so that the write processing speed viewed from the host device 2 may be reduced, and the reliability of the NAND flash memory 13 may be reduced.

Note that the erasing process in step S30 may be performed immediately before the write operation in step S14, without being performed in the NAND arrangement in step S13.

<2-3. Deletion notification processing>
Next, the SSD 3 deletion notification process will be described. Datas adopted by INCITS ATA8-ACS
et management command (commonly known as trim command). This trim command functions as a deletion notification. The deletion notification process is a method in which when the data is deleted on the host device, the logical address space in which the deleted data exists is notified to the SSD so that the area can be handled as a free area on the SSD thereafter. is there. When the deletion notification is received, the SSD accesses the management table and performs processing for invalidating the corresponding area. In the invalidation processing, the SSD management table may be rewritten to determine that the corresponding area is invalid, and the data in the corresponding area itself may or may not actually be erased.

FIG. 9 is a flowchart showing the deletion notification process of the SSD 3. The SSD 3 receives a deletion notification from the host device 2 (step S40). The deletion notification includes a deletion command and a logical address.

Subsequently, the SSD controller 11 reads the active page table from the DRAM 12 (step S41). Subsequently, the SSD controller 11 deletes, from the active page table, all pages in which all the areas in the page are subject to deletion notification among all pages corresponding to the logical address area to be deleted designated by the deletion notification (step S1). S42)
.

Subsequently, the SSD controller 11 reads the active block table from the DRAM 12 (step S43). Subsequently, the SSD controller 11 refers to the active block table, deletes a physical block having no physical block ID in the active page table, and adds it to the free block table (step S44). As described above, the SSD 3 can secure a free block not only by NAND arrangement but also by deletion notification processing.

Note that the data size for which the host device 2 performs the deletion notification may be equal to or larger than the minimum data size managed by the SSD 3 in the management table.

<2-4. Read operation>
Next, the reading operation of the SSD 3 will be described. FIG. 10 is a flowchart showing the read operation of the SSD 3. The SSD 3 receives a read command from the host device 2 (
Step S50). The read command includes a read command and a logical address.

Subsequently, the SSD controller 11 reads the active block table from the DRAM 12 (step S51). Subsequently, the SSD controller 11 refers to the physical block ID corresponding to the logical address received from the host device 2 (step S52). When the physical block ID exists in the active block table, the SSD controller 11
Thereafter, the already written area reading process is performed. That is, the SSD controller 11 reads the active page table (step S53), and the NAND flash memory 13
Are read out (step S54). Then, the SSD controller 11 transmits the read data to the host device 2 (step S55).

On the other hand, when the physical block ID corresponding to the logical address received from the host device 2 does not exist in the active block table, the SSD controller 11 performs unwritten area read processing thereafter. That is, the SSD controller 11 does not perform the reading process of the NAND flash memory 13 and transmits the data represented by the function f to the host device 2 by the data length corresponding to the request from the host device 2 (step S56). Specifically, if the data represented by the function f is an address Ai and data Di, the function f (Ai) = Di (i is 0).
The above integer) is satisfied. The function f can be set freely on the user side. In the present embodiment, a logical address (LBA: Logical Block Addressing) is used as a variable, and “function f
(LBA) = 0 ”. In this example, the SSD controller 11 is connected to the host device 2.
The data filled with “0” corresponding to the data length corresponding to the request from is sent to the host device 2.

The unwritten area read operation is not accompanied by read processing of the NAND flash memory 13. Therefore, the unwritten area read operation is performed at a higher speed than the already written area read operation accompanied by the read process of the active page table and the read process of the NAND flash memory 13.

<2-5. Optimization process>
Next, the optimization process of the host device 2 will be described. FIG. 11 is a flowchart showing optimization processing of the host device 2. The logical address space is divided by physical page size, and the divided elements are defined as logical pages.

The CPU 20 of the host device 2 monitors a specific flag (referred to as an optimization flag) (step S60), and if it is confirmed that the optimization flag is set, the current logical address is set to 0 (step S60). S61). Subsequently, the CPU 20 transmits to the SSD 3 a read command for reading data for the page size from the current logical address, that is, one logical page from the current logical address (step S62). In contrast, SSD3
Read processing is performed in accordance with a read command received from the host device 2, and read data is transmitted to the host device 2.

Subsequently, the CPU 20 receives read data for one logical page from the SSD 3 (step S63). Subsequently, the CPU 20 determines whether or not all read data is filled with “0” (step S64). When all the read data is filled with “0”, the CPU 20 issues a deletion notification to the read logical address area (
Step S65). In contrast, the SSD 3 performs the deletion notification process described above.

Subsequently, the CPU 20 adds the page size to the current logical address (step S6).
6). Subsequently, the CPU 20 repeats the processes of steps S62 to S66 until all the logical pages are read (step S67).

The logical page group to which the deletion notification process is performed by this optimization process is not associated with a physical address in the SSD 3 because the corresponding logical address is deleted from the active page table. When a read command is transmitted to the SSD 3 regarding the logical page for which the deletion notification process has been performed, the unwritten area read operation of FIG.
Since the data filled in is received from the SSD 3, the data in the SSD 3 viewed from the host device 2 is the same as before the optimization process. On the other hand, since the unwritten area read operation is performed at a higher speed than the already written area read operation, the performance of reading the logical page to be deleted is increased compared with that before the optimization process.

Note that the host device 2 may confirm that the optimization flag is set and perform the optimization process when the SSD 3 is in the idle state. As a result, the normal operation of the SSD 3 (
It is possible to prevent a decrease in performance during a write operation or the like.

This optimization process may be performed for the entire LBA area, or may be performed for a part of the LBA areas. For example, the range for performing the optimization process may be limited to a range of 10 gigabytes. In this case, for example, the first optimization process is executed over a range of LBA = 0 to 10 gigabytes, and then the optimization flag is set and the optimization process is started. It may be executed over the range of the next LBA (for example, LBA = 0 + 10 gigabytes + 1) to 10 gigabytes. In this way, by performing the optimization process by dividing the entire LBA area into a plurality of times, the time required for one optimization process can be shortened.

<2-6. Optimization flag setting process>
Next, the optimization flag setting process of the host device 2 will be described. FIG. 12 is a flowchart showing the optimization flag setting process of the host device 2.

The user starts the optimization program of the host device 2 and uses GUI (Graphical User Inter
face) Select the button or menu above to call the optimization function, or start the optimization program from the command line and enter the command to call the optimization function. FIG.
FIG. 3 is a schematic diagram illustrating an example of a GUI of an optimization program. For the GUI shown in FIG. 13, the user selects an OK button using a mouse or the like. By such user operation, the optimization function of the host apparatus 2 is called.

The CPU 20 of the host device 2 monitors whether or not the optimization function has been called by the user (step S70). When the optimization function is called by the user, the CPU 20
Sets an optimization flag in the storage area 23C of the DRAM 23. Thereafter, when it is confirmed that the optimization flag is set, the above-described optimization process is performed.

[3. effect]
According to the first embodiment described in detail above, the following effects can be obtained.

With respect to a logical address area that is a repetition of a specific pattern among valid user data on the host apparatus 2, the host apparatus 2 sends a deletion notification to the SSD 3 to delete the logical address area. Then, the user data that has been subjected to the deletion notification process is read from a specific area (such as a DRAM) other than the NAND flash memory 13, so that the user data read operation can be speeded up.

In addition, a free block can be secured by sending a deletion notification to SSD3.
After the optimization process, the performance of the write operation can be improved and the deterioration of the reliability of the SSD 3 can be reduced. This effect increases as the number of user data valid on the host device 2 is completely filled with “0”.

FIG. 14 is a graph showing the relationship between the ratio of data filled with “0” and the average read speed. The average read speed is an average of read speeds of read operations that occur during a specific period. As the ratio of data filled with “0” in the user data stored in the SSD 3 increases, the unwritten area read operation increases, and the average read speed increases.

FIG. 15 is a graph showing the relationship between the percentage of data filled with “0” and the free block amount. The raised portion of the graph of FIG. 15 is the amount of free blocks other than the user area. As the proportion of data filled with “0” in the user data stored in the SSD 3 increases, the amount of free blocks viewed from the host device 2 increases.

For example, when full formatting is performed on SSD3 in Windows (registered trademark), S
Data “0” is written to all logical addresses of SD3. As a result, the number of free blocks is significantly reduced, so that the write performance is lowered, and the reliability of the SSD 3 is likely to deteriorate in the subsequent write operation. However, when the optimization process of the present embodiment is performed on the SSD 3 after the full format, the deletion notification process is performed for all the areas filled with “0”, so that the number of free blocks is greatly increased. As a result, the writing performance is increased, and the deterioration of the reliability of the SSD 3 in the subsequent writing operation can be reduced.

This embodiment can be realized simply by installing in the host device 2 software or firmware that performs optimization processing on the SSD 3 and rewrites the management table when the SSD 3 is in an idle state. Therefore, the function for realizing the present embodiment can be easily realized at low cost. In addition, it is desirable that the optimization process is performed when the SSD 3 is in an idle state, so that even when this embodiment is realized, performance degradation such as writing to the SSD 3 is not accompanied.

In the above description, data filled with “0” is used as data transmitted from the SSD controller 11 to the host device 2 in the unwritten area read operation.
The data represented by can be freely set by the user. For example, “function f (L
By defining (BA) = 1 ”, data filled with“ 1 ”may be used.
By defining “function f (LBA) = LBA”, data represented by “LBA” may be used. Further, the logical address space is divided and managed, and data that is filled with “0” is used for the deletion notification for the first partition, and “1” is used for the deletion notification for the second partition. The filled data may be used. The function f in this case is
For example, “f (LBA) = 0 (LBA <Cap / 2, f (LBA) = 1 (LBA ≧ Cap)
/ 2) ". Cap is the storage capacity of SSD3.

In addition, the logical page read from the SSD 3 by the optimization process is a specific data pattern X (
If the data size of X is the same as the page size), the host device 2 may send a deletion notification. In this case, it is assumed that “function f = X”. When the SSD 3 receives this deletion notification, the SSD controller 11 transmits the data pattern X stored in the DRAM 12 to the host device 2 in the unwritten area read operation. The example of using a specific data pattern for the deletion notification is particularly effective when the memory system 1 frequently handles the same data. In this embodiment, for the sake of simplicity, the physical address of the NAND flash memory is directly associated with the LBA in the management information. However, for example, as shown in Japanese Patent Application Laid-Open No. 2009-217603, a natural number multiple of 2 or more of the sector size. The management unit may be used to manage the LBA, and instead of the LBA, the correspondence between the management unit and the NAND flash memory may be described in the management information. NAND like LBA management unit
The management information describing the correspondence between the physical address of the flash memory and the LBA can take various forms. Details of the form taken by the table (logical-physical conversion table) describing the correspondence between the physical address and the LBA The essence of this embodiment is not affected.

In this embodiment, when the data received from the host device is written in the NAND flash memory, the data itself (plain text) received from the host device is recorded. However, the plain text is randomized to obtain the NAND flash memory. The present embodiment can also be applied when storing in a memory or encrypting plaintext and storing it in a NAND flash memory.

[Second Embodiment]
In the optimization processing of the first embodiment, all logical addresses are read, whereas in the second embodiment, a file list is obtained by accessing the OS management area in the host device 2. However, optimization processing is performed in units of files.

FIG. 16 is a flowchart showing optimization processing of the host apparatus 2 according to the second embodiment. The CPU 20 of the host device 2 monitors the optimization flag (step S80), and confirms that the optimization flag is set. Then, the CPU 20 acquires file management information from the OS, and stores all files registered in the file management information. Add to the optimization target file list (step S8)
1). The optimization target file list is stored, for example, in the DRAM 23 of the host device 2.

Subsequently, the CPU 20 sets the first file in the optimization target file list as the current file (step S82). Subsequently, the CPU 20 transmits a read command for reading the current file to the SSD 3 (step S83). On the other hand, the SSD 3 performs read processing in accordance with the read command received from the host device 2 and transmits read data to the host device 2.

Subsequently, the CPU 20 receives read data corresponding to the current file from the SSD 3 (step S84). Subsequently, the CPU 20 determines whether or not there is a logical page filled with “0” in the read file (step S85). If there is a logical page filled with “0”, the CPU 20 notifies the corresponding logical page of deletion (step S86). In contrast, the SSD 3 performs the deletion notification process described above.

Subsequently, the CPU 20 sets the next file in the optimization target file list as the current file (step S87). Subsequently, the CPU 20 repeats the processes of steps S83 to S87 until all the files in the optimization target file list are read (step S88).
.

As described above in detail, in the second embodiment, in the optimization process, the read operation is not performed on the entire logical address space, and the read operation is performed only on the valid file on the host device 2. This reduces the amount of data targeted for optimization,
The time required for the optimization process can be shortened.

This optimization process may be performed on all files or limited to a part of the files. For example, the number of files to be optimized may be limited to 10,000. In that case, for example, the first optimization process is executed over 10,000 files from the top of the file list, and when the optimization flag is set and the optimization process is started next, the next optimization file after the last optimization was performed. It may be executed over 10,000 files from the file (for example, “0 + 10000 + 1” -th in the file list). Alternatively, the capacity of the file to be optimized may be limited to 10 gigabytes. In that case, for example, the first optimization process is executed over a total of 10 gigabytes of files from the beginning of the file list, and when the optimization flag is set and the optimization process is started next, the last optimization performed last time You may make it perform over the file for a total of 10 gigabytes from the next file of a file. In this way, by performing the optimization process by dividing all files into a plurality of times, the time required for one optimization process can be shortened.

[Third Embodiment]
The third embodiment is another example relating to the optimization flag setting process, and an optimization flag is set when the clock of the host device 2 reaches a specified time.

FIG. 17 is a flowchart showing optimization flag setting processing of the host apparatus 2 according to the third embodiment. The host device 2 has a clock, and the CPU 20 of the host device 2 monitors whether or not the clock has reached a specified time (step S90). When the prescribed time is reached, the CPU 20 sets an optimization flag in the storage area 23C of the DRAM 23. Thereafter, when it is confirmed that the optimization flag is set, the above-described optimization process is performed.

In the third embodiment, the memory system 1 can perform the optimization process in the background without an explicit instruction from the user.

[Fourth Embodiment]
The fourth embodiment is another example relating to the optimization flag setting process, and the optimization flag setting process is started by using information transmitted from the SSD 3 to the host device 2 as a trigger. Information prompting execution of the optimization process transmitted from the SSD 3 to the host device 2 is referred to as optimization start information. One example of information managed in the SSD 3 is the number of free blocks. In the fourth embodiment, optimized activation information is generated using the number of free blocks.

First, the operation on the SSD 3 side will be described. FIG. 18 shows an SSD according to the fourth embodiment.
3 is a flowchart showing a process for generating optimization start information 3. The SSD controller 11 monitors the number of free blocks (step S100), and the number of free blocks is a specified value (
For example, when the value falls below a predetermined value determined in advance at the time of design, the SSD controller 11
Transmits optimization activation information to the host device 2 (step S101). As the specified value, for example, the minimum number of free blocks necessary for keeping the writing speed at a certain level or higher in the writing operation of the SSD 3 can be used.

Next, the operation on the host device 2 side will be described. FIG. 19 is a flowchart showing the optimization flag setting process of the host device 2. The CPU 20 of the host device 2 monitors whether or not optimization activation information has been received from the SSD 3 (step S110). When the optimization activation information is received, the CPU 20 sets an optimization flag in the storage area 23C of the DRAM 23. Thereafter, when it is confirmed that the optimization flag is set, the above-described optimization process is performed.

For example, SMART, which is a memory self-diagnosis function, is used to monitor and exchange optimized startup information.
(Self-Monitoring Analysis and Reporting Technology) command SMART READ
DATA (B0h (D0h)) or a vendor-specific command may be used. Alternatively, the CPU 20 directly monitors the number of free blocks of the SSD 3 using, for example, SMART READ DATA (B0h (D0h)), which is a SMART command, or a vendor-specific command. The CPU 20 may set an optimization flag in the storage area 23 </ b> C of the DRAM 23 when the predetermined value is less than a predetermined value (held on the optimization program 23 </ b> B).

In the fourth embodiment, the optimization process can be performed at the optimal timing on the SSD 3 side.

[Fifth Embodiment]
In the fifth embodiment, the memory system 1 is configured such that the data pattern defined by the function f can be changed in accordance with the usage state of the SSD 3. FIG. 20 is a flowchart showing optimization processing of the host apparatus 2 according to the fifth embodiment. It is assumed that a specific data pattern is represented by X and “function f = X”.

Steps S120 to S123 in FIG. 20 are the same as steps S60 to S63 in FIG. Subsequently, the CPU 20 determines whether or not the read data is the same as the data pattern represented by the function f (step S124). When the read data is the same as the data pattern represented by the function f, the CPU 20 issues a deletion notification to the read logical address area (step S125). In contrast, the SSD 3 performs the deletion notification process described above.

Subsequently, the CPU 20 adds the page size to the current logical address (step S1).
26). Subsequently, the CPU 20 performs steps S122 to S1 until all logical pages are read.
26 is repeated (step S127).

Subsequently, the CPU 20 calculates the total size of the area of the data pattern X among all the logical address areas (step S128). Subsequently, the CPU 20 calculates the total size of the areas of the data pattern X ′ different from the data pattern X among all the logical address areas (step S129).

Subsequently, the CPU 20 determines whether or not the total size of the data pattern X ′ is larger than the total size of the data pattern X (step S130). When the total size of the data pattern X ′ is equal to or smaller than the total size of the data pattern X, the CPU 20 ends the optimization process. On the other hand, when the total size of the data pattern X ′ is larger than the total size of the data pattern X, “function f = X” stored in the storage area 23D of the DRAM 23 is changed to “function f = X ′”.
(Step S131) If the determination condition in Step 130 is (total size of data pattern X ′) ≧ (total size of data pattern X), the total size of data pattern X ′ and the total size of data pattern X are Since the function f is frequently rewritten when they are substantially equal, for example, a margin may be given to the determination condition as (total size of data pattern X ′) ≧ 1.1 × (total size of data pattern X). desirable.

At this time, since deletion notification processing is performed for the logical page group that is the data pattern X, the SSD 3 does not manage this logical page as an active page. Therefore, the CPU 20 transmits a write command for rewriting the data pattern X for the logical page group that is the data pattern X to the SSD 3 (step S132). CPU2
0 is “function f = X stored in the storage area 13 </ b> C of the NAND flash memory 13.
An instruction for rewriting “function f = X ′” is transmitted to the SSD 3 (step S133). For the rewriting instruction for the function f, for example, a vendor-specific command may be used.
The SSD 3 writes “function f = X ′” in the storage area 13 </ b> C of the NAND flash memory 13 and updates the function f of the DRAM 12.

As described above in detail, according to the fifth embodiment, it is possible to change a data pattern that is a deletion notification target. As a result, the unwritten area read operation is performed on the data pattern having the highest existence probability, so that the read operation can be speeded up. Note that the fifth embodiment can be applied to the second embodiment in which optimization processing is performed in file units.

[Sixth Embodiment]
In the sixth embodiment, the SSD 3 itself performs the same processing as the optimization processing described above. FIG. 21 is a block diagram of the memory system 1 according to the sixth embodiment. The SSD 3 DRAM 12 includes a storage area 12C for storing an optimization flag. Other configurations are the same as those in FIG.

In the present embodiment, the host device 2 does not have to hold the optimization program 23B, the optimization flag 23C, and the function f (23D). The SSD controller 11 sets an optimization flag in the storage area 12C of the DRAM 12.

FIG. 22 is a diagram for explaining a part of the management table stored in the DRAM 12. In order not to perform the optimization process redundantly, it is desirable that an optimized mark is added to the active page table. The optimized mark is provided corresponding to the logical address. The optimized mark is information for determining whether or not an optimization process described later has been performed.

The write operation of the SSD 3 is the same as the flowchart of FIG. 7 shown in the first embodiment. However, when the active page table is updated in step S16 of FIG. 7, an optimized mark = false is added.

Next, the optimization process will be described. FIG. 23 is a flowchart showing the optimization process of SSD3.

When the SSD controller 11 confirms that the optimization flag is set in the storage area 12C of the DRAM 12 (step S140), the SSD controller 11 sets the first line of the active page table as the current line (step S141). Subsequently, the SSD controller 11 reads the current row of the active page table from the DRAM 12 (step S142).

Subsequently, the SSD controller 11 determines whether or not the physical page indicated by the current line has been optimized by confirming the optimized mark included in the current line (step S143).
. When the physical page has been optimized (optimized mark = True), the optimization process is not performed on the physical page. On the other hand, when the physical page has not been optimized (optimized mark = false), the SSD controller 11 reads the physical page indicated by the current row from the NAND flash memory 13 (step S144).

Subsequently, the SSD controller 11 is filled with read data “0” (
It is determined whether “function f (LBA) = 0” (step S145). If the read data is filled with “0”, the SSD controller 11 deletes the current row from the active page table (step S146). Alternatively, the current line may be invalidated by writing invalid data to the element of the current line. When the read data is not filled with “0”, the SSD controller 11 writes “True” to the optimized mark of the current row (step S147). Subsequently, the SSD controller 11 sets the next row of the active page table as the current row (step S148).

Subsequently, when the SSD controller 11 receives an interrupt command from the host device 2,
In order to finish the optimization process, the process proceeds to step S151. If no interrupt command has been received from the host device 2, the SSD controller 11 repeats the processing of steps S142 to S148 until all rows of the active page table are read (step S150). Subsequently, the SSD controller 11 updates the active block table and the free block table (step S151). That is, referring to the active block table, a physical block having no element in the active page table is deleted from the active block table and added to the free block table. If the interrupt instruction is urgent, the active block table and free block table may be updated later.

The read operation and the NAND organizing operation of the SSD 3 are the same as those in the first embodiment. Accordingly, the host device 2 relates to the logical page deleted by the optimization process by the SSD 3
When the read command is transmitted to the non-written area shown in FIG. 10, the host device 2 receives data filled with “0” from the SSD 3 (“function f (LBA) = 0”.
Therefore, the data in the SSD 3 as seen from the host device 2 is the same as before the optimization process. On the other hand, the unwritten area read operation is performed at a higher speed than the already written area read operation, The logical page read performance is increased as compared with that before the optimization process, and according to the sixth embodiment, the processing load on the host apparatus 2 can be reduced.

The condition for setting the optimization flag may be a case where it is explicitly instructed from the host device 2 by, for example, a vendor-specific command, or a case where it is not instructed explicitly from the host device 2. . For example, the SSD controller 11 may set an optimization flag in the storage area 12C of the DRAM 12 when there is no access from the host device 2 for a certain period (for example, 10 seconds). Further, as in the fourth embodiment, the SSD controller 1
1 monitors the number of free blocks, and sets an optimization flag in the storage area 12C of the DRAM 12 when the number of free blocks falls below a prescribed value (for example, a prescribed value determined in advance at the time of design). May be.

The above optimization may be performed for all elements of the active page table.
You may make it carry out limiting to a one part element. For example, in each optimization operation, 1000
You may make it carry out limiting to 0 elements. In this case, the first optimization optimizes the 0th to 9999th elements of the active page table, and the next optimization is 1
0000th to 19999th elements are optimized, and the next optimization is 2000
The optimization may be performed on the 0th to 29999th elements. By dividing the active page table and performing optimization in this way, the time required for one optimization can be shortened.

In the above description, data filled with “0” is used as data transmitted from the SSD controller 11 to the host device 2 in the unwritten area read operation (when “function f (LBA) = 0”). The data represented by the function f can be freely set on the user side, and the data pattern and the function f are compared in step S145.
For example, by defining “function f (LBA) = 1”, data filled with “1” may be used. Further, by defining “function f (LBA) = LBA”, data represented by “LBA” may be used. Further, the logical address space is divided and managed, and data filled with “0” is used for optimization for the first partition, and “1” is used for optimization for the second partition. The filled data may be used. The function f in this case is, for example, “f (LBA) = 0 (LBA <Cap / 2, f (LBA) = 1”.
(LBA ≧ Cap / 2) ”. Cap is the storage capacity of SSD3.

In step S145, if the data read from the LBA in the NAND flash memory 13 is the same as the specific data pattern X (the data size of X is the same as the page size), the SSD controller 11 optimizes the data. Processing may be performed (step S146). This case corresponds to “function f = X”. When the SSD 3 receives this deletion notification, the SSD controller 11 reads the DRAM 12 in the unwritten area read operation.
The data pattern X stored in is transmitted to the host device 2. The example of using a specific data pattern for the deletion notification is particularly effective when the memory system 1 frequently handles the same data.

In this embodiment, for the sake of simplicity, the physical address of the NAND flash memory is directly associated with the LBA in the management information, but for example, JP 2009-217603 A
The LBA is managed using a management unit that is a natural number multiple of 2 or more of the sector size, such as
Instead, the correspondence between the management unit and the NAND flash memory may be described in the management information. The management information describing the correspondence between the physical address of the NAND flash memory and the LBA, such as an LBA management unit, can take various forms, but a table describing the correspondence between the physical address and the LBA (theory). The details of the form taken by the object conversion table do not affect the essence of the present embodiment.

In this embodiment, when the data received from the host device is written in the NAND flash memory, the data itself (plain text) received from the host device is recorded. However, the plain text is randomized to obtain the NAND flash memory. The present embodiment can also be applied when storing in a memory or encrypting plaintext and storing it in a NAND flash memory. At this time, the comparison of data in step S145 is preferably performed on plaintext after decryption, but may be performed on randomized data or encrypted data before decryption.

[Seventh Embodiment]
In the seventh embodiment, if the data read from the SSD in a certain logical address area from the SSD is the same as the comparison source data, the host apparatus sends a deletion notification to the logical address area. Then, by rewriting the management table so that the logical address corresponding to the logical address area corresponds to the “comparison source data area” in which the physical address on the NAND flash memory is pseudo-assigned, a repetitive pattern is obtained. The user data read operation is speeded up to increase the number of free blocks that can be used by the SSD.

[1. Memory system configuration]
When the data (comparison source data) stored in the comparison source data area repeatedly exists in the physical address area of the NAND flash memory, the LB corresponding to the physical address area
A (logical address) is all associated with the comparison source data area according to this embodiment. As a result, a physical block corresponding to “(data amount of repetitive data pattern) − (data amount of repetitive data area)” becomes a free block at the maximum, so that the free blocks usable by the SSD controller can be increased. . The comparison source data area is allocated on a RAM such as a DRAM, an LSI or RAM in an SSD controller, or a NAND flash memory. Each case will be described below. In the present embodiment, for simplicity, it is assumed that the minimum management unit of the NAND flash memory is equal to the physical page size (generally, the minimum management unit may be a natural number multiple of the physical page size).

<When the comparison source data area is on DRAM>
FIG. 24 is a block diagram of the memory system 1 when the comparison source data area is allocated on the DRAM 12. Hereinafter, for simplicity, the individual chips of the NAND flash memory 13 are omitted from the block diagrams.

The DRAM 12 includes a storage area 12D for storing comparison source data. The NAND flash memory 13 includes a storage area 13E for storing backup comparison source data (hereinafter referred to as backup comparison source data). In the storage area 13E of the NAND flash memory 13, initial data such as data filled with “0” at the time of factory shipment is written. The storage area 12D of the DRAM 12 and the storage area 13E of the NAND flash memory 13 are equal in size. When the SSD 3 starts up, the SSD controller 1
1 reads the backup comparison source data to the DRAM 12.

The present embodiment is characterized by an address mapping method for physical addresses as compared to the first embodiment. FIG. 25 is a diagram for explaining a physical address mapping table. FIG.
"H" means hexadecimal notation.

A physical block ID and a physical page address are assigned to all the physical pages of the NAND flash memory 13, and all the physical pages can be distinguished from each other. In the first embodiment, only the physical page of the NAND flash memory 13 is allocated to the physical address. However, in this embodiment, the storage area 12D (comparison source data area) on the DRAM 12 is the same as the NAND flash memory. Assigned as a physical address. The management unit of the comparison source data area is equal to the minimum management unit of the NAND flash memory 13, that is, in this embodiment, equal to the physical page size.

The pseudo physical block ID of the comparison source data area is N by the address mapping method.
It can be distinguished from the physical block ID of the AND type flash memory 13. The physical address area of the DRAM 12 to which the comparison source data area is allocated is divided and managed for each physical page size of the NAND flash memory, and a pseudo physical page address is assigned to these management units in ascending order of the physical address of the DRAM 12. Are assigned in ascending order. When the size of the comparison source data area is larger than the physical block size of the NAND flash memory 13, the physical address area of the DRAM 12 to which the comparison source data area is allocated is divided for each physical block size of the NAND flash memory 13, DRAM1
The pseudo physical block IDs are assigned in ascending order to the physical addresses of 2 in ascending order, and the physical page address is repeated from 0 to the maximum physical page address. The physical block ID assigned to the comparison source data area is registered in the active block table from the factory shipment stage.

The size of the comparison source data area is a natural number multiple of the minimum management unit of the NAND flash memory 13, that is, in this embodiment, is a natural number multiple of the physical page size.
The size of the comparison source data area may be larger or smaller than the physical block size of the NAND flash memory 13. Hereinafter, for the sake of simplicity in this embodiment, the size of the comparison source data area is assumed to be one physical page (FIG. 26). When the size of the comparison source data area is for a plurality of physical pages, for example, the comparison source data area may be divided and processed for each physical page, so the essence of the embodiment is the same.

<When the comparison source data area is on the SSD controller>
FIG. 27 is a block diagram of the memory system 1 when the comparison source data area is allocated on the SSD controller 11.

A storage area 11A (comparison source data area) for storing comparison source data in the SRAM (or other RAM) or LSI (or other ROM) on the SSD controller 11
Is assigned. When the storage area 11A is a RAM, as in the case of a DRAM, the SSD controller 11 reads backup comparison source data to the storage area 11A when the SSD 3 is activated. When the storage area 11A is a ROM, initial data such as data filled with “0” at the time of factory shipment is written in the ROM.

As shown in FIG. 28, the physical address mapping table method is the same as that of the DRAM described above. Hereinafter, for the sake of simplicity in this embodiment, the size of the comparison source data area is assumed to be one physical page (FIG. 29).

<When comparison source data area is on NAND flash memory>
FIG. 30 is a block diagram of the memory system 1 when the comparison source data area is allocated on the NAND flash memory 13.

In a partial area of the NAND flash memory 13, a storage area 1 for storing comparison source data
3E (comparison source data area) is allocated. In the comparison source data area of the NAND flash memory 13, initial data such as data filled with “0” at the time of factory shipment is written.

FIG. 31 is a diagram for explaining a physical address mapping table. A physical block ID and a physical page address are assigned to all physical pages of the NAND flash memory 13, and a part of the physical page group is assigned as a comparison source data area. The data management method such as the management unit in the comparison source data area is the same as the area on the other NAND flash memory 13. The size of the comparison source data area is a natural number multiple of the minimum management unit of the NAND flash memory 13, that is, in this embodiment, is a natural number multiple of the physical page size. The start address of the comparison source data area is equal to the start address of any physical page (must not be in the middle of a physical page). The physical block ID assigned to the comparison source data area is preferably registered in the active block table from the factory shipment stage.

The size of the comparison source data area may be larger or smaller than the physical block size of the NAND flash memory 13. Hereinafter, for the sake of simplicity in this embodiment, the size of the comparison source data area is assumed to be one physical page (FIG. 32). If the size of the comparison source data area is for a plurality of physical pages, for example, the comparison source data area may be divided and processed for each physical page.
The essence of the embodiment is the same.

The case where the comparison source data area is allocated on the DRAM 12, the SSD controller 11 or the NAND flash memory 13 has been described above. In any case, the comparison source data area is a normal NAND flash memory. Physical address mapping equivalent to that of the physical page is performed and can be handled in the same way. Therefore, the following description can be carried out in the same manner regardless of the portion (DRAM 12, SSD controller 11, or NAND flash memory 13) where the comparison source data area is secured.

Next, the configuration of the host device 2 will be described. FIG. 33 is a block diagram of the host device 2. The DRAM 23 includes a storage area 23E for storing comparison source data instead of the function f. Other configurations are the same as those in FIG. 3 of the first embodiment.

The CPU 20 of the host device 2 transmits a read command for reading the comparison source data to the SSD 3 when the host device 2 is started or when the optimization program is started, and the comparison source data read from the SSD 3 is stored in the storage area of the DRAM 23. 23E. For reading, a read command may be used, or another command such as a vendor-specific command may be used. Alternatively, the comparison source data is stored in another external storage device connected to the host device 2, and the CPU 20 of the host device 2 is activated when the host device 2 is activated or when the optimization program is activated.
However, the comparison source data may be read from the external storage device.

[2. Operation]
<2-1. NAND organization processing>
FIG. 34 is a flowchart showing NAND organization processing of SSD3. The SSD controller 11 sets the column number i = 0 and the free space accumulated amount s = 0 (step S160). Subsequently, the SSD controller 11 reads the first row of the active block table from the DRAM 12 and obtains the physical block ID (step S161). The physical block ID acquired in step S161 is referred to as a current physical block ID.

Subsequently, the SSD controller 11 determines whether or not the current physical block ID includes the comparison source data area based on the address mapping of FIG. 26, FIG. 29, or FIG. 32 (step S162). When the current physical block ID includes the comparison source data area, the SSD controller 11 increments the current physical block ID and sets the work target as the next physical block ID.
(Step S163). When the current physical block ID does not include the comparison source data area, the active page table is read from the DRAM 12 and all the physical page addresses included in the current physical block ID are acquired, as in FIG. 8 of the first embodiment (step S16
4). In this way, the comparison source data area is excluded from the NAND organization. The comparison source data area is preferably excluded from the NAND organization, but the comparison source data area is
It is good also as D rearrangement object. The subsequent flow is the same as in FIG. 8 of the first embodiment.

<2-2. Deletion notification processing>
Next, the SSD 3 deletion notification process will be described. FIG. 35 is a flowchart showing the deletion notification process of the SSD 3. The SSD 3 receives a deletion notification from the host device 2 (step S180). The deletion notification includes a deletion command and a logical address.

Subsequently, the SSD controller 11 reads the active page table from the DRAM 12 (step S181). Subsequently, the SSD controller 11 associates the logical address area (deletion target logical area) to be deleted designated by the deletion notification with the comparison source data area in the active page table (step S182). In this embodiment,
Since the size of the comparison source data area is one physical page, the deletion target logical area is one logical page.

Subsequently, the SSD controller 11 reads the active block table from the DRAM 12 (step S183). Subsequently, the SSD controller 11 refers to the active block table, deletes the physical block having no physical block ID in the active page table, and adds it to the free block table (step S184). As described above, the SSD 3 can secure a free block not only by NAND arrangement but also by deletion notification processing.

<2-3. Read operation>
Next, the reading operation of the SSD 3 will be described. FIG. 36 is a flowchart showing the read operation of the SSD 3. The SSD 3 receives a read command from the host device 2 (
Step S190). The read command includes a read command and a logical address.

Subsequently, the SSD controller 11 reads the active block table from the DRAM 12 (step S191), and refers to the physical block ID corresponding to the logical address received from the host device 2. Subsequently, the SSD controller 11 reads the active page table (step S192), and performs a corresponding physical address read operation based on the address mapping of FIG. 26, FIG. 29, or FIG. 32 (step S193). At this time, when the read target is the comparison source data area on the DRAM 12 or the SSD controller 11, the read operation is performed at a high speed, so that the read performance is improved. And S
The SD controller 11 transmits the read data to the host device 2 (step S19).
4).

<2-4. Optimization process>
Next, the optimization process of the host device 2 will be described. FIG. 37 is a flowchart showing optimization processing of the host device 2. The optimization flag setting process is the same as that of the host apparatus 2 described so far.

The CPU 20 of the host device 2 monitors the optimization flag (step S200), and if it is confirmed that the optimization flag is set, the current logical address is set to 0 (step S20).
1). Subsequently, the CPU 20 transmits to the SSD 3 a read command for reading data for the page size from the current logical address, that is, one logical page from the current logical address (step S202). On the other hand, the SSD 3 performs read processing in accordance with the read command received from the host device 2 and transmits read data to the host device 2.

Subsequently, the CPU 20 receives read data for one logical page from the SSD 3 (step S203). Subsequently, the CPU 20 determines whether or not the read data matches the comparison source data stored in the storage area 23E of the DRAM 23 (step S204).
). If the read data matches the comparison source data, the CPU 20 issues a deletion notification to the read logical address area (step S205). In contrast, the SSD 3 performs the deletion notification process described above.

Subsequently, the CPU 20 adds the page size to the current logical address (step S2).
06). Subsequently, the CPU 20 performs steps S202 to S2 until all logical pages are read.
The process of 06 is repeated (step S207). The LBA area to be optimized may be the entire LBA area or a part of the LBA area.

The logical page group subjected to the deletion notification process by this optimization process is associated with the physical address of the comparison source data area in the SSD 3. When the host device 2 sends a read command to the SSD 3 regarding the logical page on which the deletion notification process has been performed, the comparison source data is changed to SS.
Since the data is received from D3, the data in the SSD viewed from the host device 2 is the same as before the optimization process. On the other hand, when the comparison source data area exists in the DRAM or SSD controller on the SSD 3 side, the read operation of the comparison source data is performed at a high speed, so that the performance of reading the logical page to be deleted notification is higher than that before the optimization process. Increase.

[3. Effect]
According to the seventh embodiment described in detail above, the following effects can be obtained.

Regarding the logical address area that is the same as the comparison source data among the valid user data on the host device 2, the host device 2 sends a deletion notification to the SSD 3. As a result, the logical address to which the deletion notification is made is associated with the physical address of the comparison source data area in the SSD 3. Then, the user data subjected to the deletion notification process is transferred to the DRAM or SSD on the SSD 3 side.
By reading from the comparison source data area existing in the controller, it is possible to speed up the reading operation of the user data.

Further, since the free block can be secured by performing the optimization process, the performance of the write operation is improved after the optimization process, and the reliability deterioration of the SSD 3 is reduced.
The effect of the present embodiment increases as the number of pieces of user data valid on the host device 2 matches the comparison source data increases.

[4. Examples]
Example 1
The second embodiment may be applied to the eighth embodiment. That is, in the first embodiment, as in the second embodiment, the file list is acquired by accessing the management area of the OS in the host device, and the optimization process is performed on a file basis.

FIG. 38 is a flowchart illustrating the optimization process of the host device 2 according to the first embodiment. The process of receiving read data from the SSD 3 in file units is the same as that in FIG. 16 of the second embodiment.

Subsequently, the CPU 20 of the host apparatus 2 determines whether or not a logical page that matches the comparison source data exists in the read file (step S215). If there is a logical page that matches the comparison source data, the CPU 20 notifies the corresponding logical page of deletion (step S216). In contrast, the SSD 3 performs the deletion notification process described above. The subsequent processes are the same as those in FIG. 16 of the second embodiment.

As described above in detail, in the first embodiment, in the optimization process, the read operation is not performed on the entire logical address space, and the read operation is performed only on the valid file on the host device 2. For this reason, since the amount of data to be optimized can be reduced, the time required for the optimization process can be shortened. The files to be optimized may be all files or some files.

(Example 2)
The comparison source data on the SSD and the comparison source data on the host device may be rewritten after factory shipment. FIG. 39 is a flowchart illustrating the comparison source data rewriting process of the host device 2 according to the second embodiment.

For example, in response to a user instruction, the CPU 20 of the host device 2 rewrites the old comparison source data stored in the storage area 23E of the DRAM 23 with new comparison source data (step S220).
). Subsequently, the CPU 20 transmits a rewrite command for rewriting the comparison source data and the comparison source data to the SSD 3 (step S221). The rewrite command for rewriting the comparison source data may be a normal write command or other commands such as a vendor-specific command.

FIG. 40 is a flowchart illustrating the comparison source data rewriting process of the SSD 3 according to the second embodiment. The SSD 3 receives the comparison source data rewrite command and the comparison source data from the host device 2 (step S230).

When the comparison source data of the SSD 3 is rewritten, it is necessary to prevent inconsistency in the data of the NAND flash memory 13. Therefore, the SSD controller 11 writes the old comparison source data in the logical page associated with the old comparison source data (step S231).
For example, the SSD controller 11 converts the old comparison source data into the NAND flash memory 13.
Duplicate the number of logical pages associated with the old comparison source data in the logical address area of
The old comparison source data is written in these logical pages. Subsequently, the SSD controller 11
The comparison source data area of the AND flash memory 13 or DRAM 12 is rewritten with new comparison source data (step S232).

In this embodiment, for the sake of simplicity, various physical addresses are directly associated with the LBA in the management information. However, for example, as shown in Japanese Patent Application Laid-Open No. 2009-217603, a management unit that is a natural number multiple of 2 or more of the sector size is used. It is also possible to manage the LBA by using the management information and describe the correspondence between the management unit and the NAND flash memory instead of the LBA. LB
The management information describing the correspondence between the physical address of the NAND flash memory and the LBA, such as the management unit of A, can take various forms, but a table describing the correspondence between the physical address and the LBA (theory). The details of the form taken by the object conversion table do not affect the essence of the present embodiment.

In this embodiment, when the data received from the host device is written in the NAND flash memory, the data itself (plain text) received from the host device is recorded. However, the plain text is randomized to obtain the NAND flash memory. The present embodiment can also be applied when storing in a memory or encrypting plaintext and storing it in a NAND flash memory.

[Eighth Embodiment]
In the eighth embodiment, the optimization processing related to the comparison source data shown in the seventh embodiment is performed with the SSD.
3 is doing it itself. The configuration of the comparison source data area included in the SSD 3 is the same as that of the seventh embodiment, and the comparison source data area included in the SSD 3 includes a RAM such as a DRAM, an SSD.
It is allocated on the LSI or RAM in the controller, or on the NAND flash memory. Similarly to the seventh embodiment, for the sake of simplicity in this embodiment, the size of the comparison source data area is assumed to be one physical page.

FIG. 41 is a flowchart showing the optimization process of the SSD 3 according to the eighth embodiment.
When the SSD controller 11 confirms that the optimization flag is set in the storage area 12C of the DRAM 12 (step S240), the SSD controller 11 sets the first line of the active page table as the current line (step S241). Subsequently, the SSD controller 11 reads the current row of the active page table from the DRAM 12 (step S242).

Subsequently, the SSD controller 11 determines whether or not the physical address indicated by the current row is the comparison source data area (step S243). When the physical address indicated by the current row is the comparison source data area, the optimization process is not performed on the physical address. On the other hand, if the physical address indicated by the current row is not the comparison source data area, the physical page indicated by the current row is read from the NAND flash memory 13 (step S244).

Subsequently, the SSD controller 11 determines whether or not the read data matches the comparison source data (step S245). When the read data matches the comparison source data, the SSD controller 11 pseudo-assigns the physical block ID assigned to the comparison source data area to the current row of the active page table and the comparison source data area. The allocated physical page address is written (step S246). If the read data does not match the data in the comparison source data area, the optimization process is not performed on the physical page.

Thereafter, as in FIG. 23 of the sixth embodiment, optimization processing is performed on all rows of the active page table. The optimization process may be performed for some rows. Next, SSD
The controller 11 updates the active block table and the free block table (step S250). That is, the SSD controller 11 refers to the active block table, and deletes a physical block having no element in the active page table from the active block table among the areas other than the comparison source data area of the NAND flash memory 13 and free blocks. Add to table. If the interrupt instruction is urgent, the active block table and free block table may be updated later.

The read operation and the NAND organizing operation of the SSD 3 are the same as those in the seventh embodiment.

As described above in detail, according to the eighth embodiment, when there are a plurality of physical pages having the same data as the data in the comparison source data area, a plurality of logical addresses (LBA) are converted into a single comparison source by the optimization process. It will be associated with the data area. If the physical page associated with the logical address is released, and a new free block can be secured as a result, the performance of the write operation is improved after the optimization process, and the reliability of the SSD 3 is degraded. Is reduced.

If the comparison source data area is secured on the RAM or the SSD controller,
When the host device 2 transmits a read command to the SSD 3 with respect to the logical address associated with the comparison source data area, the read operation is performed at a higher speed than the read operation accompanied by the reading of the NAND flash memory 13. This improves read performance. The effect of the present embodiment increases as the data that matches the comparison source data area among the data written in the NAND flash memory 13 increases.

As in the seventh embodiment, the comparison source data on the SSD may be rewritten after factory shipment. In this case, the old comparison source data is duplicated by the number of logical pages associated with the old comparison source data in the logical address area of the NAND flash memory 13, and the old comparison source data is written to these logical pages. Thereafter, the comparison source data area of the NAND flash memory 13 or the DRAM 12 is rewritten with new comparison source data, so that no contradiction occurs in the data.

Further, the condition for setting the optimization flag is not limited to the case where the host apparatus 2 explicitly instructs it. That is, as in the sixth embodiment, the host apparatus 2 receives a certain period (for example, 1
(0 second) An optimization flag may be set when there is no access, or an optimization flag is set when the number of free blocks falls below a specified value (for example, a specified value determined in advance at the time of design). It may be.

In this embodiment, for the sake of simplicity, various physical addresses are directly associated with the LBA in the management information. It is also possible to manage the LBA by using the management information and describe the correspondence between the management unit and the NAND flash memory instead of the LBA. LB
The management information describing the correspondence between the physical address of the NAND flash memory and the LBA, such as the management unit of A, can take various forms, but a table describing the correspondence between the physical address and the LBA (theory). The details of the form taken by the object conversion table do not affect the essence of the present embodiment.

In this embodiment, when the data received from the host device is written in the NAND flash memory, the data itself (plain text) received from the host device is recorded. However, the plain text is randomized to obtain the NAND flash memory. The present embodiment can also be applied when storing in a memory or encrypting plaintext and storing it in a NAND flash memory. At this time, the comparison of data in step S245 is desirably performed on the plaintext after decryption, but may be performed on the randomized data or the encrypted data before decryption.

[Ninth Embodiment]
In the first to sixth embodiments, the function f (L in the host device (information processing device) or SSD
BA) is used for optimization processing. In the ninth embodiment, a method for determining the function f (LBA) at the time of designing an optimization program, at the time of product design, before factory shipment, or after factory shipment will be described. In particular, when the OS and system configuration to be installed on the SSD are determined, the optimal function f (LBA) for these systems should be set at the time of optimization program design, product design, or factory shipment. Is desirable.

Example 1
FIG. 42 is a diagram illustrating an example of logical addresses and data. The logical address area of the SSD is recognized by the host device by the logical address (LBA). In the logical address, one sector is the minimum management unit, and one sector is, for example, 512 bytes. As shown in FIG.
Logical addresses (0000h, 0001h, 0002h,..., Ba02h,...) Are assigned to each sector.

FIG. 42 shows an example of a data pattern that is written on an SSD before shipment in a certain SSD-equipped personal computer when the OS and other software are installed in advance and then shipped. Alternatively, the function f (LBA) may be determined as follows assuming (simulating) data that the end user will write even though it is not actually written.

The SSD is fully formatted before shipping, for example, and 0 data is written in all logical address areas. Thereafter, the OS and other software are installed in the SSD, and non-zero data is written in some sectors (LBA = 0000h, 0001h, 0003h, ba02h, etc. in FIG. 42). For a logical address that has not been written, data of one sector length (0-fill sector) filled with “0” remains written (LBA = 0002h, 000 in FIG. 42).
4h, 0005h, ba03h, etc.). After LBA = ba03h, no data is written and all 0-fill sectors remain.

By reading all logical addresses and comparing them with 0-fill sectors, 0-fill as shown in FIG.
The existence probability distribution of the sector can be derived. Further, when comparing, the total number of sectors that are 0-fill sectors is divided by the total capacity of all logical address areas, so that the net ratio of 0-fill sectors can be derived.

By setting the function f (LBA) = 000... 0 (fixed value, 0 is a repetition of 512 bytes), the block corresponding to the total number of sectors that is 0-fill sector at maximum is released and free block Can be secured.

Further, the optimization process may be performed only when the ratio of the net 0-fill sector is significant (for example, when it is 5% or more). The function f (LBA) may be rewritten only when the ratio of the net 0-fill sector is significant (for example, 5% or more). The criterion for determining whether the proportion of 0-fill sectors is significant is determined as follows, for example.

(1) The ratio of the free block to the total capacity of the free block for ensuring that the write response time is less than a predetermined value (for example, the specification presented in the specification) (2) Write Amplification (WA) is less than the predetermined value The ratio of the free block to the total capacity for guaranteeing that the above-mentioned WA is defined as follows.

WA = (total size of data written to NAND flash memory) / (total size of write data transferred from host device to SSD)
Or
WA = (total erase count of block) × (block size) / (total size of write data transferred from host device to SSD)
The “total size of data written in the NAND flash memory” preferably includes not only user data but also data other than user data such as management information, ECC redundant data, and an encryption key.

(Example 2)
FIG. 44 is an example in which the function f (LBA) does not become a fixed value. There are cases where the repeated pattern frequently occurs depending on the logical address area, such that a certain data pattern frequently occurs in a certain logical address area and another data pattern frequently occurs in a certain logical address area. As shown in FIG. 44, many 0-fill sectors exist in the first half of the logical address area, while 1-fill data (data in which all bits are filled with “1”, 16
Let us consider a case where there are a large number of data filled in with f.

The existence probability distribution of the 0-fill sector and the existence probability distribution of the 1-fill sector are as shown in FIG.
In this case, the function f (LBA) = 000... 0 (fixed value) or the function f (LBA) =
Instead of 111 ... 1 (fixed value), "function f (LBA) = 000 ... 0h (LB
A ≦ c000h), f (LBA) = 111... 1h (LBA ≦ c000h) ”.
It is preferable to set LBA) to different values in the two logical address areas because the number of sectors that can be released increases.

(Example 3)
FIG. 46 shows a case where frequently occurring data is different in the three logical address areas. The existence probability distribution of the 0-fill sector, the existence probability distribution of the 320a-fill sector (one-sector length data filled with 320a data), and the existence probability distribution of the 1-fill sector are as shown in FIG. in this case,
“Function f (LBA) = 000... 0h (LBA ≦ 7fffh), f (LBA) = 320a3
20a ... 320ah (8000h≤LBA≤c000h), f (LBA) = 111 ... 1h (
It is preferable to set the function f (LBA) to different values in the three logical address areas as the number of sectors that can be released increases as in (LBA ≦ c001h) ”.

Example 4
In the fourth embodiment, a management unit obtained by dividing the logical address area for each size L is used. The size L is not less than the minimum unit of LBA. The function f (LBA) takes a constant value inside the management unit, but the function f (LBA) may be different between different management units. In the first management unit, function f (LBA) = D0 (fixed value), and in the next management unit, function f (LBA) = D1 (fixed value),
... and so on. The function f (LBA) in this case is expressed as follows.

“F (LBA) = D0 (0 ≦ LBA ≦ L−1, f (LBA) = D1 (L ≦ LBA ≦ 2L)
−1), f (LBA) = D2 (2L ≦ LBA ≦ 3L−1,..., Dk (kL ≦ LBA ≦
(K + 1) L-1), ... "
Dk (k = 0, 1,...) Is an arbitrary data pattern having a size L.

The method of selecting Dk (k = 0, 1,...) Is arbitrary, and the optimal function f (LBA) is determined by selecting the optimal Dk. SSD pre-shipment inspection device for various Dk candidates,
The CPU of the information processing apparatus or the SSD controller performs scoring by the following method.

(1) Read data for each sector (or data may be read for each integral multiple of the sector, or data may be read for each management unit)
(2) It is determined whether or not the data read in (1) is equal to the function f (LBA) (when data is read every integer multiple of the sector, all the sectors in the data are converted to the function f (LBA). Determine if they are equal)
(3) Score 1 point if they are equal, and 0 points if they are not equal (4) Assign points to all the logical address areas of the SSD and calculate the total number of points (typically only some areas) Scoring may be done for
The Dk having the highest score is the optimum, and the optimum function f (LBA) is determined using this Dk. For example, if L = 128 MiB (MiB = 1,048,576 bytes), the entire logical address area of the SSD is equally divided into 128 MiB from the start, each partition is read, and all the data in the partition is equal to the function f (LBA). If equal, add 1 point and calculate the total score. For example, (D0, D1
, ...) = (0000 ..., 0000 ..., ...), (FFFF ..., FFFF ..., ...
), (FFFF ..., 0000 ..., ...), (0000 ..., FFFF ..., ...) Select as (D0, D1,...) To determine the function f (LBA).

(Example 5)
In the fourth embodiment, the section divided into the fixed length is used as the management unit. However, the section divided into the variable length may be used as the management unit. The function f (LBA) in this case is expressed as follows.

“F (LBA) = D0 (L0 = 0 ≦ LBA ≦ L1-1, f (LBA) = D1 (L1 ≦ L
BA ≦ L2-1), f (LBA) = D2 (L2 ≦ LBA ≦ L3-1,..., Dk (Lk
≦ LBA ≦ L (k + 1) −1),...
Dk (k = 0, 1,...) Is an arbitrary data pattern having a size L. “L0 <L
1 <L2 <... Lk <... ".

(Example 6)
Example 6 is a case where there are few sectors filled with the same data, and thus sufficient effects cannot be obtained even if the above-described method is applied, but there is a certain correlation between data and LBA. (FIG. 48). In such a case, it is preferable to use a more general method for determining the function f (LBA). For example, a plurality of function f (LBA) candidates are prepared, and each f (L
The above scoring is performed on BA), and f (LBA) with the highest score is changed to the optimum f (LB).
Determine as A). In the sixth embodiment, as shown in FIG. 49, a sector having a certain correlation between data and LBA is released.

(Example 7)
For example, for two types of parameters a and b, “function f (LBA) = a × LBA + b”
Is used as a candidate for the function f (LBA). Then, the highest score (a, b) is obtained, and the highest score (a, b) is used to obtain the function f (
LBA) may be determined.

FIG. 50 is a diagram illustrating a method for determining the function f (LBA) according to the seventh embodiment. FIG. 50 illustrates a case where the parameter b is b1, b2, and b3. When parameter a is changed,
(A, b) = (a0, b2) has the highest score. Therefore, “function f (LBA) = a0
XLBA + b2 "is calculated.

More generally, when the LBA is a general 48-bit LBA in the SATA standard, a vector with 48 columns of LBA = (d0, d1, d2,..., D47) t (
t is a transposed symbol), A is a 512 × 48 matrix, and b is a vector having 4096 columns (1 sector = 512 bytes = 4096 bits), so that “function f (LBA) = A × LBA + b
The function f (LBA) may be acquired in a vector format having 4096 columns. In this case, the first column of the vector f (LBA) is the first bit data, and the second column is the second bit. There is a corresponding relationship of.

More generally, when LBA is described with n bits, LBA = (d0, d1, d2
,..., Dn−1) t is a vector with n columns, A is an n × m matrix, and b is a vector with m columns, so that “function f (LBA) = A × LBA + b”. Function f
(LBA) may be acquired in a vector format with m columns.

The processing of the controller described in the above embodiments can also be realized by a program. As the flowchart of the program, the flowcharts shown in the above embodiments can be used.

In each of the above embodiments, the SSD has been described as an example of the nonvolatile storage device 3, but the SS
The present invention can be widely applied to non-volatile storage devices other than D, such as a hard disk drive (HDD) using a magnetic medium (platter) or a hybrid hard disk drive using both a NAND flash memory and a magnetic medium (platter) ( It is also possible to apply to a hybrid HDD).

Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 1 ... Memory system, 2 ... Host apparatus, 3 ... SSD, 10 ... Interface controller, 11 ... SSD controller, 12 ... DRAM, 13 ... NAND flash memory, 20 ... CPU, 21 ... South bridge, 22 ... North bridge, 23 ... DRAM, 24
Display controller, 25 Display, 26 BIOS-ROM, 27 LAN controller, 28 USB controller, 29 Keyboard, 30 Mouse, 31 Optical drive, 40 Information processing unit, 41 Motherboard, 42 Power supply device 50... Information processing device main body 51. Display unit 52. Power switch 53. Touchpad.

Claims (13)

A memory system having a nonvolatile storage device having a storage area for storing first data, and an information processing device communicating with the nonvolatile storage device,
The information processing apparatus includes:
Reading data from a first logical address area of the non-volatile storage device;
When the read data corresponding to the first logical address area is the same as the first data, a deletion notification for associating the first logical address area with the storage area is sent to the nonvolatile storage device. A first control circuit to perform,
The nonvolatile memory device is
A non-volatile storage medium;
A management table that associates logical and physical addresses;
Rewriting the management table to associate the logical address specified by the deletion notification with the storage area;
A second control circuit that transmits the first data to the information processing device when a logical address included in a read command received from the information processing device specifies the storage area;
A memory system comprising: The memory system according to claim 1, wherein the first control circuit executes a process related to the deletion notification when receiving an instruction from a user. The memory system according to claim 1, wherein the first control circuit executes a process related to the deletion notification when a predetermined time is reached. The second control circuit transmits a command for executing processing related to the deletion notification to the information processing device,
The memory system according to claim 1, wherein the first control circuit executes processing related to the deletion notification when the command is received from the nonvolatile memory device. A non-volatile storage device that communicates with an information processing device,
A non-volatile storage medium;
A storage area for storing the first data;
A management table that associates logical and physical addresses;
Reading data from the non-volatile storage medium;
When the read data is the same as the first data, the management table is rewritten to associate the logical address of the read data with the storage area,
A control circuit that transmits the first data to the information processing device when a logical address included in a read command received from the information processing device specifies the storage area;
A non-volatile memory device comprising: The nonvolatile memory device according to claim 5, wherein the control circuit executes a process of rewriting the management table when there is no access from the information processing device for a certain period. The nonvolatile memory device according to claim 5, wherein the control circuit executes a process of rewriting the management table when receiving an instruction from the information processing device. 6. The nonvolatile storage device according to claim 5, wherein the control circuit executes a process of rewriting the management table when the number of free blocks of the nonvolatile storage medium falls below a specified value. A method for controlling a nonvolatile storage device having a storage area for storing first data, comprising:
The control method is:
Reading the first data from the non-volatile storage device;
Reading data from a first logical address area of the non-volatile storage device;
Determining whether read data corresponding to the first logical address area is the same as the first data;
When the read data is the same as the first data, notifying the non-volatile storage device of deletion to associate the first logical address area with the storage area;
A control method for a non-volatile memory device, comprising: A program for controlling a nonvolatile storage device having a storage area for storing first data,
The program is stored in an information processing apparatus capable of communicating with the nonvolatile storage device.
A process of reading the first data from the nonvolatile storage device;
A process of reading data from a first logical address area of the nonvolatile storage device;
A process of determining whether read data corresponding to the first logical address area is the same as the first data;
When the read data is the same as the first data, a process of notifying the non-volatile storage device of deletion to associate the first logical address area with the storage area;
A program for running The program according to claim 10, wherein when an instruction is received from a user, processing related to the deletion notification is executed. The program according to claim 10, wherein a process related to the deletion notification is executed when a predetermined time is reached. The program according to claim 10, wherein when the number of free blocks in the nonvolatile storage medium falls below a specified value, the process related to the deletion notification is executed.

Images