KR20110096813A - Storage device and computing system, and data management method thereof - Google Patents

Abstract

The data management method of the present invention comprises the steps of: marking a region to be invalidated in response to an invalidation command input from a host; purging the region to be invalidated based on the marking result; And purging the purge based on the pivot data when sudden power off occurs.

Description

STORAGE DEVICE AND COMPUTING SYSTEM, AND DATA MANAGEMENT METHOD THEREOF

The present invention relates to a storage device and a computing system, and more particularly to a data management method of the storage device and the computing system.

With the advent of the information society, the amount of data that individuals need to store and move is exploding. Due to the increased demand for such information storage media, various types of personal information storage devices have been developed.

Among information storage devices, hard disk drives (HDDs) are widely used due to advantages such as high recording density, high data transfer speed, fast data access time, and low cost. However, since the hard disk drive has a complicated structure composed of mechanical parts, there is a problem that even a small shock and vibration may fail.

Recently, as an information storage device replacing a hard disk drive, a solid state disk (SSD) device employing a semiconductor memory has been developed. Semiconductor disk devices, unlike hard disk drives, have no mechanical configuration. Therefore, the semiconductor disk device can reduce latency and other mechanical driving time and perform a high speed read / write operation as compared with a hard disk drive. In addition, since a semiconductor disk device can reduce errors due to delay time and mechanical friction, there is an advantage of improving reliability of a read / write operation. In addition, since heat and noise are hardly generated during data operation and strong against external shocks, it is evaluated as a storage device suitable for a portable device as compared to a conventional hard disk drive.

An object of the present invention is to provide a storage device and a computing system capable of performing an interrupted invalidation operation from an interrupted part even if a sudden power off occurs due to an unexpected power failure, To provide a way to manage data.

In order to achieve the above object, a data management method according to the present invention comprises: marking an area to be invalidated in response to an invalidation command input from a host; Purging the area to be invalidated based on the marking result; Storing progress information of the purge as pivot data; And when the sudden power off occurs, performing the purge based on the pivot data.

In one embodiment, the step of performing the purge may be performed from the operation interrupted when the sudden power off occurs.

In one embodiment, storing the pivot data may be performed each time the purge is executed.

In one embodiment, storing the pivot data may be performed when the purge is performed a predetermined number of times or more.

In one embodiment, the pivot data may be stored in a nonvolatile memory.

In an embodiment, the storing of the pivot data may store the progress state information of the purge on a page or block basis.

In an embodiment, the storing of the pivot data may be performed at a predetermined time interval.

In one embodiment, the purge step may be performed during an idle period in which no command is input from the host.

In an embodiment, the purging may include merging or copying valid data included in the area to be invalidated to another area of the main storage based on the marking result; And removing data of the area to be invalidated.

In one embodiment, the invalidation command may include a security trim command or a security invalidation command.

In one embodiment, the invalidation command may include start and end addresses of the region to be invalidated.

In order to achieve the above object, a data management method according to the present invention comprises: marking an area to be invalidated in response to an invalidation command input from a host; Determining whether a sudden power off has occurred; Retrieving pivot data when the sudden power off has occurred; Purging the area to be invalidated based on the marking result and the search result of the pivot data; And storing the progress state information of the purge as current pivot data.

In one embodiment, the purge step may be performed from an operation interrupted when the sudden power off occurs.

In one embodiment, storing the pivot data may be performed each time the purge is executed.

In one embodiment, storing the pivot data may be performed when the purge is performed a predetermined number of times or more.

In one embodiment, the pivot data may be stored in a nonvolatile memory.

In an embodiment, the storing of the pivot data may store the progress state information of the purge on a page or block basis.

In an embodiment, the storing of the pivot data may be performed at a predetermined time interval.

In order to achieve the above object, a computing system according to the present invention includes a host; And a storage device operable in response to a control of the host, wherein the storage device includes a main storage unit and a controller configured to control the main storage unit, wherein the controller is in a progress state of an invalidation command input from the host. Information may be stored as pivot data in the main storage unit, and after a sudden power off occurs, the invalidation operation may be re-started from an operation interrupted at the sudden power off based on the pivot data.

In order to achieve the above object, the storage device according to the present invention, the main storage unit; And a controller for controlling the operation of the main storage unit, wherein the controller stores the progress state information of the invalidation command input from the host as pivot data, and after the sudden power off occurs, the pivot data. On the basis of the above, the invalidation operation can be performed again from the operation interrupted at the sudden power-off.

According to the present invention, even if sudden power off occurs during the invalidation operation, the invalidation operation can be performed from the interrupted portion. Therefore, it is not necessary to perform the invalidation operation from the beginning again, and the total time required for invalidation is reduced.

1 is a diagram illustrating a configuration of a computing system according to an exemplary embodiment of the present invention.
FIG. 2 is a diagram illustrating a configuration according to an embodiment of the controller shown in FIG. 1.
3 is a diagram illustrating a configuration according to another embodiment of the controller shown in FIG. 1.
4 is a view for explaining the invalidation operation according to the present invention.
5 is a flowchart illustrating an invalidation operation according to an exemplary embodiment of the present invention.
6 and 7 illustrate the purge operation illustrated in FIG. 5.
8 is a diagram illustrating a configuration of a memory system according to an embodiment of the present invention.
9 is a diagram illustrating a configuration of a computing system according to an embodiment of the present invention.

Exemplary embodiments of the invention may be described in detail below on the basis of reference drawings. The circuit configuration and operation of the storage device of the present invention to be described below are described by way of example, and the storage device of the present invention may be variously changed and changed without departing from the technical spirit of the present invention. For example, in the present invention, an SSD employing a flash memory as a main storage device among semiconductor memories can be described as a storage device. However, this is only an example to which the present invention is applied, and the storage device and its data storage method according to the present invention can be applied to various types of storage devices as well as SSDs.

1 is a diagram illustrating a configuration of a computing system 1000 according to an exemplary embodiment of the present invention.

Referring to FIG. 1, the computing system 1000 of the present invention may include a host 1100 and a storage device 1200. The host 1100 may be configured to control the storage device 1200. The host 1100 may include, for example, a portable electronic device such as a personal / portable computer, PDA, PMP, MP3 player, or the like. When the contents (eg, files) of data stored in the storage device 1200 are deleted from the host 1100, the host 1100 processes metadata related to a file of the deleted contents, thereby removing the contents of the deleted contents. You can invalidate the file.

The host 1100 may notify the storage device 1200 of invalidation (or deletion) of files as necessary. This may be accomplished by the host 1100 sending a specific command to the storage device 1200. Such specific commands may hereinafter be referred to as "invalidity commands". The invalidation command may include information for specifying an area to be deleted (for example, address information). The invalidation command may be used in various names such as a file delete command, a trim command, an unwrite command, and a deletion command.

The processing of the metadata about the file to be deleted may be done by the file system of the host 1100 (eg, FAT, HPFS, NTFS, UFS, FFS, MMFS, VFAT, etc.). For fast operation, the file system can delete only the metadata of the file, not the contents of the file. When any file is deleted, the file system can place one special code in the first character of the file name of the deleted file. For example, the hexadecimal byte code 'E5h' can be placed in place of the first letter of the file name of the deleted file. This is a special tag, which can mean 'This file has been deleted'.

When the metadata of a file to be deleted is processed by the file system, the content of the file to be deleted may be left as valid data in the storage device 1200 while the file system is treated as invalid data. For this reason, the storage device 1200 may recognize the memory block including the data of the deleted file as a valid block. Accordingly, the file system of the host 100 provides an invalidation command (ie, a trim command) to the storage device 1200 so that the contents of the deleted file can be substantially invalidated in the storage device 1200.

The invalidation command may be divided into a command for invalidating normal data and a security invalidation command for invalidating data requiring security (hereinafter referred to as security data). The security invalidation command is also called a secure TRIM command. The security invalidation command may provide a function of completely deleting data deleted by the host 1100 in the storage device 1200. Therefore, according to the security invalidation operation, external access to the security data can be fundamentally blocked.

The security invalidation command was originally defined for data that needs security, but can be applied to general data as well as data that needs security. The invalidation instruction to be described in the present invention and a method of executing the same will be described by taking a security invalidation instruction as an example among the invalidation instructions. However, this relates to an example to which the present invention is applied, and the characteristics of the present invention are not limited to the specific embodiments.

In an exemplary embodiment, the storage device 1200 may constitute a semiconductor disk (Solid State Disk or Solid State Drive, hereinafter referred to as SSD) device. This is an example to which the present invention is applied, and the storage device 1200 may be configured in various forms without being limited to an SSD. For example, the storage device 1200 may be integrated into one semiconductor device, such as a personal computer memory card international association (PCMCIA), a compact flash card (CF), a smart media card (SM, SMC), a memory stick, Multimedia cards (MMC, RS-MMC, MMCmicro), SD cards (SD, miniSD, microSD, SDHC), universal flash storage (UFS), and the like may also be configured.

The storage device 1200 may include a controller 1220 and a main storage unit 1240. The controller 1220 may control the main storage unit 1240 in response to a request from the host 1100. The controller 1220 may be connected to the main storage unit 1240 through a plurality of channels CH1 to CHn.

The main storage unit 1240 may be composed of a plurality of nonvolatile memory chips, for example, a plurality of flash memories. A plurality of flash memory chips may be commonly connected to each of the channels CH1 to CHn. However, this is an example to which the present invention is applied, and the main storage unit 1240 may be configured with other nonvolatile memory chips (eg, PRAM, FRAM, MRAM, etc.) instead of flash memory chips. Alternatively, the main storage unit 1240 may be configured of volatile memory such as DRAM or SRAM. In addition, the main storage unit 1240 may be configured by mixing at least two types of memories.

When the main storage unit 1240 is composed of a plurality of nonvolatile memory chips (eg, a plurality of flash memory needles), the storage device 1200 may maintain the stored data even when the power is cut off. Each of the flash memory chips constituting the main storage unit 1240 may store 1-bit data or M-bit data (M is an integer of 2 or more) per cell.

The invalidation operation according to an embodiment of the present invention may mean a security invalidation operation executed by a security invalidation command. The security invalidation operation may be composed of two steps, Step 1 and Step 2. In operation 1, when an invalidation command is provided from the host 1100, the storage device 1200 may record / store / mark only a location of an area (or files) to be deleted according to the invalidation command. The recording / storing / marking result may be stored in the controller 1220 or the main storage unit 1240.

This operation is called "logging". Subsequently, the storage device 1200 may notify the host 1100 that execution of the requested invalidation command is completed. As a result, the host 1100 recognizes that execution of the invalidation command is completed immediately, and the data of the area to be deleted can be viewed as invalid files on the host 1100 side. Here, step 1 can be executed without being limited in number.

The data of the area (or files) that were logged in step 1 may be deleted / purged later in step 2. In an exemplary embodiment, step 2 may be performed at idle time. In other words, the storage device 1200 does not immediately delete / purge the area of the files to be deleted, but records / stores / marks only the location of the area of the invalid files in the step 1 section, and subsequently invalidates the step 2 section. Delete / Purge data in the area recorded as a file. According to such a configuration, it is possible to process the request of the host 1100 quickly. That is, the response time for the request of the host 1100 may be shortened. In addition, it is possible to improve the performance of the storage device 1200 or the computing system 1000 through a quick response.

The location information of the region (or files) to be deleted may be stored in a RAM, a buffer, or a register provided in the controller 1220 during the logging operation, or may be stored in the main storage unit 1240. The location information of the invalid files recorded / stored / marked in step 1 can be recorded and managed using a bitmap structure.

On the other hand, the flash memory device may face a situation in which a reboot occurs due to a serious error during operation. A typical example is an unexpected power failure (such as a power outage). This power failure is referred to as sudden power off. If a sudden power off occurs during the invalidation operation, the invalidation operation will have to be resumed after the flash memory is rebooted.

The present invention provides a method for performing an invalidation operation from an operation interrupted by a sudden power off in order to shorten the execution time of the invalidation operation resumed after the sudden power off. In an exemplary embodiment, when the storage device 1200 of the present invention deletes / purges the data of an area recorded as an invalid file in the step 1 section in the step 2 section, the pivot data ( pivot data) may be stored in the main storage unit 1240. Pivot data may be configured in the form of a table in the main storage unit 1240, which is called a pivot table. Storage and management of pivot data may be performed by a Flash Translation Layer (FTL), which is system software or firmware executed by the SSD controller 1220.

Since the main storage unit 1240 may be configured of a nonvolatile memory such as a flash memory, the pivot data may be preserved even if the power supply is interrupted. If an error such as sudden power off occurs when the step 2 section of the invalidation operation is not finished, the pivot data will be provided to the controller 1200 during the boot operation. The controller 1200 may analyze the provided pivot data to determine how far the invalidation operation is performed. Then, in resuming the suspended invalidation operation, the controller 1200 resumes the invalidation operation from the interrupted portion without performing the invalidation operation again from the beginning.

According to this configuration, the invalidation operation resumed after the sudden power-off does not need to be performed again from the beginning, and the total time required for the invalidation is reduced.

Meanwhile, the delete / purge operation performed in step 2 of the present invention selectively merges valid data included in an area (ie, an area (or files) to be deleted) marked as an object to be invalidated into another memory block. It may include the process of copying. The region marked as invalidated will be substantially removed after the merge / copy operation on the valid data is performed.

In an exemplary embodiment, the merge / copy operation performed in the present invention may be performed on all data included in the region marked as invalidated, or only for valid data included in the region marked as invalidated. May be performed selectively. When the merge / copy operation is selectively performed only on the valid data included in the region marked as invalidated, the invalid data included in the region marked as invalidated may be excluded from the merge / copy operation. According to such a configuration, the time required for the merge / copy operation will be further reduced. In this case, the data of the area marked to be invalidated will be as if the value of 0xFF is merged / copied into the new block.

FIG. 2 is a diagram exemplarily illustrating a configuration of the controller 1220 illustrated in FIG. 1.

2, the controller 1220A may be configured as an SSD controller when the storage device 1200 constitutes an SSD. The controller 1220A may include a host interface 1222, a flash interface 1224, a processing unit 1226, and a buffer memory 1228. The configuration of the controller 1220A illustrated in FIG. 2 relates to an example to which the present invention is applied and may be changed and modified in various forms. For example, although not shown in FIG. 2, the controller 1220A may further include an error correction circuit for detecting and correcting an error of data stored in the main storage 1240.

The host interface 1222 may provide an interface with the host 1100, and the flash interface 1224 may provide an interface with the main storage unit 1240. The processing unit 1226 may control overall operations of the controller 1220A.

The buffer memory 1228 may be used to temporarily store data to be stored in or read from the main storage 1240. In addition, the buffer memory 1228 can be used to store information of the region (ie, the region to be deleted according to the invalidation command) that is the target of Step 1 described above. The buffer memory 1228 may be composed of volatile memory such as DRAM or SRAM. In an exemplary embodiment, the information of the region targeted for step 1 may be stored using an additional volatile / nonvolatile memory, a register, or the like instead of the buffer memory 1228. The method and configuration for storing the information of the area targeted for step 1 may be configured in various forms without being limited to a specific embodiment.

3 is a diagram exemplarily illustrating a configuration according to another embodiment of the controller 1220 illustrated in FIG. 1.

Referring to FIG. 3, the controller 1220B may be configured as an SSD controller when the storage device 1200 constitutes an SSD. The controller 1220B may include a host interface 1222, a flash interface 1224, a plurality of processing units 1226_1 to 1226_N, and a buffer memory 1228. A case in which a plurality of processing units 1226_1 to 1226_N are provided in the controller 1220B as shown in FIG. 3 is referred to as a multi-core processor, and a case in which one processing unit 1226 is provided in the controller 1220A as shown in FIG. 2. It is called a single core processor.

The configuration of the controller 1220B illustrated in FIG. 3 relates to an example to which the present invention is applied, and may be changed and modified in various forms. For example, although not shown in FIG. 3, the controller 1220B may further include an error correction circuit for detecting and correcting an error of data stored in the main storage unit 1240.

The controller 1220B may perform various operations through the plurality of processing units 1226_1 to 1226_N. After the controller 1220B divides the plurality of control operations by a predetermined number, the controller 1220B may allocate the divided control operations to the plurality of processing units 1226_1 to 1226_N. According to such a configuration, a plurality of control operations can be performed in parallel. In an exemplary embodiment, the plurality of processing units 1226_1 to 1226_N may correspond to each of the plurality of channels CH1 to CHn, and may be configured to perform independent control on each channel. According to such a configuration, even if the controller 1220B is driven by a low frequency clock, the performance of the controller 1220B including the plurality of processing units 1226_1 to 1226_N can be improved.

4 is a view for explaining an invalidation operation according to the present invention, and a detailed configuration of the storage device 1200 according to the present invention is shown.

Referring to FIG. 4, the storage device 1200 may include a main storage unit 1240 and a controller 1220. In the present invention, a case in which the main storage unit 1240 is configured as a flash memory will be described as an example. The main storage unit 1240 may use other nonvolatile memory (for example, PRAM, FRAM, MRAM, etc.) or volatile memory (for example, SRAM, DRAM, etc.) instead of the flash memory.

The main storage unit 1240 may be configured of a plurality of memory cells having a string structure. This set of memory cells is called a cell array. The memory cell array of the main storage unit 1240 may be composed of a plurality of blocks. Each block may consist of a plurality of pages. Each page may be composed of a plurality of memory cells sharing one word line. Memory cells corresponding to one or more pages may correspond to one word line. The main storage unit 1240 performs erase operations in block units, and read and write operations in page units. In another embodiment, the unit of reading and writing may be performed in units of a plurality of pages, or may be performed in units of subpages smaller than one page.

As described above, the main storage unit 1240 configured of the flash memory has a different unit of read / write operation and erase unit. In addition, unlike other semiconductor memory devices, the main storage unit 1240 may not be overwritten. Therefore, the erase operation must be performed before the main storage unit 1240 performs the write operation. This means that the memory block to which new data is to be written must be a memory block in an erased state (ie, an empty state in which no data is stored). By the way, the unit of data written in the flash memory is smaller than the unit of data to be erased. Therefore, in order to map the logical block address (LBA) generated by the file system (normally, the file system is stored in the software form on the host side) to the physical block address (PBA) of the memory block in the erased state. FTL is used. This operation of the FTL is called an address mapping function.

Due to the address mapping function of the FTL, the host 1100 recognizes the storage device 1200 configured as the flash memory as if it is a hard disk driver or an SRAM, and can access the flash memory device in the same manner as the hard disk driver. The address mapping information of the flash memory may be managed in the form of a mapping table. The mapping table may be stored in the main storage unit 1240, and the management of the mapping table may be performed by the FTL.

When the invalidation command input from the host 1100 is input, an area to be invalidated during the step 1 period may be marked. Here, the invalidation command may include a security invalidation command (eg, a secure TRIM). The data of the area recorded as an invalid file during the step 1 section is deleted / purged in the step 2 section.

As will be described in detail below, the storage device 1200 of the present invention may perform an invalidation operation from an interrupted operation even when a sudden power off occurs due to an unexpected power failure during the invalidation operation. It has a configuration for storing pivot data during invalidation operation. For example, the storage device 1200 of the present invention may store pivot data in the main storage unit 1240 at each stage of deletion / purge when performing deletion / purge in step 2. The pivot data may be configured in the form of a table in the main storage unit 1240, which is called a pivot table 1250. Pivot data storage and management operations for the pivot table 1250 may also be performed by the FTL.

In an exemplary embodiment, the pivot data may be configured to indicate whether each step performed in the invalidation operation (eg, each step of the merge / copy operation and the erase operation performed in step 2, etc.) is completed. . The pivot data may store the merge / copy result performed in step 2 in page or block units, and the delete / purge result performed in step 2 may be stored in block units. Pivot data may be stored in primary storage 1240 at a given time (eg, at a predetermined time interval) or as needed. The stored pivot data may be loaded from the main storage unit 1240 to the controller 1220 during the reboot operation. Meanwhile, a form and a storage method of pivot data applicable to the present invention may be configured in various forms without being limited to a specific form.

The controller 1220 may be configured to control the main storage unit 1240 when there is an access request from the host 1100. As shown in FIG. 4, the controller 1220 may include a processing unit 1226 and a work memory 1227. The work memory 1227 is also called an internal memory. The FTL may be stored in the work memory 1227.

5 is a flowchart illustrating an invalidation operation according to an exemplary embodiment of the present invention. Hereinafter, the operation of a computing system according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

4 and 5, in operation S1000, the storage device 1200 receives an invalidation command from the host 1100. Then, the step 1 operation is started in step S1100. The invalidation command received in operation S1000 may be used to delete files having invalid contents among files stored in the storage device 1200. In this case, the invalidation command may include a security invalidation command, and the invalidation command may include information indicating an area of files having invalid contents. The information indicating the area of the files with invalid content may include, for example, the start and end addresses of the area to be invalidated.

When the operation of step 1 starts, in step S1200, the controller 1220 of the storage device 1200 records / stores / marks location information of an area (or files) to be invalidated in response to an invalidation command provided from the host 1100. do. In an exemplary embodiment, the information on whether the memory block is invalidated may be marked in the form of file system information (eg, FAT, HPFS, NTFS, UFS, FFS, MMFS, VFAT, etc.). Here, the form in which the information on whether the memory block is invalidated is marked may be marked not only in the file system information form but also in a separately defined data form. The result recorded / stored / marked in operation S1200 may be stored in a RAM, a buffer, or a register provided in the controller 1220, or may be stored in the main storage 1240. The controller 1220 may notify the host 1100 of a response signal (eg, an ACK signal) after the location information or the like of the area (or files) to be invalidated is recorded / stored / marked. When the host 1100 receives a response signal from the storage device 1200, the host 1100 recognizes that execution of the invalidation command is completed immediately. At this time, the data of the region to be deleted is seen as invalidated on the host 1100 side.

In this case, the invalidation may mean processing so that data recorded in the region to be deleted is recognized as invalid data on the host 1100 side. This invalidation can be done by managing a mapping table in which the mapping between the physical and logical blocks of the main storage 1240 is recorded. For example, invalidation may be accomplished by mapping out mapping information for the region to be deleted from the mapping table or by marking the region to be deleted in the mapping table.

As described in detail below, the invalidation described above may optionally involve an operation in which information indicating an invalid block is recorded / stored / marked in the memory block (s) of the main storage unit 1240 corresponding to the region to be deleted. . Alternatively, the invalidation described above may involve an operation in which information indicating an invalid block is recorded / stored / marked in the memory block (s) of the main storage unit 1240 corresponding to the region to be deleted. Address management during write or erase operations for the flash memory chips constituting the main storage unit 1240 is performed by a flash translation layer (FTL), which is system software or firmware executed by the SSD controller 1220. Can be managed.

In operation S1300, it may be determined whether the operation of step 1 has ended. If the determination in step S1300 does not end the step 1 operation, the procedure returns to step S1100 and the step 1 operation is performed again. Here, the step 1 operation may be executed a plurality of times without being limited by the number of executions. In operation S1300, if the operation of step 1 is completed, it is determined whether sudden power-off occurs in operation S1400.

If sudden power-off has occurred as a result of the determination in step S1400, the previously stored pivot data is retrieved in step S1500. The pivot data is stored in the main storage unit 1240 during the step 2 period performed before the sudden power off. On the other hand, if sudden power off has not occurred as a result of the determination in step S1400, the procedure goes to step S1600.

Subsequently, a step 2 operation is started in step S1600. In an exemplary embodiment, step S1600 may be performed at an idle time when a command is not input from the host 1100.

In operation S1700, invalid data included in an area marked as invalid in operation S1200 is purged from the buffer memory or the flash memory. The delete / purge operation performed in operation S1700 may be performed to merge / copy valid data included in the region to be invalidated into a new memory block and to prevent data (ie, invalid data) of the region to be invalidated from being accessed from the outside. And completely removing data of an area to be invalidated from the main storage unit 1240. The operation of completely removing invalid data from the main storage unit 1240 may mean a secure erase. In a secure erase operation, the invalidated data may be completely erased so that recovery is impossible to prevent from being read from the outside. Such a secure erase operation may be usefully used for invalidation of data requiring security.

If there has been a previous power off, the erase / purge operation performed in step S1700 may be executed from the operation actually stopped using the pivot data retrieved in step S1500. Therefore, in performing the time-consuming erase / purge operation, the steps performed before the sudden power off are not duplicated. Therefore, the time taken for the delete / purge operation can be reduced, and the execution efficiency of the invalidation operation is increased.

If the previous power off has not occurred, the erase / purge operation performed in step S1700 will be sequentially performed based on the result of recording / saving / marking in step S1200. In this case, step S1500 of searching previously stored pivot data may be omitted.

The merge / copy operation that may be performed in the erase / purge operation of step S1700 may be performed in various ways. For example, the merge / copy operation may be performed using file system information (for example, FAT, HPFS, NTFS, UFS, FFS, MMFS, VFAT, etc.). In this case, valid data read for the merge / copy operation may be merged / copied in a latched state inside the flash memory without being output to the flash memory. This merge / copy operation may be performed through a data writing method of a flash memory known as a copyback operation. In addition, the merge / copy operation may selectively exclude merge / copy operations on invalid data included in an area marked as an area to be invalidated. As a result, the invalid data becomes as if copied to a value of 0xFF, and the time required for the merge / copy operation is reduced due to the selective merge / copy operation. However, this method is an example of performing the merge / copy operation of the present invention, and the data writing method that may be used in the merge / copy operation of the present invention may be variously changed and modified.

Subsequently, in step S1800, pivot data for each stage about the current delete / purge operation may be stored. Pivot data may be stored in the main storage unit 1240 configured of a nonvolatile memory such as a flash memory. Thus, the pivot data can be preserved even if the power supply is interrupted. If an error such as sudden power off occurs when the step 2 section of the invalidation operation is not finished, the pivot data will be provided to the controller 1200 during the boot operation.

On the other hand, the timing or manner in which the pivot data is stored may be configured in various forms. For example, depending on the number of times the step 2 operation is executed, the time point at which the pivot data is stored may be determined according to the capacity / size of the data for performing / executing the step 2 operation. In an exemplary embodiment, the pivot data may be configured to be stored each time the Step 2 operation is executed, or may be configured to store the pivot data when the Step 2 operation has been performed a predetermined number of times. Similarly, the pivot data may be configured to be stored whenever the capacity of the data to be processed in the step 2 operation is equal to or larger than a predetermined size.

After the pivot data is stored, in step S1900, it may be determined whether the step 2 operation is completed. As a result of the determination in step S1900, if the step 2 operation is not finished, the procedure returns to step S1700. If the step 2 operation is completed as a result of the determination in step S1900, the invalidation operation ends.

6 and 7 are diagrams for describing the purge operation S1700 illustrated in FIG. 5.

Hereinafter, an example of performing a purge operation using file system information (eg, FAT, HPFS, NTFS, UFS, FFS, MMFS, VFAT, etc.) will be described with reference to FIGS. 6 and 7. 7 illustrates a merge / copy operation when data invalidated by the host 1100 is stored in a flash memory. FIG. 7 illustrates a configuration in which valid data included in an area marked as invalid is selectively copied to an empty memory block (ie, an erased memory block) of a flash memory during a merge / copy operation.

As will be described in detail below, the purge operation merges and copies valid data included in an area to be invalidated into a new memory block, and stores data (ie, invalid data) of the area to be invalidated from the main storage unit 1240. It may include an operation of removing. In this case, since invalid data does not exist logically or physically in the main storage unit 1240, access from the outside may be prevented at the source.

First, referring to FIG. 6, the main storage unit 1240 may include a FAT area 1241, a data area 1242, a log area 1243, and a meta area 1244. Here, the FAT area 1241 is a kind of file system area.

The FAT file system includes a master boot record (MBR), partition boot record (PBR), first and second file allocation tables (primary FAT, copy FAT), and root directory. can do. The file to be stored / stored in the storage device 1200 may be identified using two pieces of information. The information is the file name of the file and the path of the directory entry that goes through to reach where the file is stored.

Each entry in a directory is, for example, 32-bytes long and stores information such as file name, extension, file attribute byte, date / time last changed, file size, and connection to the startup cluster. . In particular, in the file name, one special code may be used as the first character of the file name to indicate the deleted file. For example, you can place the hexadecimal byte code E5h in place of the first letter of the file name. This is a special tag that can tell the system that this file has been deleted. When any file is deleted, the central processing unit of the host 1100 places one special code in the first character of the file name of the deleted file, as well as an invalidation command (or deleted file data corresponding to the deleted file). Information about invalid information) may be output to the storage device 1200.

File system information (eg, FAT information) may be stored in the FAT area 1241. Log blocks of the log area 1243 may be respectively assigned to some of the memory blocks of the data area 1242. That is, when data is to be written to the memory block of the data area 1242, the data may be stored in the corresponding log block of the log area 1243 instead of being directly written to the memory block.

However, if a log block of the log area 1243 corresponding to the memory block of the data area 1242 is not specified, or there is no empty page in the log block of the log area 1243, or the host ( If there is a request from 1100, the merge operation may be performed. The merge operation allows valid pages of the log block and valid pages of the memory block to be copied to a new memory block or log block.

In addition, as shown in FIG. 5, when an invalidation command input from the host 1100 is input, an area to be invalidated may be marked in step 1. In the step 2 section, a merge operation on an area marked as an area to be invalidated and an operation of substantially removing invalid data may be performed. When the merge operation is performed, the mapping information is changed, and the changed mapping information may be stored in the meta area 1244. At the same time, the progress information for each step for the ongoing delete / purge operation may be stored as pivot data. Pivot data may be stored in the main storage unit 1240 configured of a nonvolatile memory such as a flash memory. In example embodiments, the pivot data may be stored in the meta area 1244 of the main storage unit 1240.

Since the pivot data is stored in the main storage unit 1240 composed of a nonvolatile memory such as a flash memory, the contents of the pivot data may be preserved even when the power supply is interrupted. If an error such as sudden power-off occurs when the step 2 section of the invalidation operation is not finished, the pivot data may be provided to the controller 1200 during the boot operation.

In the FTL mounted on the controller 1200, the pivot data provided from the main storage unit 1240 is retrieved during the booting operation, and any operation is performed during the invalidation operation (for example, in the step 2 section, which takes a long time during the invalidation operation). It can be determined whether or not until. Then, the invalidation operation is performed again from the interrupted operation.

The controller 1220 may be configured to control the main storage unit 1240 when there is an access request from the host 1100. As shown in FIG. 6, the controller 1220 may include a processing unit 1226 and a work memory 1227. The work memory 1227 is also called an internal memory. The FTL may be stored in the work memory 1227.

7, the storage device 1200 of the present invention can perform a purge operation using the result marked in step 1. The purge operation according to the present invention merges and copies valid data included in a region to be invalidated into a new memory block, and removes data (ie, invalid data) of the region to be invalidated from the main storage 1240. May include an action. When performing the purge operation, the storage device 1200 of the present invention may periodically store pivot data for each step of the erase / purge operation currently in progress. Pivot data may be stored in the main storage unit 1240 configured of a nonvolatile memory such as a flash memory.

As shown in FIG. 7, the valid pages 2511 and 2513 of the log block 2510 and the valid pages 2522 of the memory block 2520 are merged and copied through the merge / copy operation during the step 2 period. Can be merged / copied into

If sudden power off occurs in the storage device 1200 during the invalidation operation, sudden power off may occur in the merge / copy operation on the valid data and the deletion operation on the invalid data performed during the step 2 period. It can be controlled according to the analysis result of previously stored pivot data. For example, if a reboot is performed after a sudden power off, the storage device 1200 may analyze a progress state of a purge operation by searching previously stored pivot data. Based on the analysis result of the pivot data, the invalidation operation can be executed from the time when the power supply is stopped by sudden power-off. Therefore, the entire invalidation operation that was executed before the sudden power off need not be executed again, and the execution time of the invalidation operation is reduced.

For example, if, as a result of analyzing the pivot data, a sudden power off occurs while performing merge / copy of the first and third pages 2511 and 2513 of the log block 2510, the storage autonomous 1200 is reset. After booting, the invalidation operation may be executed from the merge / copy operation on the first and third pages 2511 and 2513. In this case, the first and third pages 2511 and 2513 of the log block 2510 marked as valid may be copied to the first and third pages 2531 and 2533 of the new memory block 2530, respectively. . The second page 2522 of the memory block 2520 marked as valid may be copied to the second page 2532 of the new memory block 2530.

In this case, the fourth page 2524 of the memory block 2520 marked as invalid may be configured not to be copied to the second page 2532 of the new memory block 2530. That is, the present invention selectively copies valid pages of the memory block 2520 to a new memory block 2530 with reference to a result (eg, FAT information 2540) marked in the step 1 section during the step 2 section. can do. In this case, the memory block 2530 that is the target of the merge operation (that is, a valid page is copied) may be configured as an empty memory block in an erased state.

In detail, the FAT information 2540 stored in the main storage unit 1240 may store information on whether the page of the memory block 2520 is allocated. For example, since the first page 2521 of the memory block 2520 is a page not used to store a file, the first page 2521 may be displayed as NA (Not Allocation) in the FAT information 2540. Since the second page 2522 of the memory block 2520 is a valid page for storing a file, the second page 2522 may be marked as A indicating allocation to the FAT information 2540. The FAT information 2540 corresponding to the third and fifth pages 2523 and 2525 of the memory block 2520 is indicated by NA.

When the fourth page 2524 of the memory block 2520 is designated as an area to be invalidated according to an invalidation command input from the host 1100, the FAT information 2540 corresponding to the fourth page 2524 of the memory block 2520 is designated. ) Is denoted by D, which indicates Delete. Since a file is stored in the fourth page 2524 of the memory block 2520, the data block 2520 is displayed as a valid page, but from a file system point of view, it corresponds to a deleted page. In other words, the data stored in the fourth page 2524 of the memory block 2520 is a valid page from an FTL perspective, but is no longer valid from a file system perspective. In this case, data stored in the fourth page 2524 of the memory block 2520 is considered invalid on the host 1100 side.

Therefore, when the fourth page 2524 of the memory block 2520 represents the erased data (ie, invalid data) from the file system perspective, the data of the corresponding page is not included in the merge / copy operation according to the first embodiment of the present invention. It is not copied to the new memory block 2530. As a result, the invalid data in the merge / copy operation according to the present invention is as if copied to a value of 0xFF.

If the invalid data is processed with don't care regardless of what value appears on the host 1100 side, the data marked as invalid in the buffer memory 1228 is selectively controlled under the control of the controller 1220. May be moved to primary storage 1240. In this case, even though data marked as invalid data is copied to the main storage unit 1240, a file related to invalid data which is data stored in the main storage unit 1240 may not be affected.

In another embodiment, in the merge / copy operation according to the first and second embodiments of the present invention, when invalid data is treated as don't care regardless of what value appears on the host 1100 side, merge / copy The memory block to be copied may be configured such that only an area corresponding to valid data maintains an empty space. That is, in the merge / copy operation of the present invention, the merge / copy operation on the invalid data may not be executed. Therefore, only the region corresponding to the valid data may be in the erased state even if the entire memory block is not erased. According to such a configuration, more flexible data allocation is possible, and the data bin storage capacity of the main storage unit 1240 can be fully utilized. In such a case, in the merge / copy operation of the present invention, an operation of erasing a block to be the object of the merge / copy operation may be selectively omitted.

8 is a diagram illustrating a configuration of a memory system 2000 according to an embodiment of the present invention.

Referring to FIG. 8, the memory system 2000 according to the present invention may include a controller 1220 and a main storage 1240.

The main storage unit 1240 illustrated in FIG. 8 may be configured in the same manner as the main storage unit illustrated in FIGS. 1, 4, and 6. In an exemplary embodiment, the main storage unit 1240 may be configured as a flash memory among nonvolatile memories. In addition, in the memory system 2000 illustrated in FIG. 8, the invalidation scheme of the present invention using the pivot data described above may be applied.

The controller 1220 may be configured to control the main storage unit 1240. The controller 1220 may be configured in the same manner as the controller illustrated in FIGS. 1, 2, 3, 4, and 6. Therefore, duplicate description of the same configuration will be omitted below.

The memory system 2000 according to the present invention may configure a memory card or an SSD by combining the main storage unit 1240 and the controller 1220. In this case, the controller 1220 performs a function as a memory controller.

The SRAM 2110 may be used as a work memory of the processing unit 2120. The host interface 2130 may include a data exchange protocol of a host connected to the memory system 2000. The error correction circuit 2140 included in the controller 1220 may detect and correct an error included in read data read from the main storage unit 1240. The memory interface 2150 may interface with the main storage unit 1240 of the present invention. The processing unit 2120 may perform various control operations for exchanging data of the controller 1220. Although not shown in the drawings, the memory system 2000 according to the present invention may further be provided with a ROM (not shown) for storing code data for interfacing with a host.

The main storage unit 1240 may be provided in a multi-chip package composed of a plurality of flash memory chips. The memory system 2000 of the present invention may constitute a highly reliable storage medium having a low probability of error occurrence. In particular, it is possible to configure a memory system such as SSD which is being actively studied recently. In this case, the controller 1220 is configured to communicate with the outside (eg, host) via one of a variety of interface protocols such as USB, MMC, PCI-E, SAS, SATA, PATA, SCSI, ESDI, and IDE. Will be. In addition, the controller 1220 may further include a configuration for controlling an invalidation operation (particularly, a security invalidation operation) for invalidating data deleted from the host.

9 is a diagram illustrating a configuration of a computing system 3000 according to an embodiment of the present invention. 9 illustrates a configuration of a computing system 3000 including a storage device 1200 in accordance with the present invention.

Referring to FIG. 9, a computing system 3000 according to the present invention may include a microprocessor 1200, a RAM 3300, a user interface 3400, a baseband chipset electrically connected to a system bus 3600. Same modem 3500, and storage 1200.

The storage device 1200 may include a controller 1220 and a main storage unit 1240. The controller 1220 provides a physical connection between the CPU 1200 and the main storage unit 1240 through the system bus 3600. That is, the controller 1220 may provide interfacing with the main storage unit 1240 in correspondence with the bus format of the CPU 1200.

The main storage unit 1240 illustrated in FIG. 9 may be configured substantially the same as the main storage unit illustrated in FIGS. 1, 4, and 6. In addition, the main storage unit 1240 illustrated in FIG. 9 may store the pivot data described above, and the invalidation method of the present invention using the pivot data may be applied. Therefore, duplicate description of the same configuration will be omitted below.

When the computing system 3000 according to the present invention is a mobile device, a battery (not shown) for supplying an operating voltage of the computing system 3000 may be additionally provided. Although not shown in the drawings, the computing system 3000 according to the present invention may further be provided with an application chipset, a camera image processor (CIS), a mobile DRAM, or the like. The storage device 1200 may configure, for example, an SSD (Solid State Drive / Disk) that uses a nonvolatile memory to store data. For example, the storage device 1200 illustrated in FIG. 9 may constitute an SSD. In this case, the controller 1220 will operate as an SSD controller.

The nonvolatile memory device and / or the memory controller according to the present invention may be mounted using various types of packages. For example, the flash memory device and / or the memory controller according to the present invention may be a package on package (PoP), ball grid arrays (BGAs), chip scale packages (CSPs), plastic leaded chip carrier (PLCC), plastic dual in- Line Package (PDIP), Die in Waffle Pack, Die in Wafer Form, Chip On Board (COB), Ceramic Dual In-Line Package (CERDIP), Plastic Metric Quad Flat Pack (MQFP), Thin Quad Flatpack (TQFP), Small Outline (SOIC), Shrink Small Outline Package (SSOP), Thin Small Outline (TSOP), Thin Quad Flatpack (TQFP), System In Package (SIP), Multi Chip Package (MCP), Wafer-level Fabricated Package (WFP), It can be implemented using packages such as Wafer-Level Processed Stack Package (WSP).

As described above, the embodiments are disclosed in the drawings and the specification. Although specific terms have been used herein, they are used only for the purpose of describing the present invention and are not used to limit the scope of the present invention as defined in the meaning or claims. Therefore, those skilled in the art will understand that various modifications and equivalent other embodiments are possible. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

1000, 3000: computing system 1100: host
1200: storage device 1220: controller
1240 main storage unit 2000 memory system

Claims (10)

Marking an area to be invalidated in response to an invalidation command input from the host;
Purging the area to be invalidated based on the marking result;
Storing progress information of the purge as pivot data; And
And re-executing the purge based on the pivot data when a sudden power off has occurred. The method of claim 1,
Re-executing the purge is performed from the operation interrupted when the sudden power off occurs. The method of claim 1,
Storing the pivot data is performed each time the purge is executed. The method of claim 1,
The storing of the pivot data is performed when the purge is performed a predetermined number of times or more. The method of claim 1,
And the pivot data is stored in a nonvolatile memory. The method of claim 1,
The storing of the pivot data may include storing the progress state information of the fuzzy in a page or block unit. The method of claim 1,
And storing the pivot data is performed at a predetermined time interval. Marking an area to be invalidated in response to an invalidation command input from the host;
Determining whether a sudden power off has occurred;
Retrieving pivot data when the sudden power off has occurred;
Purging the area to be invalidated based on the marking result and the search result of the pivot data; And
And storing the progress information of the purge as current pivot data. Host; And
A storage device operable in response to control of the host;
The storage device includes a main storage and a controller configured to control the main storage,
The controller stores the progress state information of the invalidation command input from the host as pivot data in the main storage unit, and after a sudden power off has occurred, the controller stops the operation based on the pivot data based on the pivot data. Computing systems that redo invalidate operations. Main storage; And
It includes a controller for controlling the operation of the main storage,
The controller stores the progress state information of the invalidation command input from the host as pivot data in the main storage unit, and invalidates from an operation interrupted at the sudden power-off based on the pivot data after a sudden power off occurs. Storage device to redo the operation.

Images