CN1719422A - Virtual access method of storage document data - Google Patents
Virtual access method of storage document data Download PDFInfo
- Publication number
- CN1719422A CN1719422A CNA2005100908332A CN200510090833A CN1719422A CN 1719422 A CN1719422 A CN 1719422A CN A2005100908332 A CNA2005100908332 A CN A2005100908332A CN 200510090833 A CN200510090833 A CN 200510090833A CN 1719422 A CN1719422 A CN 1719422A
- Authority
- CN
- China
- Prior art keywords
- physical
- address
- data
- sector
- block
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000015654 memory Effects 0.000 claims abstract description 60
- 238000013519 translation Methods 0.000 claims abstract description 30
- 238000006243 chemical reaction Methods 0.000 claims description 22
- 238000004321 preservation Methods 0.000 claims description 15
- 238000012217 deletion Methods 0.000 claims description 12
- 230000037430 deletion Effects 0.000 claims description 12
- 230000003111 delayed effect Effects 0.000 claims description 9
- 125000004122 cyclic group Chemical group 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 230000001133 acceleration Effects 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 2
- 238000011017 operating method Methods 0.000 abstract 1
- 230000008569 process Effects 0.000 description 16
- 230000003139 buffering effect Effects 0.000 description 6
- 238000007726 management method Methods 0.000 description 5
- 230000003068 static effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000013523 data management Methods 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 238000000151 deposition Methods 0.000 description 2
- 239000012467 final product Substances 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 238000010977 unit operation Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000000547 structure data Methods 0.000 description 1
Images
Abstract
The present invention relates to a memory file data virtual access method. Said method is characterized by introducing address translation table, physical state table and physical distribution description table and utilizing a storage equipment capable of abstracting different types of storage equipments into a consolidated addressing and operating method so as to raise data reading-writing efficiency of file system and management facility of memory, and prolong service life of memory. Said invention also provides the concrete steps of said access method.
Description
Technical field
The invention belongs to computer memory document data access and art file management technology field, specifically a kind of virtual access method of storage document data.
Technical background
File system is a software in order to store and management data, and the summation of the file data structure of on storer, setting up, as soft or hard dish, CD, internal memory, flash memory (Flash) etc.In general, file system is made up of catalogue and file, generally include parameter block, file allocation table, root directory tabulation and data field, file system is mainly finished three functions: the space of being consumed on the track record storer and the physical storage locations of free space, maintenance directory name and filename and each file of track record.
In the prior art, computing machine by the file system in the storer when the data of read-write memory, usually directly send read write command to the hardware device stores device, data in the sector in the storer are directly read, write or wipe, there is big problem in the reading/writing method of this file system and file data: 1, because partial document system, own design reasons as file allocation table FAT file system, its guiding and parameter block (BPB), the storage order of FAT table and root directory must be continuous and fixing, if bad piece appears in wherein any one, to cause whole file system to use, though can recover file data in another storer, will reuse this storer and must format.2, partial document system; do not consider power down protection as the FAT file system; this is the reason that adopts traditional data structure owing to it; though can customize to support power down protection to it; but this may cause the incompatible problem with other system, also can increase the development difficulty and the stability risk of file system simultaneously.3, partial document system, minimum operation unit as the FAT file system is incompatible with flash memory (Flash), though the FAT file system can be supported maximum 64K bunch (cluster) in theory, but this uses on the equipment of last gigabit storage space usually, and Windows Windows can't discern and use the high capacity bunch low capacity memory device for the basis generally speaking.4, partial document system, do not consider to write equilibrium problem as the FAT file system, main cause is the bad block management security deficiency of file system, because some data forms of partial document system, need frequent updating as the FAT table, but because its memory location fixes, the possibility that occurs bad piece at the storage area of some type memories increases considerably, and then causes the file system fault.
Therefore prior art adopts direct access mode to the sector, just can't use when bad piece appears in the sector, if when bad piece sector appears at the critical data memory location, can cause the collapse of file system; Prior art only can be carried out data management to a kind of physical storage device, can not be with the physical storage device of number of different types as carrying out unified data management; Prior art is not consider its life-span to the visit of sector, therefore at the part memory device particularly on the flash memory, will cause its critical data memory sector owing to the serious situation that bad piece causes file system collapse obliterated data too much appears in number of operations, it is same because erasing times is more, the sector becomes bad piece, causes using the decline of capacity; Prior art reckons without the characteristic of special memory device, as flash memory, the single sector of this kind equipment only can write once, prior art writes when having write the sector once more handling, the a plurality of sectors content that needs to close on is together wiped, the a plurality of sector datas that needed to close on before wiping read in the buffering, to write data again writes with the data of a plurality of sectors of closing in buffering, cause unnecessary writing and erase operation, reduce the life-span of storer, increased the running time, for example, general FLASH is a 512-528 byte (Bytes) in the minimum operation unit that reads and writes data, but when writing data, there have been data if desire writing position, need behind this cell position data erase, can write, generally need wipe 16-32Kbytes, when the data of desiring to write one 512 byte during in the 3rd sector, there have been data in this sector, at this moment needs the content of this 1-32 sector is read in internal memory, and the content with the 3rd sector in internal memory replaces, and wipe all the elements of 1-32 sector, write back the content of FLASH 1-32 sector again.
The write operation that the existing file system realizes has following shortcoming: 1, because each write operation is all wanted read-write memory, and the erasable speed of flash memory is slow, therefore, causes the execution efficient of whole write operation low.2,, can shorten the serviceable life of storer owing to the frequency height of read-write memory.
Summary of the invention
In order to overcome the deficiencies in the prior art part, the object of the present invention is to provide a kind of virtual access method of storage document data, to prolong the serviceable life of storer, improve read-write efficiency.
For finishing the foregoing invention purpose, the overall technical architecture that the present invention takes is: a kind of virtual access method of storage document data may further comprise the steps:
Step 1, storer are obtained from read-write or the removal request of file system to the file data of certain sector;
Step 2, central processing unit (CPU) will read and write by the inquire address conversion table or the logical sector address of removal request is converted to physical sector address;
Step 3, find the data available block device of this physical sector by the physical state record sheet;
Step 4, from this physical sector data available block device, read file data, and return to file system, or file data is written in the data available block device of this physical sector, or file data is deleted from the data available block device of this physical sector.
Two identical core description blocks are arranged on the storer in the described step 1, file data is deposited in after by the cyclic check computing in two different masses in the storer,, be used for checking mutually and support power down protection and wipe acceleration with mutual checking.
Described verification computing specifically comprises:
When system initialization, read each core description block respectively and they are carried out check code calculating;
The check code that calculates when any core description block and the check code of core description block preservation not simultaneously, this core description block is invalid;
The check code that calculates when all core description blocks and the check code of core description block preservation not simultaneously, reformatting;
When the check code that calculates when any core description block was identical with the check code of core description block preservation, this core description block was effective;
When the check code that calculates when all core description blocks is identical with the check code of core description block preservation, reference value to two core description blocks compares, when a core description block reference value during greater than another core description block reference value, get big core description block for effectively, use the data of effective core description block as each current table.
File system in the described step 1 includes address translation table and physical distribution description list, and these two tables of data are stored in the identical memory sectors equipment.
Described address translation table is used for the conversion of logical and physical block address, and every call number of table is represented logical address, and its value that comprises is a physical address.
Described step 3 also comprises the step that obtains the descriptor of this physical sector by the physical distribution description list.
Described physical distribution description list is the supplementary table of file system physical state record sheet, is used for the data access address of multiple different stand-alone memory device is coupled to the linear access address.
Described step 3 specifically comprises the steps:
Step 311, obtain first available block by the physical state record sheet, and by block delete/the write access times that counters table obtains this piece;
Step 312, obtain a plurality of available block and its access times, and compare by deleting/write counters table:
Step 313, by after all available block are traveled through, obtain the available block that access times are minimum.
First available block of acquisition in the described step 311 specifically may further comprise the steps:
Step 3111, the piece conversion table that produces bad piece is labeled as produces bad piece;
Step 3112, original value in the physical state record sheet is composed to the temporary variable in the storer;
Step 3113, new available block of acquisition;
Step 3114, last logical address of physical state record sheet is distributed to bad physical block.
This physical data block device that file data is written in the described step 4 further may further comprise the steps:
The state of step 41, judgement physical sector, if the state of this physical sector is not for using or having deleted, then obtain the device descriptor of this physical sector by the physical distribution description list, to determine the equipment and the address of this physical sector, further judge whether this sector needs to carry out erase operation, if the state of this sector is for deleting, storer need be wiped legacy data and just can write fashionablely simultaneously, then wipes this sectors of data content;
If the state of this sector of step 42 can't use for using maybe, then the inquire address conversion table obtains physical address one by one, obtain the state of physical sector address again by the physical state record sheet, until finding one not use or deleted physical sector, and be exchanged for physical sector address with the address translation table at the logic sector place of file system requests;
At first write down the logical sector address of original file system request before step 43, the inquiry, when inquiring not use or deleted physical sector address, preserve this physical sector address and be positioned at the logical sector address of address translation table, this physical sector address is saved in the address translation table of logical sector address of file system requests, the physical sector address in this table logical sector address is kept in the address translation table of sector logical address of the new physics sector of finding;
Step 44, data are write in the sector of this physical equipment.
Described step 44 further may further comprise the steps:
Step 441, in the physical state record sheet, seek and can use piece, can use piece if having, the new data that then writes direct, and this piece is labeled as uses;
If when step 442 can't search out future use block in the physical state record sheet, then seek piece, this delayed deletion piece is carried out erase operation, and write data the delayed deletion in the physical state record sheet.
File data is deleted from the equipment of this physical sector in the described step 4 is meant that specifically with the status indication of this device data block in the physical state record sheet be delayed deletion, do not carry out the actual erase operation.
When being in idle condition in system, the deletion in the described step 4 carries out.
Described step 4 also comprises the step of revising data, specifically is meant in the physical state record sheet to seek untapped, and the new data of revising is write in the future use block, and physical state record sheet mark original block is wiped for postponing, and write-in block is for using.
Method of the present invention is by call address conversion table, physical state table and distribution description list, and the memory device of number of different types is used, and reached high read-write efficiency and serviceable life.When an equipment has flash memory and internal memory, can deposit data in flash memory in advance, when file system is made amendment to the data that are positioned at flash memory, for the stability of system, can directly not revise flash memory, and the Data Position at flash memory place be moved on to the internal storage data position by address translation table, new data more, can improve the data read-write efficiency of file system by this method, prolong the serviceable life of storer, therefore have significant advantage and useful effect.
Description of drawings
Fig. 1 is file system structure figure of the present invention;
Fig. 2 is graph of a relation between each data structure of the present invention;
Fig. 3 is a virtual data access synoptic diagram of the present invention;
Fig. 4 sets up the process synoptic diagram for data structure of the present invention;
Fig. 5 is a main flow chart;
Fig. 6 is for to write data flowchart to file system;
Fig. 7 is the deleted data process flow diagram.
Embodiment
Below in conjunction with Figure of description the specific embodiment of the present invention is described.
As shown in Figure 1, file system structure figure of the present invention, file system includes parameter block, file allocation table, root directory tabulation and data field, it also includes two core data tables: address translation table and physical address assignments table, these two tables of data are stored in the memory device, and file system storage data are identical, and storer comprises soft or hard dish, CD, internal memory, flash memory (Flash) etc., the present invention is specially adapted to the storer of big data quantity access, as flash memory (Flash).
Table 1 physical distribution description list
| The physical equipment identifier | Reference position in the physical state table | End position in the physical state table | The reference position of physical equipment |
| Nonvolatile memory (disk) | 0 | 99 | 0 (piece) |
| Volatile memory (internal memory) | 100 | 200 | 0x0 (address) |
Table 1 is the physical distribution description list, and it is the supplementary table of physical state record sheet, and this table is that piece/bus address is coupled to the linear access address with the data access address of multiple different stand-alone memory device.Central processing unit (CPU) obtains the step of the descriptor of this physical sector by the physical distribution description list.Descriptor can make the indications of physical sector.
Shown in Figure 2 is relation between each data structure of the present invention, the state of each data segment of in store two the dissimilar equipment of physical state record sheet, state has 4 kinds, be respectively do not use (deleting), with, wait to delete and can't use (bad piece), address translation table is used for the conversion of each LBA (Logical Block Addressing) to physical block address.
In table 1 and Fig. 2, the physical state record sheet is described the physical state of multiple distinct device according to linear address by the physical distribution description list.The instance data of table 1 shows, the 0th to the 99th of physical state description list is the 0th to the 99th of nonvolatile memory, and the 100th to the 200th of physical state description list is the state in 0x0 to the 0xC800 address realm of volatile memory.
Address translation table is different with the conversion purpose of physical distribution description list, address translation table provides a physical address for each independent logical address, each contiguous LBA (Logical Block Addressing) can be discontinuous physical address, LBA (Logical Block Addressing) also is virtual address, it is a kind of address of being used by software, by address translation table logical address is converted to physical address, it can be with each data that are distributed in the physical address space, connect into continuous logical address space, and the physical distribution description list is based on the sector equipment in the storer, for one section continuous piece or address provide mapping.
As shown in Figure 3, reading/writing method of the present invention is the virtual memory method, file system is when read-write memory, no longer direct to memory hardware equipment transmission read write command as method of the prior art, but at first read write command is passed through address translation, logical address is converted to the physical address of memory hardware equipment.Method of the present invention is for file system provides a kind of abstract method, the purpose that file system is dealt with problems as reaching of operational store hardware device.
See also Fig. 4, at first set up data structure, data structure of the present invention is set up process as shown in Figure 4.At first, be each equipment disposition physical distribution description list, this table is disposed by the developer voluntarily according to the particular device situation, and the purpose that disposes this table is the equipment that the equipment of number of different types logically is merged into a unified type method and unified linear access address; Then, the capacity summation according to each equipment obtains maximum memory capacity; Come the required space of distribute data structure according to the memory capacity of maximum, and the data of initialization data structure; Afterwards, check the state of storer, and memory state is saved in corresponding data item in the physical state record sheet; At last, according to the every value in the natural order distribution address translation table, as the 0th sector of first logic of address translation table is the 0th sector in the physical state allocation table, just certain sector of certain equipment of in the physical distribution description list the 0th sector of physical state allocation table being described.
For example storer is one new does not have a formative flash memory (Flash), must carry out initialization before using, and its process is as follows:
At first, set up data block state transition table (BST) buffer zone of physical state record sheet just, and buffer content is put sky; Check bad piece then, and its state is set according to the numbering of bad piece.
Afterwards, set up the buffer zone that temporarily is used for data BST, and buffer contents is put sky; Carry out block sort, available block is placed the forward part of piece conversion table.
Bad piece is placed the rear section of table; BST is obtained the process check code by cyclic check (CRC) computing, BST is deposited in the first available block;
BST is obtained the process check code by the CRC computing, the copy of BST is deposited in second available block; Close the buffering of BST, set up and to be used for block delete/write counters table (BCT, BlocksErased/Wear Counter Table) buffering temporarily, and buffer content is put sky;
Next, counters table BCT is deleted/write to INIT block: with BCT by with last time the process check code carry out the CRC computing and obtain this process check code, BCT is deposited in the first available block; With BCT by with last time the process check code carry out the CRC computing and obtain this process check code, BCT is deposited in second available block, first available block size is generally the minimum read-write unit of nonvolatile memory in the system, as 2k, because BST, BCT data structure data volume is little, so can deposit same in.
At last, finish initialization.
After data structure is set up, just can use these data BST to carry out the read-write of file data and deletion action.
See also Fig. 5 main flow chart of the present invention, the present invention includes following steps:
Step 1, storer are obtained from read-write or the removal request of file system to the file data of certain sector;
Step 2, central processing unit (CPU) will read and write by the inquire address conversion table or the logical sector address of removal request is converted to physical sector address;
Step 3, find the minimum data available block device of access times of this physical sector by the physical state record sheet;
Step 4, from this physical sector data available block device, read file data, and return to file system, or file data is written in the data available block device of this physical sector, or file data is deleted from the data available block device of this physical sector.
As shown in Figure 6, it is as follows to write the step of data to file system:
Obtain from the write request of data of file system, obtain the physical address of this sector, and obtain the state of this sector, and judge the state of this sector by the physical state record sheet by the inquire address conversion table to certain sector;
If the state of this sector is not for using or having deleted, obtain the device descriptor of this sector by the physical distribution description list, to determine the physical equipment and the address of this sector, further judge whether this sector needs to carry out erase operation, because some storer need carry out erase operation to former data before writing new data, if the state of this sector is for deleting, storer need be wiped legacy data and just can write fashionablely simultaneously, wipes this sectors of data content earlier;
If the state of this sector can't use for using maybe, the inquire address conversion table obtains physical address one by one, obtains the state of physical sector address again by the physical state record sheet, up to finding one not use or deleted physical sector;
By with the address translation table exchange physical sevtor address at the logic sector place of file system requests, purpose so that the logic sector that is reached for file system requests distributes new write sector at first writes down the logical sector address A of original file system request before the inquiry; When inquiring not use or deleted physical sector address, preserve the logical sector address B that this physical sector address is positioned at address translation table; This physical sector address is saved in the address translation table of logical sector address A of file system request, the physical sector address among this table logical sector address A is kept in the address translation table of the sector logical address B that finds the new physics sector;
At last, data are write in the sector of this physical equipment.
Also comprised carrying out before the exchange of logical address and physical address and write equilibrium step, it comprises dynamically writes balance and static state is write balance, dynamically writes balance and will provide the optimal candidate available block for write operation, and implementation method is as follows:
By data block state transition table (BST) just the physical state record sheet obtain first available block, and by block delete/the write access times that counters table (BCT) obtains this piece;
Obtain n available block and its access times by BST/BCT, and compare;
After all available block are traveled through, obtain the available block that access times are minimum.
It is a technology that is of little use that static state is write balance, and dynamically writes the equilibrium phase ratio, and its running time is longer, but can put in order using data, dynamically writes balance then not to using data to put in order.
Because some data are not modified for a long time, the life-span of its place piece is higher, and dynamically writes balance and can't operate on it; And other data are often revised as core description block (CDB).Write the physical storage block that balance can be exchanged the data of the data of minimum erasable number of times and maximum erasable number of times (SWAP) place by static state.
Static state is write balance and is divided into two kinds of patterns again: write balance comprehensively and generally write balance, the former is the data of minimum erasable number of times and the data of maximum erasable number of times, the data of exchanging the physical storage block at place, and modified address conversion table; Available block and existing piece all can.Generally write balance, i.e. the physical storage block of exchanging the place with the available block data of blocks of data and maximum erasable number of times of minimum erasable number of times, and modified address conversion table.
When producing bad piece in the process of using and operating, in BST, be labeled as 11b and get final product, be i.e. bad piece of BST[physics number]=11b.
In the present invention, the operation of any read-write flash memory (Flash) all is based on the address translation table of a logic to physics.When initialization procedure, available block and bad piece have all been made continuous logical place and have been cut apart, and promptly available block is in the forward part of address translation table, bad piece and then available block in the rear section of conversion table.
Therefore but in use bad piece can constantly occur, need will the follow-up bad piece that in available block, produces to move to the rear section of table, make the conversion table forward part to continue maintenance based on available block by address translation table.Its conversion method is: the physical state record sheet that produces bad piece is labeled as produces bad piece; Give a temporary variable with original value in the data physical state record sheet; Obtain a new available block; Last logical address of data physical state record sheet is assigned as bad physical block
The present invention also supports power down protection and wipes acceleration.The power down protection technology is meant that the unexpected power down of system/after resetting, memory data still can return to the state that last success is operated.This state is a complete write state, if power down/reset in the process of flash memory programming, all data of this process will be lost, and can't recover.In order to finish this function, need on storer such as flash memory (Flash), to have set up two parts of identical core description blocks (CDB).
The operating process of read/write and modification/wipe is as follows:
When carrying out write operation, at first in BST, seek and can use piece, the new data that writes direct, and this piece BST is labeled as uses;
When carrying out erase operation, be that delayed deletion (Pending Delete) gets final product with this piece at the status indication of BST, do not carry out the actual erase operation; When carrying out write operation, but can't search out the future use block piece in BST the time, this explanation or space are not enough, major part is taken by bad piece and useful data, can't continue operation, report information is given the user, or is that rubbish does not empty, not actual erase of delayed deletion is arranged among the BST, then at first the piece of the delayed deletion among the BST is carried out erase operation, then this piece is distributed to write operation, any BCT that needs this piece that wipes adds 1, before this process of execution, can at first carry out state and preserve;
When carrying out retouching operation, in BST, to seek untapped, and the new data of revising is write in the future use block, BST mark original block is wiped for postponing, and write-in block is for using.
The state preservation mainly is the preservation to BST and reference value, and actual is with the whole preservation of core description block (CDB).All operations before state is not preserved, in case power down/reset and will not play any effect.
State is preserved operation and will be read former CDB content, and the BST in the current buffering is replaced former BST, carries out again depositing in after cyclic check (CRC) computing in two the different pieces in the flash memory, comprises copy.
On storer, set up two parts of identical core description blocks (CDB) and can play the effect of mutual checking, when depositing any one CBD in, cut off the power supply, can not influence the data mode of last time to deposit in.
Each executable operations; management system can't be judged its unit operation such as close file, deleted file etc.; therefore need provide an interface to file system; when a unit operation is finished; just notifying management system is used for the state preservation of power down protection; perhaps when the buffering row of formation/internal memory is full, automatically perform.
In common file system, the key data structure that similar CDB is so all is kept at the fixed position, as first sector.
But because CDB itself operates the most frequent one in all pieces of flash memory, so the position of the each storage of CDB also draws according to writing balance and wiping speed technology, the new piece that writes after the each modification of CDB all will be the piece of erasable least number of times in all future use blocks.
Being in formative storer such as the flash memory (Flash), contain the free area, be used to preserve ECC data check sign indicating number and other customizable data.At this, the present invention is defined as signed data with all other customizable data, and the signature of CDB data will be different with the signature of all other data.Therefore when initialization system, all pieces of storer are traveled through/enumeration operation, read the also signature of decision block,, be the CDB data when being consistent with the signature that is predefined for CDB.Therefore can dynamically obtain the CDB data structure.
Judge the validity of CDB below.The present invention has two CDB to be positioned at the diverse location of storer, and each CDB has its check code and reference value to self.When system initialization, read each CDB respectively and they are carried out check code calculating.
The check code that calculates as any CDB and the check code of CDB preservation not simultaneously, this CDB is invalid.
The check code that the check code that calculates as all CDB and CDB preserve shows the total system damage, necessary reformatting not simultaneously.
When the check code that calculates as any CDB was identical with the check code of CDB preservation, this CDB was effective.
When the check code that calculates as all CDB is identical with the check code of CDB preservation, the reference value of two CDB is compared, to get big CDB during greater than another CDB reference value effective when a CDB reference value.Use effective CDB to comprise physical distribution description list, BTT, BST, the data of BCT etc. as each current table.
See also shown in Figure 7ly, the process of deleted data is as follows:
At first, obtain from the clear data request of file system certain sector;
Then, obtain the physical address of this sector by the inquire address conversion table, the state in the inquiry physical state record sheet of this physical address of being provided with is for deleting, and finishes.
Obliterated data also can be chosen another approach, can make file system need not to wipe immediately when writing data, the new data of can writing direct, and the work of wiping can be carried out in system's idle condition, writes efficient with raising.
The present invention can be based on any file system, and the visit of the memory devices under suitable bus address such as internal memory and block address such as input and output (I/O) pattern is as flash memory and disk.
The present invention reads, writes or delete by call address conversion table, physical state table and the distribution description list storage device data to number of different types, has higher read-write efficiency and serviceable life.When an equipment has flash memory and internal memory, can deposit data in flash memory in advance, when file system is made amendment to the data that are positioned at flash memory, stability for system, can directly not revise flash memory, by address translation table the Data Position at flash memory place is moved internal memory place Data Position, both more new data.
The above is the preferred embodiments of the present invention only, is not limited to the present invention, and for a person skilled in the art, the present invention can have various changes and variation.Within the spirit and principles in the present invention all, any modification of being done, be equal to replacement, improvement etc., all should be included within the claim scope of the present invention.
Claims (14)
1, a kind of virtual access method of storage document data is characterized in that, this method may further comprise the steps:
Step 1, storer are obtained from read-write or the removal request of file system to the file data of certain sector;
Step 2, central processing unit (CPU) will read and write by the inquire address conversion table or the logical sector address of removal request is converted to physical sector address;
Step 3, find the data available block device of this physical sector by the physical state record sheet;
Step 4, from this physical sector data available block device, read file data, and return to file system, or file data is written in the data available block device of this physical sector, or file data is deleted from the data available block device of this physical sector.
2, virtual access method of storage document data according to claim 1; it is characterized in that; two identical core description blocks are arranged on the storer in the described step 1; file data is deposited in after by the cyclic check computing in two different masses in the storer; with mutual checking, be used for checking mutually and support power down protection and wipe acceleration.
3, virtual access method of storage document data according to claim 2 is characterized in that, described verification computing specifically comprises:
When system initialization, read each core description block respectively and they are carried out check code calculating;
The check code that calculates when any core description block and the check code of core description block preservation not simultaneously, this core description block is invalid;
The check code that calculates when all core description blocks and the check code of core description block preservation not simultaneously, reformatting;
When the check code that calculates when any core description block was identical with the check code of core description block preservation, this core description block was effective;
When the check code that calculates when all core description blocks is identical with the check code of core description block preservation, reference value to two core description blocks compares, when a core description block reference value during greater than another core description block reference value, get big core description block for effectively, use the data of effective core description block as each current table.
4, virtual access method of storage document data according to claim 1 is characterized in that, the file system in the described step 1 includes address translation table and physical distribution description list, and these two tables of data are stored in the identical memory sectors equipment.
5, virtual access method of storage document data according to claim 4 is characterized in that, described address translation table is used for the conversion of logical and physical block address, and every call number of table is represented logical address, and its value that comprises is a physical address.
6, virtual access method of storage document data according to claim 1 is characterized in that, described step 3 also comprises the step that obtains the descriptor of this physical sector by the physical distribution description list.
7, virtual access method of storage document data according to claim 6, it is characterized in that, described physical distribution description list is the supplementary table of file system physical state record sheet, is used for the data access address of multiple different stand-alone memory device is coupled to the linear access address.
8, virtual access method of storage document data according to claim 1 is characterized in that, described step 3 specifically comprises the steps:
Step 311, obtain first available block by the physical state record sheet, and by block delete/the write access times that counters table obtains this piece;
Step 312, obtain a plurality of available block and its access times, and compare by deleting/write counters table;
Step 313, by after all available block are traveled through, obtain the available block that access times are minimum.
9, virtual access method of storage document data according to claim 1 is characterized in that, first available block of the acquisition in the described step 311 specifically may further comprise the steps:
Step 3111, the piece conversion table that produces bad piece is labeled as produces bad piece;
Step 3112, original value in the physical state record sheet is composed to the temporary variable in the storer;
Step 3113, new available block of acquisition;
Step 3114, last logical address of physical state record sheet is distributed to bad physical block.
10, virtual access method of storage document data according to claim 1 is characterized in that, this physical data block device that file data is written in the described step 4 further may further comprise the steps:
The state of step 41, judgement physical sector, if the state of this physical sector is not for using or having deleted, then obtain the device descriptor of this physical sector by the physical distribution description list, to determine the equipment and the address of this physical sector, further judge whether this sector needs to carry out erase operation, if the state of this sector is for deleting, storer need be wiped legacy data and just can write fashionablely simultaneously, then wipes this sectors of data content;
If the state of this sector of step 42 can't use for using maybe, then the inquire address conversion table obtains physical address one by one, obtain the state of physical sector address again by the physical state record sheet, until finding one not use or deleted physical sector, and be exchanged for physical sector address with the address translation table at the logic sector place of file system requests;
At first write down the logical sector address of original file system request before step 43, the inquiry, when inquiring not use or deleted physical sector address, preserve this physical sector address and be positioned at the logical sector address of address translation table, this physical sector address is saved in the address translation table of logical sector address of file system requests, the physical sector address in this table logical sector address is kept in the address translation table of sector logical address of the new physics sector of finding;
Step 44, data are write in the sector of this physical equipment.
11, virtual access method of storage document data according to claim 10 is characterized in that, described step 44 further may further comprise the steps:
Step 441, in the physical state record sheet, seek and can use piece, can use piece if having, the new data that then writes direct, and this piece is labeled as uses;
If when step 442 can't search out future use block in the physical state record sheet, then seek piece, this delayed deletion piece is carried out erase operation, and write data the delayed deletion in the physical state record sheet.
12, virtual access method of storage document data according to claim 1, it is characterized in that, file data is deleted from the equipment of this physical sector in the described step 4 is meant that specifically with the status indication of this device data block in the physical state record sheet be delayed deletion, do not carry out the actual erase operation.
13, virtual access method of storage document data according to claim 1 is characterized in that, carries out when the deletion in the described step 4 is in idle condition in system.
14, virtual access method of storage document data according to claim 1, it is characterized in that, described step 4 also comprises the step of revising data, specifically be meant in the physical state record sheet and seek untapped, and the new data of revising write in the future use block, physical state record sheet mark original block is wiped for postponing, and write-in block is for using.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100908332A CN100426259C (en) | 2005-08-18 | 2005-08-18 | Virtual access method of storage document data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100908332A CN100426259C (en) | 2005-08-18 | 2005-08-18 | Virtual access method of storage document data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1719422A true CN1719422A (en) | 2006-01-11 |
| CN100426259C CN100426259C (en) | 2008-10-15 |
Family
ID=35931261
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100908332A Expired - Fee Related CN100426259C (en) | 2005-08-18 | 2005-08-18 | Virtual access method of storage document data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100426259C (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101390062B (en) * | 2006-02-23 | 2011-02-23 | 飞思卡尔半导体公司 | Data processing system having address translation bypass and method therefor |
| CN102004706B (en) * | 2009-09-01 | 2012-09-19 | 联芯科技有限公司 | Flash erasing power-fail protection method based on FTL(Flash Translation Layer) |
| CN103034592A (en) * | 2012-12-05 | 2013-04-10 | 华为技术有限公司 | Data processing method and device |
| CN104281539A (en) * | 2013-07-10 | 2015-01-14 | 北京旋极信息技术股份有限公司 | Cache managing method and device |
| CN106339327A (en) * | 2015-07-06 | 2017-01-18 | 上海大学 | Computer system and blade server cabinet |
| CN108763559A (en) * | 2018-05-25 | 2018-11-06 | 广东电网有限责任公司 | A kind of date storage method, system, equipment and storage medium based on big data |
| CN109918228A (en) * | 2019-01-28 | 2019-06-21 | 浙江中控技术股份有限公司 | A kind of data power-down retaining method and system |
| CN110019073A (en) * | 2017-12-29 | 2019-07-16 | 北京忆恒创源科技有限公司 | The method and its file system of file system are provided using KV storage equipment |
| CN111061649A (en) * | 2019-10-28 | 2020-04-24 | 宁波三星智能电气有限公司 | Self-adaptive allocation method for storage space of memory |
| CN111742291A (en) * | 2018-02-07 | 2020-10-02 | 阿里巴巴集团控股有限公司 | Method and system for user space storage I/O stack with user space flash translation layer |
| CN112947862A (en) * | 2021-03-10 | 2021-06-11 | 歌尔科技有限公司 | Device, Flash memory and data storage method thereof |
| CN116166194A (en) * | 2023-01-16 | 2023-05-26 | 广州辰创科技发展有限公司 | Equipment guarantee maintenance and management method and system |
| CN116185563A (en) * | 2022-12-27 | 2023-05-30 | 上海千顾汽车科技有限公司 | Software simulation algorithm based on vehicle-gauge microcontroller data flash memory |
| CN116483287A (en) * | 2023-06-21 | 2023-07-25 | 宝德计算机系统股份有限公司 | Data security storage method and system for Internet users |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3417525B2 (en) * | 1996-10-31 | 2003-06-16 | 株式会社エヌ・ティ・ティ・データ | Composite card and composite card access system |
| US6205531B1 (en) * | 1998-07-02 | 2001-03-20 | Silicon Graphics Incorporated | Method and apparatus for virtual address translation |
| US6321276B1 (en) * | 1998-08-04 | 2001-11-20 | Microsoft Corporation | Recoverable methods and systems for processing input/output requests including virtual memory addresses |
| CN1258715C (en) * | 2003-03-20 | 2006-06-07 | 中国科学院计算技术研究所 | Virtual shared storage device and method |
-
2005
- 2005-08-18 CN CNB2005100908332A patent/CN100426259C/en not_active Expired - Fee Related
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101390062B (en) * | 2006-02-23 | 2011-02-23 | 飞思卡尔半导体公司 | Data processing system having address translation bypass and method therefor |
| CN102004706B (en) * | 2009-09-01 | 2012-09-19 | 联芯科技有限公司 | Flash erasing power-fail protection method based on FTL(Flash Translation Layer) |
| CN103034592A (en) * | 2012-12-05 | 2013-04-10 | 华为技术有限公司 | Data processing method and device |
| CN103034592B (en) * | 2012-12-05 | 2016-09-28 | 华为技术有限公司 | Data processing method and device |
| CN104281539A (en) * | 2013-07-10 | 2015-01-14 | 北京旋极信息技术股份有限公司 | Cache managing method and device |
| CN106339327A (en) * | 2015-07-06 | 2017-01-18 | 上海大学 | Computer system and blade server cabinet |
| CN110019073A (en) * | 2017-12-29 | 2019-07-16 | 北京忆恒创源科技有限公司 | The method and its file system of file system are provided using KV storage equipment |
| CN111742291A (en) * | 2018-02-07 | 2020-10-02 | 阿里巴巴集团控股有限公司 | Method and system for user space storage I/O stack with user space flash translation layer |
| CN111742291B (en) * | 2018-02-07 | 2024-04-05 | 阿里巴巴集团控股有限公司 | Method and system for user space storage I/O stack with user space flash translation layer |
| CN108763559A (en) * | 2018-05-25 | 2018-11-06 | 广东电网有限责任公司 | A kind of date storage method, system, equipment and storage medium based on big data |
| CN108763559B (en) * | 2018-05-25 | 2021-10-01 | 广东电网有限责任公司 | Data storage method, system, equipment and storage medium based on big data |
| CN109918228A (en) * | 2019-01-28 | 2019-06-21 | 浙江中控技术股份有限公司 | A kind of data power-down retaining method and system |
| CN111061649B (en) * | 2019-10-28 | 2023-09-29 | 宁波三星智能电气有限公司 | Storage space self-adaptive allocation method of memory |
| CN111061649A (en) * | 2019-10-28 | 2020-04-24 | 宁波三星智能电气有限公司 | Self-adaptive allocation method for storage space of memory |
| CN112947862A (en) * | 2021-03-10 | 2021-06-11 | 歌尔科技有限公司 | Device, Flash memory and data storage method thereof |
| CN116185563A (en) * | 2022-12-27 | 2023-05-30 | 上海千顾汽车科技有限公司 | Software simulation algorithm based on vehicle-gauge microcontroller data flash memory |
| CN116185563B (en) * | 2022-12-27 | 2024-04-16 | 上海千顾汽车科技有限公司 | Software simulation algorithm based on vehicle-gauge microcontroller data flash memory |
| CN116166194B (en) * | 2023-01-16 | 2023-11-03 | 广州辰创科技发展有限公司 | Equipment guarantee maintenance and management method and system |
| CN116166194A (en) * | 2023-01-16 | 2023-05-26 | 广州辰创科技发展有限公司 | Equipment guarantee maintenance and management method and system |
| CN116483287A (en) * | 2023-06-21 | 2023-07-25 | 宝德计算机系统股份有限公司 | Data security storage method and system for Internet users |
| CN116483287B (en) * | 2023-06-21 | 2024-02-06 | 宝德计算机系统股份有限公司 | Data security storage method and system for Internet users |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100426259C (en) | 2008-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1719422A (en) | Virtual access method of storage document data | |
| Chang et al. | Efficient management for large-scale flash-memory storage systems with resource conservation | |
| CN110678836B (en) | Persistent memory for key value storage | |
| KR102401596B1 (en) | A heuristic interface for enabling a computer device to utilize data property-based data placement inside a nonvolatile memory device | |
| CN1277213C (en) | A flash memory file system management method | |
| Wu et al. | An efficient B-tree layer implementation for flash-memory storage systems | |
| CN106708424B (en) | Apparatus and method for performing selective underlying exposure mapping on user data | |
| TWI514140B (en) | Non-volatile memory apparatus and operating method thereof | |
| JP5530012B2 (en) | Method and system for storage address remapping for storage devices | |
| CN103186350B (en) | The moving method of mixing storage system and hot spot data block | |
| Chiang et al. | Using data clustering to improve cleaning performance for flash memory | |
| CN1223945C (en) | Improved flash file system | |
| CN1295622C (en) | Address mapping method and mapping information managing method for flash memory, and flash memory using the same | |
| CN106708423B (en) | Multimode storage management system | |
| US20110099324A1 (en) | Flash memory storage system and flash memory controller and data processing method thereof | |
| CN103577339A (en) | Method and system for storing data | |
| CN102298555A (en) | Modular flash management system based on Not AND (NAND) technology | |
| TWI463313B (en) | Memory management method and memory controller and memory storage device using the same | |
| CN101030165A (en) | Magnetic disk space management and managing system | |
| Na et al. | Dynamic In-Page Logging for B⁺-tree Index | |
| CN110968269A (en) | SCM and SSD-based key value storage system and read-write request processing method | |
| Choudhuri et al. | Performance improvement of block based NAND flash translation layer | |
| Kim et al. | Exploiting compression-induced internal fragmentation for power-off recovery in SSD | |
| CN1896960A (en) | Fast photographic system and method | |
| CN1447242A (en) | Control device suitable to quick flash memory card and its construction methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20081015 Termination date: 20120818 |