CN101093463A - Novel algorithm in conversion layer for flash memory to implement fast writing in speed - Google Patents
Novel algorithm in conversion layer for flash memory to implement fast writing in speed Download PDFInfo
- Publication number
- CN101093463A CN101093463A CN 200610088088 CN200610088088A CN101093463A CN 101093463 A CN101093463 A CN 101093463A CN 200610088088 CN200610088088 CN 200610088088 CN 200610088088 A CN200610088088 A CN 200610088088A CN 101093463 A CN101093463 A CN 101093463A
- Authority
- CN
- China
- Prior art keywords
- data
- flash memory
- mirror image
- flash
- piece
- 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.)
- Pending
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A new flash memory conversion layer algorithm used for realizing quick writing-in speed of flash memory includes directly finding out a clean block in flash memory to write in new data when data of a sector size in flash memory block is revised, changing original flash memory block containing old data to be mirror image block correspondingly, treating that effective data placed separately in two blocks is still continuous data by PC point of view through data management function of flash memory conversion layer.
Description
Affiliated technical field
The present invention relates to the technical field of memory of consumption electronic product, refer in particular to the writing speed method that improves flash memory.Because the physical characteristics of flash memory, reading speed is generally very fast, and the present invention does not relate to.Annotate in addition: " flash translation layer (FTL) " abbreviates " FTL " in the text as.
Background technology
Over the past two years, consumption electronic product such as MP3 player enjoyed market to make a pet of.These electronic data such as song, picture even little film mainly all are on the flash memory (flash) that is stored in electronic product, and the user is by flash memory and computer exchange electronic data (as downloading, uploading).This just needs the flash memory of product is the storage medium the same with disk at computer.Computer uses the file system management disk, and between file system and flash memory, provide a flash translation layer (FTL) (Flash Translation Layer) exactly with its very natural method that operates on the flash memory, its function is exactly that flash memory with bottom is modeled to the same storage medium of disk (seeing Fig. 1, file system/flash memory hierarchical structure).
The work that FTL will do mainly is following two aspects:
1, organization space: the unit of file system deal with data is the sector.Flash memory is made up of big gauge block.Before the physical characteristics of flash memory required to write, the space of being write was a state of wiping back (totally), and the unit of wiping is a piece.Generally speaking, the size of piece is the size of several sectors, so FTL at first needs flash memory space is organized into the read-write space of file system needs.
2, management data: the data in the flash memory are that the scattered difference that is stored in is local, and the data that file system reads need be continuously linear storages.It is the mapping management that flash memory is arrived in the sector that FTL adopts, and makes the data of flash memory stored it seems it all is continuously linear the storage.
General, to write a sector data in flash memory the time when file system, FTL will do following operation and finish this write request (see Fig. 2, file system is write the operating process of a sector data to flash memory):
1, with in the whole buffer memory (buffer) of reading PC memory of the data of this place, sector flash block
2, content corresponding with this sector in the buffer memory is replaced with new content
3, this flash block is carried out erase operation
4, with whole this flash block that is written back to of the data in the buffer memory (comprising amended sector data)
The shortcoming of this write operation is clearly:
1, efficient is low: the renewal to a sector will be moved whole, cause the very big waste of data bandwidth, thereby speed is low.
2, the life-span is short: but these pieces in the flash memory have the maximum erasing times, and surpassing then becomes bad piece.According to this write operation mode, some piece can frequently be used and be become bad piece (when the user uses electronic product, promptly can repeatedly read, wipe, write flash memory that this file uses to the read-write that some files are frequent some), this flash memory can not re-use if bad number of blocks is crossed at most.
Summary of the invention
In order to solve the major defect of general FTL algorithm, the present invention proposes a kind of FTL algorithm of novelty, the writing speed of flash memory is more accelerated, more extend the serviceable life of flash memory.
The technical solution adopted for the present invention to solve the technical problems: (setting forth) in conjunction with Fig. 3, Fig. 4, Fig. 5, Fig. 6
Among Fig. 3, Fig. 4, Fig. 5, Fig. 6, A is that a quilt is write piece in the flash memory, B is a clean piece in the flash memory, and C is a simulated block (not being true piece, only for saying something usefulness), the 1st, sector to be revised, a is the first data segment that writes to sector to be changed a little of piece, and b is that sector to be changed writes the data segment that a little arrives the piece tail, and 1 ' is corresponding 1 amended sector, the 2nd, another sector to be changed (being included in the data segment a), 2 ' is corresponding 2 amended sectors; A ', b ' are the sectors after corresponding a, b move.
1 will be modified among the A, and FTL searches out B at random, will be 1 ' corresponding to 1 content update, is placed among the B.Because the data management function of FTL, this moment, a, 1 ', b still were a continuous data at PC, and a among the A, b resemble among Fig. 4 shown in the C in B seemingly.Thus, we are called the mirror image piece with A.
Data among a, the b also might be modified in the near future, so we exist when keeping A and B, in case the data among a and the b are modified, can be analogous to the operation of modification 1, new data are write the corresponding position of B, as Fig. 5.
Data in A are not made an amendment for a long time, and perhaps write operation once more no longer satisfies FTL algorithm condition, and FTL will eliminate mirror image piece A, if promptly also have valid data among the A then store into B, wipe A then, and A becomes clean piece, as Fig. 6.
This algorithm by the PC buffer memory, is not directly looked for clean piece in flash memory when flash memory is write data, and reduced manyly wipe, read-write operation, thereby improved the flash memory writing speed greatly.
The invention has the beneficial effects as follows:
The present invention utilizes mirror image piece notion, and the process of having avoided classic method to look for new piece, repetitive erasing repeatedly, write has repeatedly improved the speed that writes greatly.
The invention enables relatively more balanced being assigned on all pieces of the write operation of flash memory, thus balance wearing and tearing, thereby prolonged serviceable life of flash memory.
Description of drawings
Fig. 1 is file system/flash memory hierarchical structure synoptic diagram;
Fig. 2 is that file system is write the operational flowchart of a sector data to flash memory;
Fig. 3 is the synoptic diagram of mirror image piece A and storage block B original state;
Fig. 4 is mirror image piece A, the synoptic diagram of storage block B and simulated block C;
Fig. 5 is a synoptic diagram of revising data segment a;
Fig. 6 wipes the synoptic diagram that mirror image piece A becomes clean piece;
Fig. 7 is implementing procedure figure of the present invention.
Embodiment
Fig. 7 is implementing procedure figure of the present invention.
Below in conjunction with Fig. 7 embodiment is described:
When 1, flash memory being write a sector, check the state of this sector write-in block, monoblock totally then writes, and finishes; Do not change following 2;
2, checking whether current block has the mirror image piece, is to change down 3, does not change down 5;
3, checking and could directly write, do not change down 4, is to judge that then the mirror image piece that whether satisfies the FTL algorithm this moment eliminates condition, and the words of denying only need be upgraded the mirror image block message, write data, finish; The words that are are write this piece with mirror image piece valid data, wipe the mirror image piece then and include it in clean piece formation;
4, look for clean piece in addition, write data is also duplicated this piece and other valid data of mirror image piece, wipes this piece and mirror image piece then, includes them in clean piece formation, writes data, finishes;
5, check corresponding position, sector clean state in the piece, totally then write, finish; Do not change following 6;
6, whether check system also has available mirror image piece resource, is just to find a clean piece, writes the sector new data, and piece is first motionless to writing data a little and writing the data that a little arrive the piece tail, becomes the mirror image piece and generates the mirror image block message, end with former; Not, the FTL algorithm can discharge an available mirror image piece resource, carries out the same operation of this above-mentioned "Yes" then.
Claims (3)
1, a kind of flash translation layer (FTL) algorithm of realizing the novelty of the quick writing speed of flash memory, it is characterized in that: the notion of introducing the mirror image piece, that is: when revising the data of a sector-size in the flash block, directly in flash memory, look for clean piece to write new data, originally the corresponding mirror image piece that becomes of the flash block that comprised legacy data, by the data management function of flash translation layer (FTL), dividing the valid data of putting in two pieces still is continuous at PC.
2, a kind of flash translation layer (FTL) algorithm of realizing the novelty of the quick writing speed of flash memory according to claim 1, it is characterized in that: flash block and its mirror image piece exist simultaneously, when the data that also are not modified in the mirror image piece are modified again, amended data are write the clean position of the correspondence of flash block, upgrade the mirror image block message.
3, a kind of flash translation layer (FTL) algorithm of realizing the novelty of the quick writing speed of flash memory according to claim 1, it is characterized in that: the data in the mirror image piece are not made an amendment for a long time, perhaps write operation once more no longer satisfies flash translation layer (FTL) algorithm condition, flash translation layer (FTL) will be eliminated the mirror image piece, if also have valid data in the mirror image piece this moment, then it is stored into former flash block corresponding to the mirror image piece, discharge the mirror image block space then, become clean piece, include it in clean piece formation, recycle.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 200610088088 CN101093463A (en) | 2006-06-23 | 2006-06-23 | Novel algorithm in conversion layer for flash memory to implement fast writing in speed |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 200610088088 CN101093463A (en) | 2006-06-23 | 2006-06-23 | Novel algorithm in conversion layer for flash memory to implement fast writing in speed |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101093463A true CN101093463A (en) | 2007-12-26 |
Family
ID=38991736
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 200610088088 Pending CN101093463A (en) | 2006-06-23 | 2006-06-23 | Novel algorithm in conversion layer for flash memory to implement fast writing in speed |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101093463A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101364438B (en) * | 2007-08-08 | 2011-04-06 | 奇岩电子股份有限公司 | Device and method for improving NAND gate array flash memory access |
CN102346691A (en) * | 2010-07-30 | 2012-02-08 | 联芯科技有限公司 | Method for manufacturing flash image files and device |
CN104699623A (en) * | 2015-02-12 | 2015-06-10 | 北京天诚盛业科技有限公司 | Method and device for fastening Flash write operation |
CN105845182A (en) * | 2016-03-18 | 2016-08-10 | 华南理工大学 | File system level non-volatile memory wear balancing free block management method |
-
2006
- 2006-06-23 CN CN 200610088088 patent/CN101093463A/en active Pending
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101364438B (en) * | 2007-08-08 | 2011-04-06 | 奇岩电子股份有限公司 | Device and method for improving NAND gate array flash memory access |
CN102346691A (en) * | 2010-07-30 | 2012-02-08 | 联芯科技有限公司 | Method for manufacturing flash image files and device |
CN102346691B (en) * | 2010-07-30 | 2013-07-24 | 联芯科技有限公司 | Method for manufacturing flash image files and device |
CN104699623A (en) * | 2015-02-12 | 2015-06-10 | 北京天诚盛业科技有限公司 | Method and device for fastening Flash write operation |
CN104699623B (en) * | 2015-02-12 | 2018-09-11 | 北京天诚盛业科技有限公司 | Accelerate the method and device of Flash write operations |
CN105845182A (en) * | 2016-03-18 | 2016-08-10 | 华南理工大学 | File system level non-volatile memory wear balancing free block management method |
CN105845182B (en) * | 2016-03-18 | 2019-07-12 | 华南理工大学 | The non-volatility memorizer abrasion equilibrium free time block management method of file system level |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CA2574756C (en) | Systems, methods, computer readable medium and apparatus for memory management using nvram | |
CN106527969B (en) | A life-balanced NandFlash memory reading and writing method | |
CN105190526B (en) | To the memory re-mapping for non-volatile memory device | |
TWI382421B (en) | Hybrid density memory storage device | |
TWI398770B (en) | Data accessing method for flash memory and storage system and controller using the same | |
CN101727295B (en) | Data writing and reading method based on virtual block flash memory address mapping | |
CN108021510A (en) | The method for operating the storage device being managed to multiple name space | |
CN100501868C (en) | Implementation method of file system based on NAND Flash memory | |
CN101571832B (en) | Data writing method, flash storage system using the method and its controller | |
CN106354658B (en) | A method of it reducing mapping table memory source in mixed-use developments algorithm and occupies | |
CN103559138A (en) | Solid state disk (SSD) and space management method thereof | |
CN102779096A (en) | Page, block and face-based three-dimensional flash memory address mapping method | |
CN107239526A (en) | File system implementation method, scrap cleaning method, operating position localization method | |
CN103544110A (en) | Block-level continuous data protection method based on solid-state disc | |
CN110018966A (en) | A kind of memory, storage system, host and data manipulation, rubbish recovering method | |
CN101093463A (en) | Novel algorithm in conversion layer for flash memory to implement fast writing in speed | |
CN101625661B (en) | Data management method, storage system and controller for flash memory | |
CN100470677C (en) | Method for increasing speed of writing data into flash memory disk | |
CN101216806B (en) | Data updating method and device | |
US6532513B1 (en) | Information recording and reproduction apparatus | |
KR101020781B1 (en) | How to Manage Logs for Database Systems Based on Flash Memory | |
CN105630406A (en) | Storage system using MRAM as editing buffer and editing buffer method | |
CN111949212B (en) | File system and file management method based on self-defined open channel SSD | |
CN107688442A (en) | A kind of virtual block management method for solid state hard disc | |
TWI354892B (en) | Method of immediate data update with flash memory |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C57 | Notification of unclear or unknown address | ||
DD01 | Delivery of document by public notice |
Addressee: Tang Xiaoyan Document name: Written notice of preliminary examination of application for patent for invention |
|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |