`Gorobets et al.
`
`(10) Patent No.:
`(45) Date of Patent:
`
`US 8,250,333 B2
`Aug. 21, 2012
`
`USOO8250333B2
`
`2005/0172082 A1* 8, 2005 Liu et al. ....................... 711/144
`2007/0101.095 A1
`5/2007 Gorobets
`2007/0300037 A1* 12/2007 Rogers et al. ................. T11 202
`2008. O162792 A1* 7, 2008 Wu et al. .....
`... 711,103
`2008. O162793 A1* 7, 2008 Chu et al. ....
`... 711,103
`2008. O177935 A1* 7, 2008 Lasser et al.
`... 711,103
`2010.0030999 A1
`2/2010 Hinz ............................. T11 206
`
`(54)
`54) MLAPPING ADDRESS TABLE MAINTENANCE
`INA MEMORY DEVICE
`
`(75) Inventors: Sergey Anatolievich Gorobets,
`Edinburgh (GB); Alexander Paley,
`Kfar-Saba (IL); Eugene Zilberman,
`Richmond Hill (CA); Alan David
`Bennett, Edinburgh (GB); Shai Traister,
`San Jose, CA (US)
`
`(*) Notice:
`
`73) Assignee: Sandisk Technologies Inc., Plano, TX
`9.
`g
`(US)
`Subject to any disclaimer, the term of this
`past lSolisted under 35
`M
`YW-
`y
`yS.
`(21) Appl. No.: 12/348,782
`(22) Filed:
`Jan. 5, 2009
`
`OTHER PUBLICATIONS
`International Search Report issued in International Application No.
`PCT/US2010/020027, mailed Apr. 23, 2010 (6 pages).
`Written Opinion issued in International Application No. PCT/
`US2010/020027, mailed Apr. 23, 2010 (7 pages).
`* cited by examiner
`Primary Examiner — Kaushikkumar Patel
`(74) Attorney, Agent, or Firm — Brinks Hofer Gilson &
`Lione
`ABSTRACT
`(57)
`A method and system maintains an address table for mapping
`logical groups to physical addresses in a memory device. The
`method includes receiving a request to set an entry in the
`address table and selecting and flushing entries in an address
`table cache depending on the existence of the entry in the
`cache and whether the cache meets a flushing threshold cri
`teria. The flushed entries include less than the maximum
`capacity of the address table cache. The flushing threshold
`criteria includes whether the address table cache is full or if a
`page exceeds a threshold of changed entries. The address
`table and/or the address table cache may be stored in a non
`Volatile memory and/or a random access memory. Improved
`performance may result using this method and system due to
`U.S. PATENT DOCUMENTS
`the reduced number of write operations and time needed to
`ck
`595. f ck 3.283 FGS et al... 3.06 partially flush the address table cache to the address table.
`2004/008.84.74 A1* 5, 2004 Lin ............................... T11 103
`2004/O186946 A1
`9, 2004 Lee
`22 Claims, 9 Drawing Sheets
`
`(65)
`
`Prior Publication Data
`US 2010/O1748.69 A1
`Jul. 8, 2010
`s
`
`(51) Int. Cl.
`(2006.01)
`G06F 2/12
`(52) U.S. Cl. ......................... 711/207 711/135: 711/103
`(58) Field of Classification Search s
`s
`None
`See application file for com lete search history
`pp
`p
`ry.
`References Cited
`
`(56)
`
`800
`
`request to Set GATEntry
`802.
`
`
`
`
`
`Yes
`
`GATEntry
`Exists in GAT
`Delta?
`804
`
`
`
`GAT Delta Meets
`Criteria?
`
`Select GATPage to update
`808
`
`Flush Changed entries from
`GAT Delta to GAT 810
`
`Allocate entry in GATDalia
`812
`
`-- Set GATEntry in GATD
`
`Ed
`
`Micron Ex. 1053, p. 1
`Micron v. Vervain
`IPR2021-01550
`
`
`
`U.S. Patent
`
`Aug. 21, 2012
`
`Sheet 1 of 9
`
`US 8,250,333 B2
`
`-OS is:
`
`
`
`EtoxY SYSex: x
`
`808: i:
`
`W
`
`is settery 3:
`as 8tettery
`
`*:::gr:8:3:38
`8:::::::::::: 88883:y
`
`Micron Ex. 1053, p. 2
`Micron v. Vervain
`IPR2021-01550
`
`
`
`U.S. Patent
`
`Aug.21, 2012
`
`Sheet 2 of 9
`
`US 8,250,333 B2
`
`HOST 70
`
`Hostsiie Memory Manager Sopllonals
`
`Application
`
`OS!
`File System
`
`husieralLogical sectons}
`
`| Logica! sectora
`
`MEMORY SYSTEM20
`
`Flash Memory260
`
`Manzacar
`
`
`MMaetablock Link
`
`
`FIG. 2
`
`Micron Ex. 1053, p. 3
`Micron v. Vervain
`IPR2021-01550
`
`i Logeal io Pirgsinal
`Adkireas
`Translation
`
`petao
`
`1.
`
`oe TOU
`
`Manager
`Meraory-side Merary
`
`
`
`
`
` Uncale Binck
`idarger
`
`
` Erased Block
` $70
`
`Manager
`
`Micron Ex. 1053, p. 3
`Micron v. Vervain
`IPR2021-01550
`
`
`
`U.S. Patent
`
`Aug. 21, 2012
`
`Sheet 3 of 9
`
`US 8,250,333 B2
`
`i.cgieai stat:
`
`{i}
`
`
`
`trysikai groxip
`itetaticsek.
`
`3giea & exts
`
`
`
`ity-sixai Srok
`exixitxik
`
`i.ogicai ix3
`hysica:
`irectories
`
`FIG. 3B
`
`Micron Ex. 1053, p. 4
`Micron v. Vervain
`IPR2021-01550
`
`
`
`U.S. Patent
`
`Aug. 21, 2012
`
`Sheet 4 of 9
`
`US 8,250,333 B2
`
`a -a aaa aaaa-, -a-, -a-
`
`i.&ge:388:38
`
`controller too
`
`-(8 xer: Act is:
`
`
`
`
`
`
`
`
`Xs: 8 (383-838.
`8:888.88:8888
`
`exi Setia
`
`
`
`
`
`
`
`
`
`*iasis is try :
`... 3
`3:3 &xiki:88
`tabies {&A
`
`A psia
`cat Dera
`
`-
`
`FIG. 4
`
`Micron Ex. 1053, p. 5
`Micron v. Vervain
`IPR2021-01550
`
`
`
`U.S. Patent
`
`
`
`US 8,250,333 B2
`
`
`
`00G
`
`Z09
`
`Micron Ex. 1053, p. 6
`Micron v. Vervain
`IPR2021-01550
`
`
`
`U.S. Patent
`
`Aug. 21, 2012
`
`Sheet 6 of 9
`
`eedLVS
`
`saolpu|aBed1VD
`
`
`eyedLVSt
`
`(€99L"8rZLOT
`
`¢a6edLV
`
`
`
`
`
`
`(3(éve,"%e897
`(0(Les"9lyOT
` Zzabed1VD|abed
`
`1VD
`
`(Sbr097
`
`ee Ee ee
`
`US 8,250,333 B2
`
`9Sid
`
`Micron Ex. 1053, p. 7
`Micron v. Vervain
`IPR2021-01550
`
`Micron Ex. 1053, p. 7
`Micron v. Vervain
`IPR2021-01550
`
`
`
`
`
`
`
`
`
`
`U.S. Patent
`U.S. Patent
`
`Aug.21, 2012
`
`Sheet 7 of 9
`
`eyedLVD
`
`SyOO|G-e}aWMAN
`
`
`
`SQ]USUM-3y
`
`
`
`009
`
`
`
`abed>4y8eX@pU|
`
`
`
`
`
`
` “00s
`
`
`
`ad5)dqv'(Ja)Is701gaely|Loeeneeennneeoeeeeeketelewaead
`
`
`US 8,250,333 B2
`US 8,250,333 B2
`
`Micron Ex. 1053, p. 8
`Micron v. Vervain
`IPR2021-01550
`
`
`
`
`
`9)(es"9lrOT)|Abed1VvO|.
`
`(SL7097)0abedLVD
`
`Micron Ex. 1053, p. 8
`Micron v. Vervain
`IPR2021-01550
`
`
`
`
`
`U.S. Patent
`
`Aug. 21, 2012
`
`Sheet 8 of 9
`
`US 8,250,333 B2
`
`800
`
`Request to Set GAT Entry
`802
`
`Yes
`
`
`
`
`
`GAT Entry
`Exists in GAT
`Delta?
`804
`NO
`
`
`
`GAT Delta Meets\ N
`Criteria?
`806
`
`
`
`Yes
`
`Select GAT Page to Update
`808
`
`Flush Changed Entries from
`GAT Delta to GAT 810
`
`Allocate Entry in GAT Delta
`812
`
`Set GAT Entry in GAT Dglia
`
`End
`
`FIG. 8
`
`Micron Ex. 1053, p. 9
`Micron v. Vervain
`IPR2021-01550
`
`
`
`U.S. Patent
`
`Aug. 21, 2012
`
`Sheet 9 Of 9
`
`US 8,250,333 B2
`
`900
`
`Write to GAT Delta Copy in
`RAM
`800
`
`
`
`
`
`Flush GAT Delta
`Copy in RAM to
`GAT Delta in Flash?
`
`
`
`Update GAT Delta in Flash
`With GAT Delta in RAM
`
`FIG. 9
`
`Micron Ex. 1053, p. 10
`Micron v. Vervain
`IPR2021-01550
`
`
`
`US 8,250,333 B2
`
`1.
`MAPPING ADDRESS TABLE MAINTENANCE
`INA MEMORY DEVICE
`
`TECHNICAL FIELD
`
`This application relates generally to memory devices.
`More specifically, this application relates to maintenance of
`logical to physical mapping address tables in reprogram
`mable non-volatile semiconductor flash memory.
`
`BACKGROUND
`
`10
`
`15
`
`25
`
`30
`
`35
`
`2
`pages. The method also includes determining whether the
`entry exists in an address table cache that stores changes to the
`address table and determining whether the address table
`cache meets a flushing threshold criteria. If the entry does not
`exist in the address table cache and the address table cache
`meets the flushing threshold criteria, a quantity of pages of the
`address table is selected. The quantity of pages selected is less
`than the total number of pages in the address table. The pages
`in the address table include changed and unchanged entries.
`Changed entries in the selected pages are flushed from the
`address table cache to the address table. The requested entry
`is allocated and set in the address table cache.
`In some embodiments, the flushing threshold criteria may
`include the maximum capacity of the address table cache, and
`determining whether the address table cache meets the crite
`ria includes determining whether the number of entries in the
`address table cache is at the maximum capacity of the address
`table cache. In other embodiments, the flushing threshold
`criteria may include a threshold of changed entries, and deter
`mining whether the address table cache meets the criteria
`includes determining whether a number of changed entries in
`the pages of the address table exceeds the threshold of
`changed entries.
`Selecting the quantity of pages to flush in the address table
`may include selecting the pages with the greatest number of
`changed entries. Alternatively, selecting the quantity of pages
`to flush in the address table may include selecting the pages
`that have a number of changed entries above a predetermined
`threshold of changed entries. The quantity of pages selected
`to be flushed may be one. Flushing changed entries from the
`address table cache to the address table may include updating
`the entries in the address table with the changed entries in the
`address table cache.
`The method may further include updating the existing
`entry for the logical group in the address table cache if the
`entry already exists in the address table cache. The method
`may also include allocating and setting the entry for the
`logical group in the address table cache if the entry does not
`exist in the address table cache and the address table cache
`does not meet the flushing threshold criteria. The address
`table and/or the address table cache may be stored in one or
`more of a non-volatile memory or a random access memory.
`According to another aspect, a memory device includes an
`address table for mapping logical groups to physical address
`in the memory device, an address table cache that stores
`changes to the address table, and a controller. The controller
`is configured to receive a request to set an entry in the address
`table, where the entry maps a logical group to a physical
`address. The address table includes a plurality of pages. The
`controller is also configured to determine whether the entry
`exists in an address table cache and determine whether the
`address table cache meets a flushing threshold criteria. If the
`controller determines the entry does not exist in the address
`table cache and that the address table cache meets the flushing
`threshold criteria, a quantity of pages of the address table is
`selected. The quantity of pages selected is less than the total
`number of pages in the address table. The pages in the address
`table include changed and unchanged entries. The controller
`flushes changed entries in the selected pages from the address
`table cache to the address table. The requested entry is allo
`cated and set in the address table cache by the controller.
`In some embodiments, the flushing threshold criteria may
`include the maximum capacity of the address table cache, and
`determining whether the address table cache meets the crite
`ria includes the controller being configured to determine
`whether the number of entries in the address table cache is at
`the maximum capacity of the address table cache. In other
`
`Non-volatile memory systems, such as flash memory, have
`been widely adopted for use in consumer products. Flash
`memory may be found in different forms, for example in the
`form of a portable memory card that can be carried between
`host devices or as a solid state disk (SSD) embedded in a host
`device. When writing data to a conventional flash memory
`system, a host typically writes data to, and reads data from,
`addresses within a logical address space of the memory sys
`tem. The memory system then commonly maps data between
`the logical address space and the physical blocks or meta
`blocks of the memory, where data is stored in fixed logical
`groups corresponding to ranges in the logical address space.
`Generally, each fixed logical group is stored in a separate
`physical block of the memory system. The memory system
`keeps track of how the logical address space is mapped into
`the physical memory but the host is unaware of this. The host
`keeps track of the addresses of its data files within the logical
`address space but the memory system generally operates
`without knowledge of this mapping.
`An address table in the memory system includes the map
`ping of the logical address space to the physical memory. In
`particular, the address table includes pages indicating the
`mapping of logical groups to physical blocks in the memory
`system. When the host writes data to logical groups that have
`already been mapped, the address table may be updated with
`the pertinent mapping information.
`Some memory systems contain a cache to the address table
`to temporarily store changes to the address table when data is
`written. Writing to an address table cache instead of to the
`address table may save some time and write operation over
`head. Memory systems with an address table cache may
`periodically synchronize the changed entries in the cache
`with the address table by updating the entire cache to the
`address table, regardless of the amount of actual changed
`entries in the cache. However, in a large memory system that
`may have over one hundred pages in the address table, updat
`ing the entire address table with changed entries in the address
`table cache may negatively affect performance and delay
`other operations in the memory system. Pages in the address
`table may be unnecessarily rewritten if no changes were
`made. The flash memory cells used to store the address table
`may also be worn out prematurely when the entire address
`table is written.
`
`40
`
`45
`
`50
`
`55
`
`SUMMARY
`
`In order to address the problems noted above, a method and
`system for maintaining an address table mapping logical
`groups to physical addresses is disclosed.
`According to a first aspect of the invention, a method is
`disclosed for maintaining an address table for mapping logi
`cal groups to physical addresses in a memory device. The
`method includes receiving a request to set an entry in the
`address table, where the entry maps a logical group to a
`physical address. The address table includes a plurality of
`
`60
`
`65
`
`Micron Ex. 1053, p. 11
`Micron v. Vervain
`IPR2021-01550
`
`
`
`US 8,250,333 B2
`
`10
`
`15
`
`25
`
`30
`
`35
`
`3
`embodiments, the flushing threshold criteria may include a
`threshold of changed entries, and determining whether the
`address table cache meets the criteria includes the controller
`being configured to determine whether a number of changed
`entries in the pages of the address table exceeds the threshold
`of changed entries.
`Selecting the quantity of pages to flush in the address table
`may include selecting the pages with the greatest number of
`changed entries. Alternatively, selecting the quantity of pages
`to flush in the address table may include selecting the pages
`that have a number of changed entries above a predetermined
`threshold of changed entries. The quantity of pages selected
`to be flushed may be one. Flushing changed entries from the
`address table cache to the address table may include updating
`the entries in the address table with the changed entries in the
`address table cache.
`The controller may be further configured to update the
`existing entry for the logical group in the address table cache
`if the entry already exists in the address table cache. The
`controller may also be configured to allocate and set the entry
`for the logical group in the address table cache if the entry
`does not exist in the address table cache and the address table
`cache does not meet the flushing threshold criteria. The
`address table and/or the address table cache may be stored in
`one or more of a non-volatile memory or a random access
`memory.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 illustrates the main hardware components of a
`memory system Suitable for implementing embodiments of
`the invention.
`FIG. 2 illustrates the memory being organized into physi
`cal groups of sectors (or metablocks) and managed by a
`memory manager of the controller, according to an embodi
`ment.
`FIG. 3A illustrates the mapping between a logical group
`and a metablock, according to an embodiment.
`FIG. 3B illustrates the mapping between logical groups
`and metablocks.
`FIG. 4 is a schematic block diagram of the metablock
`management system as implemented in the controller and
`flash memory.
`FIG. 5 illustrates the group address table block in an initial
`State.
`FIG. 6 illustrates the group address table block after data is
`re-written.
`FIG. 7 illustrates the group address table block after a
`group address table cache update.
`FIG. 8 is a flow diagram illustrating a method of maintain
`ing a group address table and group address table cache that
`map logical groups to physical addresses.
`FIG. 9 is a flow diagram illustrating a method of maintain
`ing a group address table cache in random access memory and
`flash memory.
`
`4
`prises one or more array of non-volatile memory cells distrib
`uted over one or more integrated circuit chips. The controller
`100 includes an interface 110, a processor 120, an optional
`coprocessor 121, ROM 122 (read only memory), RAM 130
`(random access memory) and optionally programmable non
`volatile memory 124. The interface 110 has one component
`interfacing the controller to a host and another component
`interfacing to the memory 200. Firmware stored in nonvola
`tile ROM 122 and/or the optional nonvolatile memory 124
`provides code for the processor 120 to implement the func
`tions of the controller 100. Error correction codes may be
`processed by the processor 120 or the optional coprocessor
`121. In an alternative embodiment, the controller 100 is
`implemented by a state machine (not shown). In yet another
`embodiment, the controller 100 is implemented within the
`host.
`FIG. 2 illustrates the memory being organized into physi
`cal groups of sectors (or metablocks) and managed by a
`memory manager of the controller, according to an embodi
`ment. The memory 200 is organized into metablocks
`MB, ..., MB, where each metablock is a group of physical
`sectors So..., S. that are erasable together.
`The host 10 accesses the memory 200 when running an
`application under a file system or operating system. Typically,
`the host system addresses data in units of logical sectors
`where, for example, each sector may contain 512 bytes of
`data. Also, it is usual for the host to read or write to the
`memory system in units of logical clusters, each consisting of
`one or more logical sectors. In some host systems, an optional
`host-side memory manager may exist to perform lower level
`memory management at the host. In most cases during read or
`write operations, the host 10 essentially issues a command to
`the memory system 20 to read or write a segment containing
`a string of logical sectors of data with contiguous addresses.
`A memory-side memory manager is implemented in the
`controller 100 of the memory system 20 to manage the stor
`age and retrieval of the data of host logical sectors among
`metablocks of the flash memory 200. In the preferred embodi
`ment, the memory manager contains a number of Software
`modules for managing erase, read and write operations of the
`metablocks. The memory manager also maintains system
`control and directory data associated with its operations
`among the flash memory 200 and the controller RAM 130.
`FIGS. 3A(i)-3A(iii) illustrate the mapping between a logi
`cal group and a metablock, according to an embodiment. The
`metablock of the physical memory has N physical sectors for
`storing N logical sectors of data of a logical group. FIG.3A(i)
`shows the data from a logical group LG, where the logical
`sectors are in contiguous logical order 0, 1,...,N-1. FIG.
`3A(ii) shows the same data being stored in the metablock in
`the same logical order. The metablock when stored in this
`manner is said to be "sequential. In general, the metablock
`may have data stored in a different order, in which case the
`metablock is said to be “non-sequential or “chaotic.”
`There may be an offset between the lowest address of a
`logical group and the lowest address of the metablock to
`which it is mapped. In this case, the logical sector address
`wraps around as a loop from the bottom back to the top of the
`logical group within the metablock. For example, in FIG.
`3A(iii), the metablock stores in its first location beginning
`with the data of logical sector k. When the last logical sector
`N-1 is reached, it wraps around to sector 0 and finally stores
`data associated with logical sector k-1 in its last physical
`sector. In the preferred embodiment, a page tag is used to
`
`40
`
`45
`
`50
`
`55
`
`BRIEF DESCRIPTION OF THE PRESENTLY
`PREFERRED EMBODIMENTS
`
`FIG. 1 illustrates the main hardware components of a
`memory system Suitable for implementing embodiments of
`the invention. The memory system 20 typically operates with
`a host 10 through a host interface. The memory system is
`typically in the form of a memory card or an embedded
`memory system, such as a solid state disk (SSD) drive. The
`memory system 20 includes a memory 200 whose operations
`are controlled by a controller 100. The memory 200 com
`
`60
`
`65
`
`Micron Ex. 1053, p. 12
`Micron v. Vervain
`IPR2021-01550
`
`
`
`5
`identify any offset, such as identifying the starting logical
`sector address of the data stored in the first physical sector of
`the metablock. Two blocks will be considered to have their
`logical sectors stored in similar order when they only differ by
`a page tag.
`FIG. 3B illustrates the mapping between logical groups
`and metablocks. Each logical group is mapped to a unique
`metablock, except for a small number of logical groups in
`which data is currently being updated. After a logical group
`has been updated, it may be mapped to a different metablock.
`The mapping information is maintained in a set of logical to
`physical directories, such as a group address table and group
`address table cache, as described below.
`FIG. 4 is a block diagram of the metablock management
`system as implemented in the controller and flash memory.
`The metablock management system comprises various func
`tional modules implemented in the controller 100 and main
`tains various control data in tables and lists in the flash
`memory 200 and the controller RAM 130. The function mod
`ules implemented in the controller 100 includes an interface
`module 110 and a logical-to-physical address translation
`module 140. The interface 110 allows the metablock manage
`ment system to interface with a host system. The logical to
`physical address translation module 140 maps the logical
`address from the host to a physical memory location.
`During operation the metablock management system gen
`erates and works with control data Such as addresses, control
`and status information. Since much of the control data tends
`to be frequently changing data of Small size, it cannot be
`readily stored and maintained efficiently in a flash memory
`with a large block structure. A hierarchical and distributed
`scheme may be employed to store the more static control data
`in the nonvolatile flash memory while locating the smaller
`amount of the more varying control data in controller RAM
`for more efficient update and access. In the event of a power
`35
`shutdown or failure, the scheme allows the control data in the
`volatile controller RAM to be rebuilt quickly by scanning a
`Small set of control data in the nonvolatile memory.
`The non-volatile flash memory 200 may store control data
`such as the group address table (GAT) 210 and the group
`address table cache (GAT Delta) 220. The GAT 210 keeps
`track of the mapping between logical groups of sectors and
`their corresponding metablocks. The GAT 210 contains one
`entry for each logical group, ordered sequentially according
`to logical address. The GAT 210 includes a plurality of pages
`with each page including entries defining metablock
`addresses for every logical group in the memory system. The
`GAT Delta 220 acts as a cache that is a list of changed entries
`in the mappings of the GAT 210. In one embodiment, the GAT
`210 and GAT Delta 220 are both stored in the flash memory
`50
`200. Flushing of changed entries from the GAT Delta 220 to
`the GAT 210 take place within the flash memory 200 in this
`embodiment.
`In some embodiments, the RAM 130 may include a GAT
`Delta Copy 132. The GAT Delta Copy 132 may contain the
`55
`same list of changed entries as in the GAT Delta 220. Peri
`odically, the controller may synchronize the GAT Delta Copy
`132 and the GAT Delta 220 so that they contain the same
`information. This process is detailed more below in reference
`to FIG. 9.
`FIGS. 5-7 illustrate the group address table (GAT) block
`500 in (1) an initial state with an empty GAT Delta; (2) after
`logical groups are re-written with an updated GAT Delta; and
`(3) after the GAT Delta is partially flushed to the GAT. FIG. 8
`is a flow diagram illustrating a method 800 of maintaining a
`group address table and group address table cache that map
`logical groups to physical addresses. Each of the steps
`
`30
`
`40
`
`45
`
`60
`
`65
`
`US 8,250,333 B2
`
`10
`
`15
`
`25
`
`6
`described in FIG. 8 for the method 800 may be performed
`alone or in combination with other steps.
`FIG. 5 illustrates the group address table (GAT) block 500
`in an initial state. The GAT block 500 includes a master index
`page 502 containing a free block list (FBL) 504, a GAT Delta
`506, and a GAT 508 that includes a plurality of pages, where
`each page includes entries mapping metablockaddresses for
`logical groups that have been written to. The FBL 504 lists
`available free blocks that may be later mapped to logical
`groups. The FBL 504 may be in the order the free blocks were
`previously allocated. In FIG. 5, the exemplary FBL 504 lists
`metablocks F, G, H, J, and Kas free blocks.
`In an initial state of the memory system, written logical
`groups are already assigned to physical metablocks in entries
`of the pages in the GAT508. The exemplary GAT508 in FIG.
`5 includes Pages 0, 1, 2, and 3 that each has 416 entries
`corresponding to logical groups. The GAT Delta 506 is empty
`in the initial state because no changes have been made to the
`GAT508 yet. In other words, in the initial state shown in FIG.
`5, the GAT 508 contains the most updated mapping for the
`logical groups to physical metablocks.
`FIG. 6 illustrates the GAT block 500 after data is re-written
`by the host. The old copy of the master index page 502 shown
`in FIG.5 is not shown in FIG. 6. When data is written from the
`host, a request to set an entry in the GAT 508 mapping a
`logical group to a physical metablock may be received at step
`802 of the method 800 shown in FIG. 8. Because the GAT
`Delta 506 may contain more recent mapping information than
`the GAT508, the GAT Delta 506 is checked to see if the entry
`specified in the set request already exists at step 804. If the
`entry already exists at step 804, then it is updated with the new
`mapping information from the request at step 814. In some
`embodiments, if the entry already exists in the GAT Delta
`506, the corresponding page in the GAT 508 containing the
`logical group for the already existing entry may be updated
`immediately from the GAT Delta 506. In other embodiments,
`if the entry already exists in the GAT Delta 506, a new entry
`may be allocated and set in the GAT Delta 506 for the logical
`group specified in the request.
`However, if the entry does not already exist in the GAT
`Delta 506 at step 804, the entry may be allocated and set if a
`flushing threshold criteria is not met at step 806. In this case,
`the entry is allocated and set in the GAT Delta 506 based on
`the request at steps 812 and 814. The flushing threshold
`criteria includes the GAT Delta 506 reaching its maximum
`capacity, if one or more pages in the GAT 508 have a number
`of changed entries above a threshold, or after a certain time
`period has elapsed. Flushing threshold criteria may also
`include doing a preemptive flush if the master index page 502
`is updated for other reasons.
`Another flushing threshold criteria may include a case
`when the GAT 508 is compacted. In one embodiment, when
`the GAT 508 is compacted, valid pages in the GAT508 are
`copied from the compacted GAT block and the updated indi
`ces for the GAT508 are written to the master index page 502.
`In another embodiment, when the GAT 508 is compacted,
`pages in the GAT 508 are copied and updated with changed
`entries from the GAT Delta 506. In this compaction case, the
`GAT Delta 506 may be partially or fully flushed after com
`paction and updating of the GAT508. At a minimum, in this
`compaction case, the GAT Delta 506 would no longer include
`any entries for the pages in the GAT508 that were compacted.
`Therefore, if the entry does not exist in the GAT Delta 506
`and the flushing threshold criteria is not met at step 806, the
`GAT Delta 506 is not flushed and synchronized with the GAT
`508. When the host rewrites data to logical groups, the cor
`responding physical metablock the data is written to is
`
`Micron Ex. 1053, p. 13
`Micron v. Vervain
`IPR2021-01550
`
`
`
`7
`recorded in entries in the GAT Delta 506 at steps 812and814
`instead of directly in the GAT508. For example, in FIG. 6, the
`host rewrites logical groups 410, 411, 520, 413, and 1101. A
`request to set entries for these logical groups may be received
`at step 802. Because the GAT Delta 506 is empty at step 804,
`the entries for these logical groups do not already exist in the
`GAT Delta 506 of FIG. 5. In addition to the entries non
`existence in the GAT Delta 506, because the flushing thresh
`old criteria for the GAT Delta 506 is not met at step 806, the
`entries for the logical groups are allocated and set at steps 812
`and 814. Following steps 812and814, the GAT Delta 506 is
`no longer empty and includes entries mapping logical groups
`410,411,520, 413, and 1101 to physical metablocks F, G, H,
`J, and K, respectively. The memory controller writes the data
`for these blocks into free blocks F, G, H, J, and K, respec
`tively, based on the available free blocks listed in FBL 504.
`The logical groups 410, 411, 520, 413, and 1101 had pre
`viously been mapped to physical metablocks A, B, C, D, and
`E, respectively. At this point, the GAT 508 still contains this
`original mapping, however, the entries for these logical
`groups in the GAT 508 are now superseded by the entries in
`the GAT Delta 506. FIG. 6 shows the superseded physical
`metablocks A, B, C, D, and E in the GAT 508 as grayed out.
`Because physical metablocks A, B, C, D, and E no longer
`contain valid data, the FBL 504 now lists these metablocks as
`free blocks.
`The masterindex page 502 that includes the FBL 504, GAT
`Delta 506 and GAT 508 may be updated in a single write
`operation and/or be contained within a single data structure.
`The master index page 502 may also include other informa
`tion that is updated in the single write operation, e.g., update
`block information (e.g., logical groups that have been
`updated, block locations, and written length), pointers to
`active binary cache blocks, indices for pages of the GAT508,
`wear leveling counters, and other information. The master
`index page 502 provides a synchronous snapshot of the tables
`contained within and does not need to be entirely rewritten
`after every write to a page of the GAT508.
`All physical block references in the master index page 502
`are updated in the single write operation Such that there are no
`lost blocks or double-referenced blocks. All physical blocks
`are referenced by the GAT508, the master index page 502, or
`in control blocks (e.g., pointers to blocks in the GAT 508,
`binary cache blocks, etc.). When a block is taken from the
`FBL 504, the block may optionally be referenced in the
`update block information in the master index page 502. The
`block taken from the FBL 504 is then referenced by the GAT
`Delta 506 and the GAT 508. The reference for a block is
`changed in multiple places when a single write operation is
`performed on the master index page 502, e.g., in the FBL 504,
`where a new allocated block disappears and a new free block
`appears, and in the GAT 508 as a new reference. Therefore,
`instead of updating the FBL 504 and the FBL 508 at the same
`time using an atomic write operation with separate writes, a
`single write operation on the master index page 502 collec
`tively updates the FBL 504, the GAT508, block information,
`and the logical-to-physical table.
`FIG. 7 illustrates the GAT block 500 after a GAT update,
`including the old copy of the master index page 502 from FIG.
`6. The flushing threshold criteria may trigger a partial update
`of the GAT508 with the corresponding changed entries from
`the GAT Delta 506. For example, as shown in the GAT block
`500 of FIG. 7, if (1) a request to set an entry in the GAT508
`mapping a logical group to a physical metablock is received at
`step 802; (2) the entry does not already exist in the GAT Delta
`506 at step 804; and (3) the flushing threshold criteria is met
`at step 806, then a partial update of the GAT 508 occurs at
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`US 8,250,333 B2
`
`10
`
`