U.S. patent application number 09/841522 was filed with the patent office on 2001-10-11 for identification and verification of a sector within a block of mass storage flash memory.
Invention is credited to Estakhri, Petro, Iman, Berhanu.
Application Number | 20010029564 09/841522 |
Document ID | / |
Family ID | 27125428 |
Filed Date | 2001-10-11 |
United States Patent
Application |
20010029564 |
Kind Code |
A1 |
Estakhri, Petro ; et
al. |
October 11, 2001 |
Identification and verification of a sector within a block of mass
storage flash memory
Abstract
A method and apparatus is disclosed for identifying a block
being stored within flash memory devices using a cluster address
for each block, the block being selectively erasable and having one
or more sectors, the cluster address being stored in one of the
sectors of the block. In an alternative embodiment, the cluster
address is stored in at least two different sectors within the same
block for ensuring that the information last written to the block
is valid. Further disclosed is a novel way to use a defect flag for
each block stored within the flash memory device for efficiently
identifying non-defective blocks upon system power-up.
Inventors: |
Estakhri, Petro;
(Pleasanton, CA) ; Iman, Berhanu; (Sunnyvale,
CA) |
Correspondence
Address: |
Law Offices of Imam & Associates
111 North Market St., Suite 1010
San Jose
CA
95113
US
|
Family ID: |
27125428 |
Appl. No.: |
09/841522 |
Filed: |
April 23, 2001 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
09841522 |
Apr 23, 2001 |
|
|
|
09520903 |
Mar 7, 2000 |
|
|
|
6223308 |
|
|
|
|
09520903 |
Mar 7, 2000 |
|
|
|
09156951 |
Sep 18, 1998 |
|
|
|
6128695 |
|
|
|
|
09156951 |
Sep 18, 1998 |
|
|
|
08858847 |
May 19, 1997 |
|
|
|
5838614 |
|
|
|
|
08858847 |
May 19, 1997 |
|
|
|
08831266 |
Mar 31, 1997 |
|
|
|
5907856 |
|
|
|
|
08831266 |
Mar 31, 1997 |
|
|
|
08509706 |
Jul 31, 1995 |
|
|
|
5845313 |
|
|
|
|
Current U.S.
Class: |
711/103 ;
711/156; 711/203 |
Current CPC
Class: |
G06F 3/0613 20130101;
G06F 3/064 20130101; G06F 3/0652 20130101; G06F 12/023 20130101;
G11C 16/08 20130101; G06F 11/1068 20130101; G11C 29/82 20130101;
G06F 3/061 20130101; G06F 3/0619 20130101; G06F 2212/7201 20130101;
G11C 29/765 20130101; G06F 12/0246 20130101; G06F 3/0679 20130101;
G11C 16/102 20130101 |
Class at
Publication: |
711/103 ;
711/203; 711/156 |
International
Class: |
G06F 012/00 |
Claims
What is claimed is:
1. A storage device comprising: nonvolatile memory coupled to a
host for storing information assigned by the host, said nonvolatile
memory being organized into blocks, at least a portion of each said
block being identified by a physical block address (PBA); a memory
device coupled to said nonvolatile memory for identifying a
particular block having one or more rows within said nonvolatile
memory in which information is, or is to be stored; and means
associated with said memory device for accessing a block identified
a PBA within said nonvolatile memory and having a block address
associated therewith, said associated means accesses at least a
portion of a particular PBA to read data therefrom, or to write
data into the accessed block, said associated means writes the
block address into two different locations within said accessed
block, and upon accessing of any block within the nonvolatile
memory having data stored therein, said associated means compares
the block address written into the two different locations within
the block and if they match, said associated means determines that
the block is valid for accessing thereof.
2. A storage device as recited in claim 1 wherein if the block
addresses do not match, said associated means being operative to
erase the block prior to re-use thereof.
3. A storage device as recited in claim 1 wherein each block
includes a defect flag that if set, prevents the block from being
re-used.
Description
CROSS REFERENCE TO RELATED APPLICATION
[0001] This application is a continuation of our prior co-pending
application Ser. No. 09/520,903, filed on Mar. 7, 2000, entitled
"IDENTIFICATION AND VERIFICATION OF A SECTOR WITHIN A BLOCK OF MASS
STORAGE FLASH MEMORY", which is a continuation of our prior
co-pending application Ser. No. 09/156,951, filed on Sep. 18, 1998,
entitled "IDENTIFICATION AND VERIFICATION OF A SECTOR WITHIN A
BLOCK OF MASS STORAGE FLASH MEMORY" which is a continuation of
prior U.S. Pat. No. 5,838,614, issued on Nov. 17, 1998, entitled
"IDENTIFICATION AND VERIFICATION OF A SECTOR WITHIN A BLOCK OF MASS
STORAGE FLASH MEMORY" which is a continuation-in-part of prior U.S.
Pat. No. 5,907,856, issued on May 25, 1999, entitled "MOVING
SECTORS WITHIN A BLOCK OF INFORMATION IN A FLASH MEMORY MASS
STORAGE ARCHITECTURE", which is a Continuation-in-Part of prior
U.S. Pat. No. 5,845,313, issued on Dec. 1, 1998, and entitled
"DIRECT LOGICAL BLOCK ADDRESSING FLASH MEMORY MASS STORAGE
ARCHITECTURE."
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] This invention relates to the field of mass storage for
computers. More particularly, this invention relates to an
architecture for replacing a hard disk with a semiconductor
nonvolatile memory and in particular flash memory.
[0004] 2. Description of the Prior Art
[0005] Computers conventionally use rotating magnetic media for
mass storage of documents, data, programs and information. Though
widely used and commonly accepted, such hard disk drives suffer
from a variety of deficiencies. Because of the rotation of the
disk, there is an inherent latency in extracting information from a
hard disk drive.
[0006] Other problems are especially dramatic in portable
computers. In particular, hard disks are unable to withstand many
of the kinds of physical shock that a portable computer will likely
sustain. Further, the motor for rotating the disk consumes
significant amounts of power decreasing the battery life for
portable computers.
[0007] Solid state memory is an ideal choice for replacing a hard
disk drive for mass storage because it can resolve the problems
cited above. Potential solutions have been proposed for replacing a
bard disk drive with a semiconductor memory. For such a system to
be truly useful, the memory must be nonvolatile and alterable. The
inventors have determined that FLASH memory is preferred for such a
replacement.
[0008] FLASH memory is a transistor memory cell which is
programmable through hot electron, source injection, or tunneling,
and erasable through Fowler-Nordheim tunneling. The programming and
erasing of such a memory cell requires current to pass through the
dielectric surrounding floating gate electrode. Because of this,
such types of memory have a finite number of erase-write cycles.
Eventually, the dielectric deteriorates. Manufacturers of FLASH
cell devices specify the limit for the number of erase-write cycles
between 100,000 and 1,000,000.
[0009] One requirement for a semiconductor mass storage device to
be successful is that its use in lieu of a rotating media hard disk
mass storage device be transparent to the designer and the user of
a system using such a device. In other words, the designer or user
of a computer incorporating such a semiconductor mass storage
device could simply remove the hard disk and replace it with a
semiconductor mass storage device. All presently available
commercial software should operate on a system employing such a
semiconductor mass storage device without the necessity of any
modification.
[0010] SanDisk proposed an architecture for a semiconductor mass
storage using FLASH memory at the Silicon Valley PC Design
Conference on Jul. 9, 1991. That mass storage system included
read-write block sizes of 512 Bytes to conform with commercial hard
disk sector sizes. Earlier designs incorporated erase-before-write
architectures. In this process, in order to update a file on the
media, if the physical location on the media was previously
programmed, it has to be erased before the new data can be
reprogrammed.
[0011] This process would have a major deterioration on overall
system throughput. When a host writes a new data file to the
storage media, it provides a logical block address to the
peripheral storage device associated with this data file. The
storage device then translates this given logical block address to
an actual physical block address on the media and performs the
write operation. In magnetic hard disk drives, the new data can be
written over the previous old data with no modification to the
media. Therefore, once the physical block address is calculated
from the given logical block address by the controller, it will
simply write the data file into that location. In solid state
storage, if the location associated with the calculated physical
block address was previously programmed, before this block can be
reprogrammed with the new data, it has to be erased. In one
previous art, in erase-before-write architecture where the
correlation between logical block address given by the host is one
to one mapping with physical block address on the media. This
method has many deficiencies. First, it introduces a delay in
performance due to the erase operation before reprogramming the
altered information. In solid state flash, erase is a very slow
process.
[0012] Secondly, hard disk users typically store two types of
information, one is rarely modified and another which is frequently
changed. For example, a commercial spread sheet or word processing
software program stored on a user's system are rarely, if ever,
changed. However, the spread sheet data files or word processing
documents are frequently changed. Thus, different sectors of a hard
disk typically have dramatically different usage in terms of the
number of times the information stored thereon is changed. While
this disparity has no impact on a hard disk because of its
insensitivity to data changes, in a FLASH memory device, this
variance can cause sections of the mass storage to wear out and be
unusable significantly sooner than other sections of the mass
storage.
[0013] In another architecture, the inventors previously proposed a
solution to store a table correlating the logical block address to
the physical block address. The inventions relating to that
solution are disclosed in U.S. patent application Ser. No.
08/038,668 filed on Mar. 26, 1993 and U.S. patent application Ser.
No. 08/037,893 also filed on Mar. 26, 1993. Those applications are
incorporated herein by reference.
[0014] The inventors' previous solution discloses two primary
algorithms and an associated hardware architecture for a
semiconductor mass storage device. It will be understood that "data
file" in this patent document refers to any computer file including
commercial software, a user program, word processing software
document, spread sheet file and the like. The first algorithm in
the previous solution provides means for avoiding an erase
operation when writing a modified data file back onto the mass
storage device. Instead, no erase is performed and the modified
data file is written onto an empty portion of the mass storage.
[0015] The semiconductor mass storage architecture has blocks sized
to conform with commercial hard disk sector sizes. The blocks are
individually erasable. In one embodiment, the semiconductor mass
storage can be substituted for a rotating hard disk with no impact
to the user, so that such a substitution will be transparent. Means
are provided for avoiding the erase-before-write cycle each time
information stored in the mass storage is changed.
[0016] According to the first algorithm, erase cycles are avoided
by programming an altered data file into an empty block. This would
ordinarily not be possible when using conventional mass storage
because the central processor and commercial software available in
conventional computer systems are not configured to track
continually changing physical locations of data files. The previous
solution includes a programmable map to maintain a correlation
between the logical address and the physical address of the updated
information files.
[0017] All the flags, and the table correlating the logical block
address to the physical block address are maintained within an
array of CAM cells. The use of the CAM cells provides very rapid
determination of the physical address desired within the mass
storage, generally within one or two clock cycles. Unfortunately,
as is well known, CAM cells require multiple transistors, typically
six. Accordingly, an integrated circuit built for a particular size
memory using CAM storage for the tables and flags will need to be
significantly larger than a circuit using other means for just
storing the memory.
[0018] The inventors proposed another solution to this problem
which is disclosed in U.S. patent application Ser. No. 08/131,495
filed on Oct. 4, 1993. That application is incorporated herein by
reference.
[0019] This additional previous solution invented by these same
inventors is also for a nonvolatile memory storage device. The
device is also configured to avoid having to perform an
erase-before-write each time a data file is changed by keeping a
correlation between logical block address and physical block
address in a volatile space management RAM. Further, this invention
avoids the overhead associated with CAM cell approaches which
require additional circuitry.
[0020] Like the solutions disclosed above by these same inventors,
the device includes circuitry for performing the two primary
algorithms and an associated hardware architecture for a
semiconductor mass storage device. In addition, the CAM cell is
avoided in this previous solution by using RAM cells.
[0021] Reading is performed in this previous solutions by providing
the logical block address to the memory storage. The system
sequentially compares the stored logical block addresses until it
finds a match. That data file is then coupled to the digital
system. Accordingly, the performance offered by this solution
suffers because potentially all of the memory locations must be
searched and compared to the desired logical block address before
the physical location of the desired information can be
determined.
[0022] Additionally, a solution was disclosed in a U.S. patent
application entitled "MOVING SECTORS WITHIN A BLOCK OF INFORMATION
IN A FLASH MEMORY MASS STORAGE ARCHITECTURE", filed on Mar. 31,
1997 by Petro Estakhri, Berhanu Iman and Ali Ganjuei, to which this
application is a continuation-in-part. In the foregoing parent
application, the disclosure of which is herein incorporated by
reference, a method and apparatus was presented for efficiently
moving sectors within a block from a first area within the
nonvolatile memory to an unused area within the nonvolatile memory
and marking the first area as "used".
[0023] What is however, needed in all of the above referenced
solutions is to increase the efficiency of the system by quickly
detecting defective blocks within the flash memory and to introduce
various ways of identifying each block within the flash memory
devices without foregoing memory capacity and performance.
SUMMARY OF THE INVENTION
[0024] The present invention is for a nonvolatile memory storage
device. The device is configured to avoid having to perform an
erase-before-write each time a data file is changed. Further, to
avoid the overhead associated with CAM cells, this approach
utilizes a RAM array. The host system maintains organization of the
mass storage data by using a logical block address. The RAM array
is arranged to be addressable by the same address as the logical
block addresses (LBA) of the host. Each such addressable location
in the RAM includes a field which holds the physical address of the
data in the nonvolatile mass storage expected by the host. This
physical block address (PBA) information must be shadowed in the
nonvolatile memory to ensure that the device will still function
after resuming operation after a power down because RAMs are
volatile memory devices. In addition, status flags are also stored
for each physical location. The status flags can be stored in
either the nonvolatile media or in both the RAM and in the
nonvolatile media.
[0025] The device includes circuitry for performing two primary
algorithms and an associated hardware architecture for a
semiconductor mass storage device. The first algorithm provides a
means for mapping of host logical block address to physical block
address with much improved performance and minimal hardware
assists. In addition, the second algorithm provides means for
avoiding an erase-before-write cycle when writing a modified data
file back onto the mass storage device. Instead, no erase is
performed and the modified data file is written onto an empty
portion of the mass storage.
[0026] Reading is performed in the present invention by providing
the logical block address to the memory storage. The RAM array is
arranged so that the logical block address selects one RAM
location. That location contains the physical block address of the
data requested by the host or other external system. That data file
is then read out to the host.
[0027] According to the second algorithm, erase cycles are avoided
by programming an altered data file into an altered data mass
storage block rather than itself after an erase cycle of the block
as done on previous arts.
[0028] In an alternative embodiment of the present invention, a
method and apparatus is presented for efficiently moving sectors
within a block from a first area within the nonvolatile memory to
an unused area within the nonvolatile memory and marking the first
area as "used".
[0029] Briefly, A preferred embodiment of the present invention
includes a method and apparatus for storing mapping information for
mapping a logical block address identifying a block being accessed
by a host to a physical block address, identifying a free area of
nonvolatile memory, the block being selectively erasable and having
one or more sectors that may be individually moved. The mapping
information including a virtual physical block address for
identifying an "original" location, within the nonvolatile memory,
wherein a block is stored and a moved virtual physical block
address for identifying a "moved" location, within the nonvolatile
memory, wherein one or more sectors of the stored block are moved.
The mapping information further including status information for
use of the "original" physical block address and the "moved"
physical block address and for providing information regarding
"moved" sectors within the block being accessed.
IN THE DRAWINGS
[0030] FIG. 1 shows a schematic block diagram of an architecture
for a semiconductor mass storage according to the present
invention.
[0031] FIG. 2 shows an alternative embodiment to the physical block
address 102 of the RAM storage of FIG. 1.
[0032] FIG. 3 shows a block diagram of a system incorporating the
mass storage device of the present invention.
[0033] FIGS. 4 through 8 show the status of several of the flags
and information for achieving the advantages of the present
invention.
[0034] FIG. 9 shows a flow chart block diagram of the first
algorithm according to the present invention.
[0035] FIG. 10 shows a high-level block diagram of a digital
system, such as a digital camera, including a preferred embodiment
of the present invention.
[0036] FIGS. 11-19 illustrate several examples of the state of a
mapping table that may be stored in the digital system of FIG. 10
including LBA-PBA mapping information.
[0037] FIG. 20 depicts an example of a nonvolatile memory device
employed in the preferred embodiment of FIG. 10.
[0038] FIG. 21 shows a high-level flow chart of the general steps
employed in writing a block of information to the nonvolatile
devices of FIG. 10.
[0039] FIG. 22 shows an example of the contents of flash memory
devices in an alternative embodiment of the present invention using
a novel defect flag and LBA address means.
[0040] FIG. 23 is an example of the contents of flash memory
devices wherein another alternative embodiment of the present
invention stores the LBA address for each block in two different
sector locations within the block.
[0041] FIG. 24 shows an example of yet another alternative
embodiment of the present invention wherein the contents of the
flash memory devices and the SPM RAM block are depicted to
illustrate the correlation between the LBA and PBA addressing as
employed by the system of FIG. 10.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
[0042] FIG. 1 shows an architecture for implementation of a solid
state storage media according to the present invention. The storage
media is for use with a host or other external digital system. The
mass storage is partitioned into two portions, a volatile RAM array
100 and a nonvolatile array 104. According to the preferred
embodiment, all of the nonvolatile memory storage is FLASH. The
FLASH may be replaced by EEPROM. The RAM can be of any convenient
type.
[0043] The memory storage 104 is arranged into N blocks of data
from zero through N-1. Each of the blocks of data is M Bytes long.
In the preferred embodiment, each data block is 512 Bytes long to
correspond with a sector length in a commercially available hard
disk drive plus the extra numbers of bytes to store the flags and
logical block address (LBA) information and the associated ECC. The
memory 104 can contain as much memory storage as a user desires. An
example of a mass storage device might include 100 M Byte of
addressable storage.
[0044] There are a plurality of RAM locations 102. Each RAM
location 102 is uniquely addressable by controller using an
appropriate one of the logical block addresses provided by the host
system or the actual physical address of the nonvolatile media. The
RAM location 102 contains the physical block address of the data
associated with the logical block address and the flags associated
with a physical block address on the nonvolatile media.
[0045] It is possible that the physical block address (PBA) can be
split into two fields as shown in FIG. 2. These fields can be used
for cluster addresses of a group of data blocks. The first such
field 290 is used to select a cluster address and the second such
field 292 can be used to select the start address of the logical
block address associated with this cluster.
[0046] A collection of information flags is also stored for each
nonvolatile memory location 106. These flags include an old/new
flag 110, a used/free flag 112, a defect flag 114, and a
single/sector flag 116. Additionally, there is also a data store
122.
[0047] When writing data to the mass storage device of the present
invention, a controller determines the first available physical
block for storing the data. The RAM location 102 corresponding to
the logical block address selected by the host is written with the
physical block address where the data is actually stored within the
nonvolatile memory array in 104 (FIG. 1).
[0048] Assume for example that a user is preparing a word
processing document and instructs the computer to save the
document. The document will be stored in the mass storage system.
The host system will assign it a logical block address. The mass
storage system of the present invention will select a physical
address of an unused block or blocks in the mass storage for
storing the document. The address of the physical block address
will be stored into the RAM location 102 corresponding to the
logical block address. As the data is programmed, the system of the
present invention also sets the used free flag 112 in 104 and 293
to indicate that this block location is used. One used/free flag
112 is provided for each entry of the nonvolatile array 104.
[0049] Later, assume the user retrieves the document, makes a
change and again instructs the computer to store the document. To
avoid an erase-before-write cycle, the system of the present
invention provides means for locating a block having its used/free
flag 112 in 100 unset (not programmed) which indicates that the
associated block is erased. The system then sets the used/free flag
for the new block 112 of 106 and 293 of 100 and then stores the
modified document in that new physical block location 106 in the
nonvolatile array 104. The address of the new physical block
location is also stored into the RAM location 102 corresponding the
logical block address, thereby writing over the previous physical
block location in 102. Next, the system sets the old/new flag 110
of the previous version of the document indicating that this is an
old unneeded version of the document in 110 of 104 and 293 of 109.
In this way, the system of the present invention avoids the
overhead of an erase cycle which is required in the
erase-before-write of conventional systems to store a modified
version of a previous document.
[0050] Because of RAM array 100 will lose its memory upon a power
down condition, the logical block address with the active physical
block address in the media is also stored as a shadow memory 108 in
the nonvolatile array 104. It will be understood the shadow
information will be stored into the appropriate RAM locations 102
by the controller. During power up sequence, the RAM locations in
100 are appropriately updated from every physical locations in 104,
by reading the information 106 of 104. The logical address 108 of
106 is used to address the RAM location of 100 to update the actual
physical block address associated with the given logical block
address. Also since 106 is the actual physical block address
associated with the new data 122, the flags 110, 112, 114, and 116
arc updated in 293 of 102 with the physical block address of 106 in
100. It will be apparent to one of ordinary skill in the art that
the flags can be stored in either the appropriate nonvolatile
memory location 106 or in both the nonvolatile memory location and
also in the RAM location 102 associated with the physical block
address.
[0051] During power up, in order to assign the most recent physical
block address assigned to a logical block address in the volatile
memory 100, the controller will first read the Flags 110, 112, 114,
and 116 portion of the nonvolatile memory 104 and updates the flags
portion 293 in the volatile memory 100. Then it reads the logical
block address 108 of every physical block address of the
nonvolatile media 104 and by tracking the flags of the given
physical block address in the volatile memory 100, and the read
logical block address of the physical block address in the
nonvolatile memory 104, it can update the most recent physical
block address assigned to the read logical block address in the
volatile memory 100.
[0052] FIG. 3 shows a block diagram of a system incorporating the
mass storage device of the present invention. An external digital
system 300 such as a host computer, personal computer and the like
is coupled to the mass storage device 302 of the present invention.
A logical block address is coupled via an address bus 306 to the
volatile RAM array 100 and to a controller circuit 304. Control
signals are also coupled to the controller 304 via a control bus
308. The volatile RAM array 1.00 is coupled for providing the
physical block address to the nonvolatile RAM array 400. The
controller 304 is coupled to control both the volatile RAM 100, the
nonvolatile array 104, and for the generation of all flags.
[0053] A simplified example, showing the operation of the write
operation according to the present invention is shown in FIGS. 4
through 8. Not all the information flags are shown to avoid
obscuring these features of the invention in excessive detail. The
data entries are shown using decimal numbers to further simplify
the understanding of the invention. It will be apparent to one of
ordinary skill in the art that in a preferred embodiment binary
counting will be used.
[0054] FIG. 4 shows an eleven entry mass storage device according
to the present invention. There is no valid nor usable data stored
in the mass storage device of FIG. 4. Accordingly, all the physical
block addresses are empty. The data stored in the nonvolatile mass
storage location `6` is filled and old. Additionally, location `9`
is defective and cannot be used.
[0055] The host directs the mass storage device of the example to
write data pursuant to the logical block address `3` and then to
`4`. The mass storage device will first write the data associated
with the logical block address `3`. The device determines which is
the first unused location in the nonvolatile memory. In this
example, the first empty location is location `0`. Accordingly,
FIG. 5 shows that for the logical block address `3`, the
corresponding physical block address `0` is stored and the used
flag is set in physical block address `0`. The next empty location
is location `1`. FIG. 6 shows that for the logical block address
`4`, the corresponding physical block address `1` is stored and the
used flag is set in physical block address `1`.
[0056] The host instructs that something is to be written to
logical block address `3` again. The next empty location is
determined to be location `2`. FIG. 7 shows that the old flag in
location `0` is set to indicate that this data is no longer usable,
the used flag is set in location `2` and the physical block address
in location `3` is changed to `2`.
[0057] Next, the host instructs that something is to be written to
logical block address `4` again. The next empty location is
determined to be location `3`. FIG. 8 shows that the old flag in
location `1` is set to indicate that this data is no longer usable,
the used flag is set in location `3` and the physical block address
in location `4` is changed to `3`. (Recall that there is generally
no relation between the physical block address and the data stored
in the same location.)
[0058] FIG. 9 shows algorithm 1 according to the present invention.
When the system of the present invention receives an instruction to
program data into the mass storage (step 200), then the system
attempts to locate a free block (step 202), i.e., a block having an
unset (not programmed) used/free flag. If successful, the system
sets the used/free flag for that block and programs the data into
that block (step 206).
[0059] If on the other hand, the system is unable to locate a block
having an unset used/free flag, the system erases the flags
(used/free and old/new) and data for all blocks having a set
old/new flag and unset defect flag (step 204) and then searches for
a block having an unset used/free flag (step 202). Such a block has
just been formed by step 204. The system then sets the used/flag
for that block and programs the data file into that block (step
206).
[0060] If the data is a modified version of a previously existing
file, the system must prevent the superseded version from being
accessed. The system determines whether the data file supersedes a
previous data file (step 208). If so, the system sets the old/new
flag associated with the superseded block (step 210). If on the
other hand, the data file to be stored is a newly created data
file, the step of setting the old/new flag (step 210) is skipped
because there is no superseded block. Lastly, the map for
correlating the logical address 308- to the physical addresses
updated (step 212).
[0061] By following the procedure outlined above, the overhead
associated with an erase cycle is avoided for each write to the
memory 104 except for periodically. This vastly improves the
performance of the overall computer system employing the
architecture of the present invention.
[0062] In the preferred embodiment of the present invention, the
programming of the flash memory follows the procedure commonly
understood by those of ordinary skill in the art. In other words,
the program impulses are appropriately applied to the bits to be
programmed and then compared to the data being programmed to ensure
that proper programming has occurred. In the event that a bit fails
to be erased or programmed properly, a defect flag 148 is set which
prevent that block from being used again.
[0063] FIG. 10 depicts a digital system 500 such as a digital
camera employing an alternative embodiment of the present
invention. Digital system 500 is illustrated to include a host 502,
which may be a personal computer (PC) or simply a processor of any
generic type commonly employed in digital systems, coupled to a
controller circuit 506 for storing in and retrieving information
from non-volatile memory unit 508. The controller circuit 506 may
be a semiconductor (otherwise referred to as an "integrated
circuit" or "chip") or optionally a combination of various
electronic components. In the preferred embodiment, the controller
circuit is depicted as a single chip device. The non-volatile
memory unit 508 is comprised of one or more memory devices, which
may each be flash or EEPROM types of memory. In the preferred
embodiment of FIG. 10, memory unit 508 includes a plurality of
flash memory devices, 510-512, each flash device includes
individually addressable locations for storing information. In the
preferred application of the embodiment in FIG. 10, such
information is organized in blocks with each block having one or
more sectors of data. In addition to the data, the information
being stored may further include status information regarding the
data blocks, such as flag fields, address information and the
like.
[0064] The host 502 is coupled through host information signals 504
to a controller circuit 506. The host information signals comprise
of address and data busses and control signals for communicating
command, data and other types of information to the controller
circuit 506, which in turn stores such information in memory unit
508 through flash address bus 512, flash data bus 514, flash
signals 516 and flash status signals 518 (508 and 512-516
collectively referred to as signals 538). The signals 538 may
provide command, data and status information between the controller
506 and the memory unit 508.
[0065] The controller 506 is shown to include high-level functional
blocks such as a host interface block 520, a buffer RAM block 522,
a flash controller block 532, a microprocessor block 524, a
microprocessor controller block 528, a microprocessor storage block
530, a microprocessor ROM block 534, an ECC logic block 540 and a
space manager block 544. The host interface block 520 receives host
information signals 504 for providing data and status information
from buffer RAM block 522 and microprocessor block 524 to the host
502 through host information signals 504. The host interface block
520 is coupled to the microprocessor block 524 through the
microprocessor information signals 526, which is comprised of an
address bus, a data bus and control signals.
[0066] The microprocessor block 524 is shown coupled to a
microprocessor controller block 528, a microprocessor storage block
530 and a microprocessor ROM block 534, and serves to direct
operations of the various functional blocks shown in FIG. 10 within
the controller 506 by executing program instructions stored in the
microprocessor storage block 530 and the microprocessor ROM block
534. Microprocessor 524 may, at times, execute program instructions
(or code) from microprocessor ROM block 534, which is a
non-volatile storage area. On the other hand, microprocessor
storage block 530 may be either volatile, i.e., read-and-write
memory (RAM), or non-volatile, i.e., EEPROM, type of memory
storage. The instructions executed by the microprocessor block 524,
collectively referred to as program code, are stored in the storage
block 530 at some time prior to the beginning of the operation of
the system of the present invention. Initially, and prior to the
execution of program code from the microprocessor storage location
530, the program code may be stored in the memory unit 508 and
later downloaded to the storage block 530 through the signals 538.
During this initialization, the microprocessor block 524 can
execute instructions from the ROM block 534.
[0067] Controller 506 further includes a flash controller block 532
coupled to the microprocessor block 524 through the microprocessor
information signals 526 for providing and receiving information
from and to the memory unit under the direction of the
microprocessor. Information such as data may be provided from flash
controller block 532 to the buffer RAM block 522 for storage (may
be only temporary storage) therein through the microprocessor
signals 526. Similarly, through the microprocessor signals 526,
data may be retrieved from the buffer RAM block 522 by the flash
controller block 532.
[0068] ECC logic block 540 is coupled to buffer RAM block 522
through signals 542 and further coupled to the microprocessor block
524 through microprocessor signals 526. ECC logic block 540
includes circuitry for generally performing error coding and
correction functions. It should be understood by those skilled in
the art that various ECC apparatus and algorithms are commercially
available and may be employed to perform the functions required of
ECC logic block 540. Briefly, these functions include appending
code that is for all intensive purposes uniquely generated from a
polynomial to the data being transmitted and when data is received,
using the same polynomial to generate another code from the
received data for detecting and potentially correcting a
predetermined number of errors that may have corrupted the data.
ECC logic block 540 performs error detection and/or correction
operations on data stored in the memory unit 508 or data received
from the host 502.
[0069] The space manager block 544 employs a preferred apparatus
and algorithm for finding the next unused (or free) storage block
within one of the flash memory devices for storing a block of
information, as will be further explained herein with reference to
other figures. As earlier discussed, the address of a block within
one of the flash memory devices is referred to as PBA, which is
determined by the space manager by performing a translation on an
LBA received from the host. A variety of apparatus and method may
be employed for accomplishing this translation. An example of such
a scheme is disclosed in U.S. Pat. No. 5,485,595, entitled "Flash
Memory Mass Storage Architecture Incorporating Wear Leveling
Technique Without Using CAM Cells", the specification of which is
herein incorporated by reference. Other LBA to PBA translation
methods and apparatus may be likewise employed without departing
from the scope and spirit of the present invention.
[0070] Space manager block 544 includes SPM RAM block 548 and SPM
control block 546, the latter two blocks being coupled together.
The SPM RAM block 548 stores the LBA-PBA mapping information
(otherwise herein referred to as translation table, mapping table,
mapping information, or table) under the control of SPM control
block 546. Alternatively, the SPM RAM block 548 may be located
outside of the controller, such as shown in FIG. 3 with respect to
RAM array 100.
[0071] In operation, the host 502 writes and reads information from
and to the memory unit 508 during for example, the performance of a
read or write operation through the controller 506. In so doing,
the host 502 provides an LBA to the controller 506 through the host
signals 504. The LBA is received by the host interface block 520.
Under the direction of the microprocessor block 524, the LBA is
ultimately provided to the space manager block 544 for translation
to a PBA and storage thereof, as will be discussed in further
detail later.
[0072] Under the direction of the microprocessor block 524, data
and other information are written into or read from a storage area,
identified by the PBA, within one of the flash memory devices
510-512 through the flash controller block 532. The information
stored within the flash memory devices may not be overwritten with
new information without first being erased, as earlier discussed.
On the other hand, erasure of a block of information (every time
prior to being written), is a very time and power consuming
measure. This is sometimes referred to as erase-before-write
operation. The preferred embodiment avoids such an operation by
continuously, yet efficiently, moving a sector (or multiple
sectors) of information, within a block, that is being re-written
from a PBA location within the flash memory to an unused PBA
location within the memory unit 508 thereby avoiding frequent
erasure operations. A block of information may be comprised of more
than one sector such as 16 or 32 sectors. A block of information is
further defined to be an individually-erasable unit of information.
In the past, prior art systems have moved a block stored within
flash memory devices that has been previously written into a free
(or unused) location within the flash memory devices. Such systems
however, moved an entire block even when only one sector of
information within that block was being re-written. In other words,
there is waste of both storage capacity within the flash memory as
well as waste of time in moving an entire block's contents when
less than the total number of sectors within the block are being
re-written. The preferred embodiments of the present invention, as
discussed herein, allow for "moves" of less than a block of
information thereby decreasing the number of move operations of
previously-written sectors, consequently, decreasing the number of
erase operations.
[0073] Referring back to FIG. 10, it is important to note that the
SPM RAM block 548 maintains a table that may be modified each time
a write operation occurs thereby maintaining the LBA-PBA mapping
information and other information regarding each block being stored
in memory unit 508. Additionally, this mapping information provides
the actual location of a sector (within a block) of information
within the flash memory devices. As will be further apparent, at
least a portion of the information in the mapping table stored in
the SPM RAM block 548 is "shadowed" (or copied) to memory unit 508
in order to avoid loss of the mapping information when power to the
system is interrupted or terminated. This is, in large part, due to
the use of volatile memory for maintaining the mapping information.
In this connection, when power to the system is restored, the
portion of the mapping information stored in the memory unit 508 is
transferred to the SPM RAM block 548.
[0074] It should be noted, that the SPM RAM block 548 may
alternatively be nonvolatile memory, such as in the form of flash
or EEPROM memory architecture. In this case, the mapping table will
be stored within nonvolatile memory thereby avoiding the need for
"shadowing" because during power interruptions, the mapping
information stored in nonvolatile memory will be clearly
maintained.
[0075] When one or more sectors are being moved from one area of
the flash memory to another area, the preferred embodiment of the
present invention first moves the sector(s) from the location where
they are stored in the flash memory devices, i.e., 510-512, to the
buffer RAM block 522 for temporary storage therein. The moved
sector(s) are then moved from the buffer RAM block 522 to a free
area within one of the flash memory devices. It is further useful
to note that the ECC code generated by the ECC logic block 540, as
discussed above, is also stored within the flash memory devices
510-512 along with the data, as is other information, such as the
LBA corresponding to the data and flag fields.
[0076] FIGS. 11-19 are presented to show examples of the state of a
table 700 in SPM RAM block 548 configured to store LBA-PBA mapping
information for identification and location of blocks (and sectors
within the blocks) within the memory unit 508. Table 700 in all of
these figures is shown to include an array of columns and rows with
the columns including virtual physical block address locations or
VPBA block address locations 702, move virtual physical address
locations or MVPBA block address locations 704, move flag locations
706, used/free flag locations 708, old/new flag locations 710,
defect flag locations 712 and sector move status locations 714.
[0077] The rows of table include PBA/LBA rows 716, 718 through 728
with each row having a row number that may be either an LBA or a
PBA depending upon the information that is being addressed within
the table 700. For example, row 716 is shown as being assigned row
number `00` and if PBA information in association with LBA `00` is
being retrieved from table 700, then LBA `00` may be addressed in
SPM RAM block 548 at row 716 to obtain the associated PBA located
in 730. On the other hand, if status information, such as flag
fields, 706-712, regarding a block is being accessed, the row
numbers of rows 716-728, such as `00`, `10`, `20`, `30`, `40`,
`50`, `N-1` represent PBA, as opposed to LBA, values. Furthermore,
each row of table 700 may be thought of as a block entry wherein
each entry contains information regarding a block. Furthermore,
each row of table 700 may be addressed by an LBA.
[0078] In the preferred embodiment, each block is shown to include
16 sectors. This is due to the capability of selectively erasing an
entire block of 16 sectors (which is why the block size is
sometimes referred to as an "erase block size". If an erase block
size is 16 sectors, such as shown in FIGS. 11-19, each block entry
(or row) includes information regarding 16 sectors. Row 716
therefore includes information regarding a block addressed by LBA
`00` through LBA `15` (or LBA `00` through LBA `0F` in Hex.
notation). The next row, row 718, includes information regarding
blocks addressed by LBA `16` (or `10` in Hex.) through LBA `31` (or
`1F` in Hex.) The same is true for PBAs of each block.
[0079] It should be noted however, other block sizes may be
similarly employed. For example, a block may include 32 sectors and
therefore an erase block size 32. In the latter situation, each
block entry or row, such as 716, 718, 720 . . . , would include
information regarding 32 sectors. where each block comprises of 16
sectors (other than 16-sector block sizes may be similarly
employed).
[0080] The VPBA block address locations 702 of table 700 stores
information generally representing a PBA value corresponding to a
particular LBA value. The MVPBA block address locations 704 store
information representing a PBA value identifying, within the memory
unit 508, the location of where a block (or sector portions
thereof) may have been moved. The move flag locations 706 store
values indicating whether the block being accessed has any sectors
that may have been moved to a location whose PBA is indicated by
the value in the MVPBA block address location 704 (the PBA value
within 704 being other than the value indicated in VPBA block
address 702 wherein the remaining block address information may be
located). The used/new flag location 708 stores information to
indicate whether the block being accessed is a free block, that is,
no data has been stored since the block was last erased. The
old/new flag location 710 stores information representing the
status of the block being accessed as to whether the block has been
used and re-used and therefore, old. The defect flag location 712
stores information regarding whether the block is defective. If a
block is declared defective, as indicated by the value in the
defect flag location 712 being set, the defective block can no
longer be used. Flags 708-712 are similar to the flags 110-114
shown and described with respect to FIG. 1.
[0081] Sector move status location 714 is comprised of 16 bits
(location 714 includes a bit for each sector within a block so for
different-sized blocks, different number of bits within location
714 are required) with each bit representing the status of a sector
within the block as to whether the sector has been moved to another
block within the memory unit 508. The moved block location within
the memory unit 508 would be identified by a PBA that is other than
the PBA value in VPBA block address location 702. Said differently,
the status of whether a sector within a block has been moved, as
indicated by each of the bits within 714, suggests which one of
either the VPBA block address locations 702 or the MBPBA block
address locations 704 maintain the most recent PBA location for
that sector.
[0082] Referring still to FIG. 11, an example of the status of the
table 700 stored in SPM RAM block 548 (in FIG. 10) is shown when,
by way of example, LBA `0` is being written. As previously noted,
in the figures presented herein, a block size of sixteen sectors
(number 0-15 in decimal notation or 0-10 in hexadecimal notation)
is used to illustrate examples only. Similarly, N blocks (therefore
N LBAs) are employed, numbered from 0-N-1. The block size and the
number of blocks are both design choices that may vary for
different applications and may depend upon the memory capacity of
each individual flash memory device (such as 510-512) being
employed. Furthermore, a preferred sector size of 512 bytes is used
in these examples whereas other sector sizes may be employed
without departing from the scope and spirit of the present
invention.
[0083] Assuming that the operation of writing to LBA `0` is
occurring after initialization or system power-up when all of the
blocks within the flash memory devices 510-512 (in FIG. 10) have
been erased and are thus free. The space manager block 548 is
likely to determine that the next free PBA location is `00`.
Therefore, `00` is written to 730 in VPBA block address 702 of row
716 wherein information regarding LBA `0` is maintained, as
indicated in table 700 by LBA row number `00`. Since no need exists
for moving any of the sectors within the LBA 0 block, the MVPBA
block address 704 for row 716, which is shown as location 732 may
include any value, such as an initialization value (in FIG. 11,
`XX` is shown to indicate a "don't care" state).
[0084] The value in 734 is at logic state `0` to show that LBA `0`
block does not contain any moved sectors. Location 736 within the
used flag 708 column of row 716 will be set to logic state `1`
indicating that the PBA `0` block is in use. The state of location
738, representing the old flag 710 for row 716, is set to `0` to
indicate that PBA `0` block is not "old" yet. Location 740
maintains logic state `0` indicating that the PBA `0` block is not
defective and all of the bits in move status location 714 are at
logic state `0` to indicate that none of the sectors within the LBA
`0` through LBA `15` block have been moved.
[0085] In FIG. 11, the status information for LBA `0` in row 716,
such as in move flag location 706, used flag location 708, old flag
location 710, defect flag location 712 and move status location 714
for all remaining rows, 716-728, of table 700 are at logic state
`0`. It is understood that upon power-up of the system and/or after
erasure of any of the blocks, the entries for the erased blocks,
which would be all blocks upon power-up, in table 700, are all set
to logic state `0`.
[0086] At this time, a discussion of the contents of one of the
flash memory devices within the memory unit 508, wherein the LBA
`0` block may be located is presented for the purpose of a better
understanding of the mapping information shown in table 700 of FIG.
11.
[0087] Turning now to FIG. 20, an example is illustrated of the
contents of the flash memory device 510 in accordance with the
state of table 700 (as shown in FIG. 11). LBA `0`, which within the
memory unit 508 is identified at PBA `0` by controller 506 (of FIG.
10) is the location wherein the host-identified block is written. A
PBA0 row 750 is shown in FIG. 22 to include data in sector data
location 752. An ECC code is further stored in ECC location 754 of
PBA0 row 750. This ECC code is generated by the ECC logic block 540
in association with the data being written, as previously
discussed. Flag field 756 in PBA0 row 750 contains the move, used,
old and defect flag information corresponding to the sector data of
the block being written. In this example, in flag field 756, the
"used" flag and no other flag is set, thus, flag field 756
maintains a logic state of `0100` indicating that PBA `0` is "used"
but not "moved", "old" or "defective".
[0088] PBA0 row 750 additionally includes storage location for
maintaining in LBA address location 758, the LBA number
corresponding to PBA `0`, which in this example, is `0`. While not
related to the example at hand, the remaining PBA locations of LBA
`0` are stored in the next 15 rows following row 750 in the flash
memory device 510.
[0089] It will be understood from the discussion of the examples
provided herein that the information within a PBA row of flash
memory device 510 is enough to identify the data and status
information relating thereto within the LBA `0` block including any
moves associated therewith, particularly due to the presence of the
"move" flag within each PBA row (750, 762, 764, . . . ) of the
flash memory. Nevertheless, alternatively, another field may be
added to the first PBA row of each LBA location within the flash,
replicating the status of the bits in the move status location 714
of the corresponding row in table 700. This field is optionally
stored in sector status location 760 shown in FIG. 20 to be
included in the first PBA row of each LBA block, such as row 750,
780 and so on. Although the information maintained in location 760
may be found by checking the status of the "move" flags within the
flag fields 756 of each PBA row, an apparent advantage of using
location 760 is that upon start-up (or power-on) of the system, the
contents of table 700 in SPM RAM block 548 may be updated more
rapidly due to fewer read operations (the reader is reminded that
table 700 is maintained in SPM RAM 548, which is volatile memory
whose contents are lost when the system is power-down and needs to
be updated upon power-up from non-volatile memory, i.e. memory unit
508).
[0090] That is, rather than reading every PBA row (altogether 16
rows in the preferred example) to update each LBA entry of the
table 700 upon power-up, only the first PBA row of each LBA must be
read from flash memory and stored in SPM RAM 548 thereby saving
time by avoiding needless read operations. On the other hand,
clearly more memory capacity is utilized to maintain 16 bits of
sector status information per LBA.
[0091] In the above example, wherein location 760 is used, the
value in sector status location 760 would be all `0`s (or `0000` in
hexadecimal notation).
[0092] In flash memory device 510, each of the rows 750, 762, 764,
768 . . . , is a PBA location with each row having a PBA row number
and for storing data and other information (data and other
information are as discussed above with respect to row 750) for a
sector within a block addressed by a particular LBA. Furthermore,
every sixteen sequential PBA rows represents one block of
information. That is, PBA rows 750, 762, 764 through 768, which are
intended to show 16 PBA rows correspond to LBA 0 (shown as row 716
in table 700 of FIG. 11) and each of the PBA rows maintains
information regarding a sector within the block. The next block of
information is for the block addressed by LBA `10` (in Hex.) whose
mapping information is included in row 718 of table 700, and which
is stored in locations starting from `10` (in hexadecimal notation,
or `16` in decimal notation) and ending at `1F` (in hexadecimal
notation, or `31`) in the flash memory device 510 and so on.
[0093] Continuing on with the above example, FIG. 12 shows an
example of the state of table 700 when LBA 0 is again being written
by the host. Since LBA 0 has already been written and is again
being written without first being erased, another free location
within the memory unit 508 (it may serve helpful to note here that
the blocks, including their sectors, are organized sequentially and
continuously through each of the flash memory devices of memory
unit 508 according to their PBAs such that for example, the next
flash memory device following device 510 picks up the PBA-addressed
blocks where flash memory device 510 left off, an example of this
is where flash memory device 510 includes PBAs of 0-FF (in Hex.)
and the next flash memory device, which may be 512, may then
include 100-1FF (in Hex.)) is located by space manager 544 for
storage of the new information. This free location is shown to be
PBA `10` (in Hexadecimal notation, or 16 in decimal notation). In
row 718, where the entries for LBA `10` will remain the same as
shown in FIG. 11 except the used flag in location 742 will be set
(in the preferred embodiment, a flag is set when it is at logic
state `1` although the opposite polarity may be used without
deviating from the present invention) to indicate that the PBA `10`
is now "in use".
[0094] The entries in row 716 are modified to show `10` in MVPBA
block address location 732, which provides the PBA address of the
moved portion for the LBA `00` block. The move flag in location 734
is set to logic state `1` to indicate that at least a portion (one
or more sectors) of the LBA `00` block have been moved to a PBA
location other than the PBA location indicated in location 730 of
table 700. Finally, the bits of the move status location 714 in row
716 are set to `1000000000000000` (in binary notation, or `8000` in
hexadecimal notation), reflecting the status of the moved sectors
within the block LBA `00`. That is, in this example, `8000`
indicates that the first sector, or sector `0`, within LBA `00`
block has been moved to a different PBA location.
[0095] Referring now to FIG. 20, the state of table 700 in FIG. 12
will affect the contents of the flash memory device 510 in that the
moved sector of the LBA `0` block will now be written to PBA `10`
in row 780. Row 780 will then include the data for the moved
sector, which is 512 bytes in size. With respect to the moved
sector information, row 780 further includes ECC code, a copy of
the values in flag locations 734-740 of table 700 (in FIG. 12), and
LBA `00` for indicating that the data in row 780 belongs to LBA
`00` and may further include the move status for each of the
individual sectors within the LBA `0` block.
[0096] While not specifically shown in the figure, the move flag
within location 756 of PBA row 750 is set to indicate that at least
a portion of the corresponding block has been moved. The value
stored in the move status location 714 of row 716 (in FIG. 12),
which is `8000` in Hex., is also stored within location 760 of the
row 750. As earlier noted, this indicates that only sector `0` of
PBA `0` was marked "moved" and the new block LBA `0` was written to
PBA `10` in flash memory. Without further detailed discussions of
FIG. 22, it should be appreciated that the examples to follow
likewise affect the contents of the flash memory device 510.
[0097] FIG. 13 shows the status of table 700 when yet another write
operation to LBA `00` is performed. The values (or entries) in row
716 remain the same as in FIG. 12 except that the value in location
732 is changed to `20` (in Hex. Notation) to indicate that the
moved portion of block LBA `00` is now located in PBA location `20`
(rather than `10` in FIG. 12). As in FIG. 12, the value in move
status location 714, `8000`, indicates that the first sector (with
PBA `00`) is the portion of the block that has been moved.
[0098] Row 718 is modified to show that the LBA `10` block is now
old and can no longer be used before it is erased. This is
indicated by the value in location 744 being set to logic state
`1`. The entries for LBA `20`, row 720, remain unchanged except
that location 746 is modified to be set to logic state `1` for
reflecting the state of the PBA `20` block as being in use. It is
understood that as in FIGS. 11 and 12, all remaining values in
table 700 of FIG. 13 that have not been discussed above and are not
shown as having a particular logic state in FIG. 13 are all
unchanged (the flags are all set to logic state `0`).
[0099] FIGS. 14-16 show yet another example of what the state of
table 700 may be after either power-up or erasure of the blocks
with the memory unit 508. In FIGS. 14 and 15, the same write
operations as those discussed with reference to FIGS. 11 and 12 are
performed. The state of table 700 in FIGS. 16 and 17 resembles that
of FIGS. 11 and 12, respectively (the latter two figures have been
re-drawn as FIGS. 14 and 15 for the sole convenience of the
reader). Briefly, FIG. 14 shows the state of table 700 after a
write to LBA `0` and FIG. 15 shows the state of table 700 after
another write to LBA `0`.
[0100] FIG. 16 picks up after FIG. 15 and shows the state of table
700 after the host writes to LBA `5`. As indicated in FIG. 16, LBA
`5` has been moved to PBA `10` where LBA `0` is also located. To
this end, MBPBA block address location 732 is set to `10` in row
716 and the move flag is set at location 734 in the same row.
Moreover, the state of move status location 714 in row 716 is set
to `8400` (in Hex.) indicating that LBA `0` and LBA `5` have been
moved, or that the first and fifth sectors within LBA `00` are
moved. Being that these two sectors are now located in the PBA `10`
location of the flash memory device 510, the move flag for each of
the these sectors are also set in the flash memory device 510. It
should be understood that LBA `5` was moved to PBA `10` because
remaining free sectors were available in that block. Namely, even
with LBA `0` of that block having been used, 15 other sectors of
the same block were available, from which the fifth sector is now
in use after the write to LBA `5`.
[0101] Continuing on with the example of FIG. 16, in FIG. 17, the
state of the table 700 is shown after the host writes yet another
time to LBA `0`. According to the table, yet another free PBA
location, `20`, is found where both the LBA `5` and LBA `0` are
moved. First, LBA `5` is moved to the location PBA `10` to PBA `20`
and then the new block of location LBA `0` is written to PBA `20`.
As earlier discussed, any time there is a move of a block (for
example, here the block of LBA `5` is moved) it is first moved from
the location within flash memory where it currently resides to a
temporary location within the controller 506, namely with the
buffer RAM block 522, and then it is transferred from there to the
new location within the flash memory devices.
[0102] The used flag in location 746 of row 720 is set to reflect
the use of the PBA `20` location in flash memory and the old flag
in location 744 is set to discard use of PBA `10` location until it
is erased. Again, in flash memory, the state of these flags as well
as the state of the move flag for both the LBA `0` and LBA `5`
sectors are replicated.
[0103] FIG. 18 picks up from the state of the table 700 shown in
FIG. 16 and shows yet another state of what the table 700 may be
after the host writes to LBA `5`. In this case, the block of LBA
`0` is first moved from location PBA `10` within the flash memory
device 510 wherein it is currently stored to location PBA `20` of
the flash memory. Thereafter, the new block being written to LBA
`5` by the host is written into location PBA `20` of the flash
memory. The flags in both table 700 and corresponding locations of
the flash memory device 510 are accordingly set to reflect these
updated locations.
[0104] FIG. 19 also picks up from the state of the table 700 shown
in FIG. 16 and shows the state of what the table 700 may be after
the host writes to LBA `7`. In this case, the new block is simply
written to location PBA `10` of the flash memory since that
location has not yet been used. Additionally, three of the bits of
the move status location 714 in row 716 are set to show that LBA
`0`, LBA `5` and LBA `7` have been moved to another PBA location
within the flash memory. Location 732 shows that the location in
which these three blocks are stored is PBA `10`.
[0105] As may be understood from the discussion presented thus far,
at some point in time, the number of sectors being moved within a
block makes for an inefficient operation. Thus, the need arises for
the user to set a threshold for the number of sectors within a
block that may be moved before the block is declared "old" (the old
flag is set) and the block is no longer used, until it is erased.
This threshold may be set at, for example, half of the number of
sectors within a block. This is demonstrated as follows: For a
block having 16 sectors, when 8 of the sectors are moved into
another block, the "original" block and the "moved" block (the
block in which the moved sectors reside) are combined into the same
PBA block. The combined PBA block may be stored in a new block
altogether or, alternatively, the "original" block may be combined
with and moved into the "moved" block. In the latter case, the
"original" block is then marked as "old" for erasure thereof. If
the combined PBA block is stored in a new block, both of the
"original" and the "moved" blocks are marked as "old".
[0106] FIG. 21 depicts a general flow chart outlining some of the
steps performed during a write operation. It is intended to show
the sequence of some of the events that take place during such an
operation and is not at all an inclusive presentation of the method
or apparatus used in the preferred embodiment of the present
invention. The steps as outlined in FIG. 21 are performed under the
direction of the microprocessor block 524 as it executes program
code (or firmware) during the operation of the system.
[0107] When the host writes to a block of LBA M, step 800, the
space manager block 544, in step 802, checks as to whether LBA M is
in use by checking the state of the corresponding used flag in
table 700 of the SPM RAM block 548. If not in use, in step 804, a
search is performed for the next free PBA block in memory unit 508.
If no free blocks are located, an "error" state is detected in 808.
But where a free PBA is located, in step 806, its used flag is
marked (or set) in table 700 as well as in flash memory. In step
810, the PBA of the free block is written into the VPBA block
address 702 location of the corresponding LBA row in table 700.
[0108] Going back to step 802, if the LBA M block is in use, search
for the next free PBA block is still conducted in step 812 and upon
the finding of no such free block, at 814, an "error" condition is
declared. Whereas, if a free PBA location is found, that PBA is
marked as used in table 700 and flash memory, at step 816. Next, in
step 818, the state of the block is indicated as having been moved
by setting the move flag as well as the setting the appropriate bit
in the move status location 714 of table 700. The new location of
where the block is moved is also indicated in table 700 in
accordance with the discussion above.
[0109] Finally, after steps 818 and 810, data and all corresponding
status information, ECC code and LBA are written into the PBA
location within the flash memory.
[0110] As earlier indicated, when a substantial portion of a block
has sectors that have been moved (in the preferred embodiment, this
is eight of the sixteen sectors), the block is declared "old" by
setting its corresponding "old" flag. Periodically, blocks with
their "old" flags set, are erased and may then be re-used (or
re-programmed, or re-written).
[0111] As can be appreciated, an advantage of the embodiments of
FIGS. 10-21 is that a block need not be erased each time after it
is accessed by the host because if for example, any portions (or
sectors) of the block are being re-written, rather than erasing the
block in the flash memory devices or moving the entire block to a
free area within the flash, only the portions that are being
re-written need be transferred elsewhere in flash, i.e. free
location identified by MVPA block address. In this connection, an
erase cycle, which is time consuming is avoided until later and
time is not wasted in reading an entire block and transferring the
same.
[0112] In an alternative embodiment, there may be less information
pertaining to each sector that is stored in the flash memory.
Likewise, the table 700 in the SPM RAM block 548 (in FIG. 10)
maintains less information such as explained below.
[0113] FIG. 22 shows an example of the information a flash memory
chip (such as flash memory device 510) may store. In FIG. 22 is
shown, N blocks, blocks 1000, 1002, . . . , 1004, with each block
having up to m sectors Using block 1000 as an example, each of the
sectors 1006, 1008, . . . , 1010 within block 1000 includes a data
field 1052 and an ECC field 1014. In the preferred embodiment, the
data field 1052 contains 512 bytes and the ECC field 1014 contains
4 bytes of information although other sizes of both data and ECC
may be employed without departing from the spirit of the present
invention.
[0114] The first sector in each block, for example sector 1006 of
block 1000, additionally includes a defect flag 1016 indicating
whether the block is defective or not. In the preferred embodiment,
if any of the sectors within a block are defective, the entire
block is marked defective as further explained below.
[0115] Each sector within a block may further contain a spare area
1018. In each block, there is included an LBA 1020 for identifying
the block within the flash memory unit, as will be explained in
greater detail shortly. It is however, important to note that the
LBA 1020 while shown in FIG. 22 to be located within the last
sector 1010 of block 1000, may be alternatively located within any
other sector of block 1000 and further within any area within the
sector in which it is located. For example, LBA 1020 may be located
within the spare area 1022 of sector 1008 or it may be located at
1024, which is right before the location where data is stored for
the sector 1008 of block 1000. The same sector organization that is
used for block 1000 is also used for the remaining blocks such as
blocks 1002 and 1004 within the flash memory devices.
[0116] Similarly, the location of the defect flag 1012 may be
anywhere within the block. For example, the defect flag 1012 may be
alternatively stored at 1024 or at 1022. But there is only one
defect flag stored per block. An aspect of the present invention
relates to the way in which the defect flag is employed.
[0117] During manufacturing of flash memory chips, defects within
the memory are commonly identified and marked by the chip
manufacturer. This is typically done by writing a predefined
pattern (byte-wide) in a predetermined location within a defective
block. That is, the manufacturer will erase the flash chip and a
value, such as all `1`s, will then be carried by each memory cell
within the flash chip that was successfully erased. If a sector or
a cell within a block is defective, the manufacturer will set a
manufacturing defect flag located somewhere within the defective
block (not shown in FIG. 22) to a predetermined value. This
manufacturing defect flag is not the same as defect flag 1012 and
its location is determined by the manufacturer and not by design
choice. These manufacturer-identified defective blocks are then
ignored (or not used) during system operation. That is, the space
manager block 544 (in FIG. 10) keeps track of these defective
blocks and knows not to use them when it searches for a free block
within the flash memory devices.
[0118] Other than defects detected during manufacturing of flash
chips, there may be additional defects developed during operation
of the chips due to wearing as discussed earlier. These defects are
sometimes referred to as "grown defects" by the industry at-large
and must be accounted for by a system in which using flash memory
devices are employed.
[0119] The case of "grown defects" will be explained in the context
of system operations. In a preferred embodiment of the present
invention, after erasure of the block 1000, the defect flag 1012
(in FIG. 22) is programmed to indicate whether the erased block is
defective or not. That is, if the cells in the erased block 1000
were successfully erased, the defect flag 1012 is set to a
predefined value such as a byte-wide value of `55` (in Hex.)
indicating that the block is not defective. If on the other hand,
the block is not successfully erased, a value other than `55`, such
as `00`, is written to the defect flag 1012 or alternatively, no
value is written and whatever the state of the defect flag was
after erasure (commonly, an erased state is all `1`s (or `FF` in
Hex.)) will be maintained. In the latter case however, the erased
state of the defect flag must be a value other than `55` in order
to distinguish successful erasures.
[0120] The use of the defect flag 1012 is especially noted after
power-up of the system. The space manager block 544 takes note of
which blocks are defective by quickly scanning the defect flags of
each block and identifying those blocks that are not defective for
later by searching for the value `55` in the defect flag 1012 of
each block. If a block is not defective after power-up and later
becomes defective, its defect flag is modified to so designate and
the next time the system is powered-up, the defective block is
noted by the space manager and any use of the block is avoided.
[0121] During a write operation, if a failure of one or more cells
of a block occurs, the block is first erased and then marked as
being defective by writing a value other than `55` (such as `00`)
to the defect flag 1012 of the defective block. Alternatively,
there is no value written to the defect flag 1012 of the defective
block.
[0122] One of the advantages of having a defect flag such as
described thus far is that during system power-up, the space
manager block 544 (in FIG. 10) is able to quickly find blocks that
are not defective.
[0123] In an alternative embodiment, as shown in FIG. 23, the LBA
for each block within the flash memory devices is stored within two
different sectors of the same block. For example, after block 1000
has been erased and during the first write operation of block 1000
following its erasure, the space manager block 544 determines that
block 1000 or a portion thereof is free to be written into. The LBA
associated with block 1000 is then programmed into two locations
1030 and 1020 (within sectors 1006 and 1010, respectively) of block
1000.
[0124] Thereafter, each time power to the system is temporarily
interrupted, or upon power-up, the two LBAs in locations 1030 and
1020 are compared to each other and if they match and the defect
flag 1012 indicates that the block is not defective, the block
continues to be used for information storage under the direction of
the space manager block 544. If however, the defect flag 1012
indicates that the block is not defective but the two LBAs do not
match, then the block 1000 will still continue to be used but it
will first be erased prior to further reuse. The latter case may
arise, for example, when the system was writing to block 1000 and
the first several sectors starting with sector 0 were written but
then there was a power interruption and information was therefore
not written to the remaining sectors of block 1000. In this case,
the LBA values in locations 1030 and 1020 will likely not match
because sector m-1, or sector 1010, of block 1000 was not written
into prior to the power interruption. Upon power restoration, since
not completely written, the block 1000 should not be used in its
state as it was prior to the power interruption. Accordingly, upon
detecting a mismatch between the two LBA values in locations 1030
and 1020, the controller 506 (in FIG. 10) will erase the block 1000
prior to its re-use.
[0125] Where only one LBA is used per block, as shown in FIG. 22,
obviously, the comparison of two LBAs in each block is not
performed yet detection of blocks that are defective is maintained
through the use of the defect flag as explained earlier.
Additionally, the LBA may be written last, that is, after all of
the sectors of the block are written. Writing the LBA last is an
added measure of successfully having written to the block.
[0126] To illustrate another aspect of the present invention, FIG.
24 shows, on the left-hand side, an example of the contents of
several blocks of the flash memory device (this is the same
information that is illustrated in FIG. 22) and, on the right-hand
side, an example of the contents of the table 700, which
corresponds to the block information that is on the left-hand side
of the figure. The purpose of showing the block contents is only to
serve as convenience to the reader in understanding the correlation
between the LBA-PBA addressing of the blocks with respect to the
table 700. The example shown in FIG. 24 is to illustrate the
mapping of this alternative embodiment between the addresses of the
blocks as they are stored in flash memory and the addresses of the
same blocks as that information is stored in the SPM RAM block
548.
[0127] In the table 700, there is shown to be stored in a column, a
virtual PBA field 1036 and a flag field 1038 including a defect
flag 1012 for each row. Rows 1040, 1042, 1044, . . . , 1050 each
correspond to an LBA and are addressed by `0`, `1`, `2`, . . . `Z`,
respectively. The virtual PBA field 1036 for each row actually
contains an LBA address pointer that points to a block within the
flash memory having 16 sectors stored therein. In this respect, the
virtual PBA serves as the PBA for 16 sectors (the number of sectors
may be other than sixteen in alternative embodiments).
[0128] If the host sends a command to read for example, LBA 05, the
controller 506, in FIG. 10, first masks the four least significant
bits (LSBs) of the value `05` to obtain the value `00`. Then using
`00` as the row address for table 700, the row 1040 in FIG. 24 is
addressed. Note that the reason the four LSBs of the LBA value sent
by the host is masked is because there are 16 sectors being
represented by each row of table 700. 16 sectors is two to the
power of 4 in binary terms which translates to 4 bits. If for
example each row represented 32 sectors, then the 5 LSBs of the LBA
sent by the host would be masked.
[0129] Once the row 1040 in the SPM RAM block 548 has been
addressed, the value in the virtual PBA field 1036 for the row 1040
is retrieved, which in this case is `00`. `00` is then used as the
pointer to the blocks stored within the flash memory devices. That
is, in this example, block 0, or the block 1000 (on the left-hand
side of FIG. 24) is addressed. But to get to the sector that was
intended to be read, the 4 bits that were initially masked from the
LBA value sent by the host, are used to specifically address sector
`5` (not shown), which is the sixth sector, within the block 1040.
The data stored in the data field 1052 of sector 5 (not shown) is
then retrieved or read. Next, the ECC field 1014 of that same
sector is read.
[0130] The LBA location 1020 should have the value `00`, which is
the same address value that is stored in the virtual PBA field 1036
of the row 1040 of table 700. This is intended for identifying the
block 1000 in the flash memory devices as block `00` during the
power-up routine, to update the space manager. Although, the LBA
location 1020 does not represent the LBA value sent by the host as
noted above. To avoid such confusion, the value in location 1020
may be referred to as a `cluster` address rather than an LBA
address.
[0131] To illustrate another example in the context of a write
operation, if the host commands the controller 506 to write to LBA
`17` (in decimal notation), the controller masks the LSBs of the
hexadecimal notation of `17` (which is `11`). The masked version of
`17` is `1` in hexadecimal notation. Thus, the row 1042 is
addressed in table 700 and a the address of a free block, as found
by the space manager block 544, is placed in the virtual PBA field
1036 of the row 1042. In this case, block 2 is found as the next
free block. Within the flash memory devices, block 2, which is the
block 1002, is addressed. The sector where information is actually
written into is sector 1 of the block 1002 (the second sector in
that block) because the 4 LSBs of the LBA value `11` sent by host.
The data and ECC are then written into the sector 1 of block
1002.
[0132] If the preferred embodiment example of FIG. 24 is used with
the embodiment discussed with reference to FIG. 23, the LBA (or
cluster) address is written in two places within each of the
blocks.
[0133] Although the present invention has been described in terms
of specific embodiments, it is anticipated that alterations and
modifications thereof will no doubt become apparent to those
skilled in the art. It is therefore intended that the following
claims be interpreted as covering all such alterations and
modification as fall within the true spirit and scope of the
invention.
* * * * *