Hardware White Paper

Designing Hardware for Microsoft® Operating Systems

Microsoft Extensible Firmware Initiative

FAT32 File System Specification
FAT: General Overview of On-Disk Format

Version 1.03, December 6, 2000

Microsoft Corporation

The FAT (File Allocation Table) file system has its origins in the late 1970s and early1980s
and was the file system supported by the Microsoft® MS-DOS® operating system. It was
originally developed as a simple file system suitable for floppy disk drives less than 500K in
size. Over time it has been enhanced to support larger and larger media. Currently there are
three FAT file system types: FAT12, FAT16 and FAT32. The basic difference in these FAT
sub types, and the reason for the names, is the size, in bits, of the entries in the actual FAT
structure on the disk. There are 12 bits in a FAT12 FAT entry, 16 bits in a FAT16 FAT entry
and 32 bits in a FAT32 FAT entry.

Contents
Notational Conventions in this Document ..............................................................................................7
General Comments (Applicable to FAT File System All Types)............................................................7
Boot Sector and BPB...............................................................................................................................7
FAT Data Structure ...............................................................................................................................13
FAT Type Determination ......................................................................................................................14
FAT Volume Initialization ....................................................................................................................19
FAT32 FSInfo Sector Structure and Backup Boot Sector.....................................................................21
FAT Directory Structure .......................................................................................................................22
FAT Long Directory Entries .................................................................................................................25
Name Limits and Character Sets ...........................................................................................................29
Name Matching In Short & Long Names..............................................................................................30
Naming Conventions and Long Names.................................................................................................30
Effect of Long Directory Entries on Down Level Versions of FAT .....................................................32
Validating The Contents of a Directory ................................................................................................32
Other Notes Relating to FAT Directories..............................................................................................33

Microsoft, MS_DOS, Windows, and Windows NT are trademarks or registered trademarks of Microsoft Corporation in the United States
and/or other countries. Other product and company names mentioned herein may be the trademarks of their respective owners.
© 2000 Microsoft Corporation. All rights reserved.
 FAT: General Overview of On-Disk Format—Page 2

Microsoft Extensible Firmware Initiative FAT32 File System Specification

IMPORTANT-READ CAREFULLY: This Microsoft Agreement (“Agreement”) is a legal agreement

between you (either an individual or a single entity) and Microsoft Corporation (“Microsoft”) for the
version of the Microsoft specification identified above which you are about to download
(“Specification”). BY DOWNLOADING, COPYING OR OTHERWISE USING THE
SPECIFICATION, YOU AGREE TO BE BOUND BY THE TERMS OF THIS AGREEMENT. IF
YOU DO NOT AGREE TO THE TERMS OF THIS AGREEMENT, DO NOT DOWNLOAD, COPY,
OR USE THE SPECIFICATION.

The Specification is owned by Microsoft or its suppliers and is protected by copyright laws and
international copyright treaties, as well as other intellectual property laws and treaties.

1. LIMITED LICENSE AND COVENANT NOT TO SUE.

(a) Provided that you comply with all terms and conditions of this Agreement and subject to
the limitations in Sections 1(c) - (f) below, Microsoft grants to you the following non-exclusive,
worldwide, royalty-free, non-transferable, non-sublicenseable license under any copyrights owned
or licensable by Microsoft without payment of consideration to unaffiliated third parties, to
reproduce the Specification solely for the purposes of creating portions of products which comply
with the Specification in unmodified form.

(b) Provided that you comply with all terms and conditions of this Agreement and subject to
the limitations in Sections 1(c) - (f) below, Microsoft grants to you the following non-exclusive,
worldwide, royalty-free, non-transferable, non-sublicenseable, reciprocal limited covenant not to
sue under its Necessary Claims solely to make, have made, use, import, and directly and
indirectly, offer to sell, sell and otherwise distribute and dispose of portions of products which
comply with the Specification in unmodified form.
For purposes of sections (a) and (b) above, the Specification is “unmodified” if there are no
changes, additions or extensions to the Specification, and “Necessary Claims” means claims of a
patent or patent application which are (1) owned or licenseable by Microsoft without payment of
consideration to an unaffiliated third party; and (2) have an effective filing date on or before
December 31, 2010, that must be infringed in order to make a portion(s) of a product that
complies with the Specification. Necessary Claims does not include claims relating to
semiconductor manufacturing technology or microprocessor circuits or claims not required to be
infringed in complying with the Specification (even if in the same patent as Necessary Claims).

(c) The foregoing covenant not to sue shall not extend to any part or function of a product
which (i) is not required to comply with the Specification in unmodified form, or (ii) to which there
was a commercially reasonable alternative to infringing a Necessary Claim.
(d) Each of the license and the covenant not to sue described above shall be unavailable to
you and shall terminate immediately if you or any of your Affiliates (collectively “Covenantee
Party”) “Initiates” any action for patent infringement against: (x) Microsoft or any of its Affiliates
(collectively “Granting Party”), (y) any customers or distributors of the Granting Party, or other
recipients of a covenant not to sue with respect to the Specification from the Granting Party
(“Covenantees”); or (z) any customers or distributors of Covenantees (all parties identified in (y)
and (z) collectively referred to as “Customers”), which action is based on a conformant
implementation of the Specification. As used herein, “Affiliate” means any entity which directly or
indirectly controls, is controlled by, or is under common control with a party; and control shall
mean the power, whether direct or indirect, to direct or cause the direction of the management or
policies of any entity whether through the ownership of voting securities, by contract or otherwise.
“Initiates” means that a Covenantee Party is the first (as between the Granting Party and the
Covenantee Party) to file or institute any legal or administrative claim or action for patent
infringement against the Granting Party or any of the Customers. “Initiates” includes any situation
in which a Covenantee Party files or initiates a legal or administrative claim or action for patent

© 2000 Microsoft Corporation. All rights reserved. 2

 FAT: General Overview of On-Disk Format—Page 3

infringement solely as a counterclaim or equivalent in response to a Granting Party first filing or

instituting a legal or administrative patent infringement claim against such Covenantee Party.

(e) Each of the license and the covenant not to sue described above shall not extend to your
use of any portion of the Specification for any purpose other than (a) to create portions of an
operating system (i) only as necessary to adapt such operating system so that it can directly
interact with a firmware implementation of the Extensible Firmware Initiative Specification v. 1.0
(“EFI Specification”); (ii) only as necessary to emulate an implementation of the EFI Specification;
and (b) to create firmware, applications, utilities and/or drivers that will be used and/or licensed
for only the following purposes: (i) to install, repair and maintain hardware, firmware and portions
of operating system software which are utilized in the boot process; (ii) to provide to an operating
system runtime services that are specified in the EFI Specification; (iii) to diagnose and correct
failures in the hardware, firmware or operating system software; (iv) to query for identification of a
computer system (whether by serial numbers, asset tags, user or otherwise); (v) to perform
inventory of a computer system; and (vi) to manufacture, install and setup any hardware,
firmware or operating system software.

(f) Microsoft reserves all other rights it may have in the Specification and any intellectual
property therein. The furnishing of this document does not give you any license or covenant not
to sue with respect to any other Microsoft patents, trademarks, copyrights or other intellectual
property rights.

2. ADDITIONAL LIMITATIONS AND OBLIGATIONS.

(a)The foregoing license and covenant not to sue is applicable only to the version of the
Specification which you are about to download. It does not apply to any additional versions of or
extensions to the Specification.
(b)Without prejudice to any other rights, Microsoft may terminate this Agreement if you fail to
comply with the terms and conditions of this Agreement. In such event you must destroy all
copies of the Specification.

3. INTELLECTUAL PROPERTY RIGHTS. All ownership, title and intellectual property rights
in and to the Specification are owned by Microsoft or its suppliers.

4. U.S. GOVERNMENT RIGHTS. Any Specification provided to the U.S. Government pursuant
to solicitations issued on or after December 1, 1995 is provided with the commercial rights and
restrictions described elsewhere herein. Any Specification provided to the U.S. Government
pursuant to solicitations issued prior to December 1, 1995 is provided with RESTRICTED
RIGHTS as provided for in FAR, 48 CFR 52.227-14 (JUNE 1987) or DFAR, 48 CFR 252.227-
7013 (OCT 1988), as applicable.

5. EXPORT RESTRICTIONS. Export of the Specification, any part thereof, or any process
or service that is the direct product of the Specification (the foregoing collectively referred to as
the “Restricted Components”) from the United States is regulated by the Export Administration
Regulations (EAR, 15 CFR 730-744) of the U.S. Commerce Department, Bureau of Export
Administration (“BXA”). You agree to comply with the EAR in the export or re-export of the
Restricted Components (i) to any country to which the U.S. has embargoed or restricted the
export of goods or services, which currently include, but are not necessarily limited to Cuba, Iran,
Iraq, Libya, North Korea, Sudan, Syria and the Federal Republic of Yugoslavia (including Serbia,
but not Montenegro), or to any national of any such country, wherever located, who intends to
transmit or transport the Restricted Components back to such country; (ii) to any person or entity
who you know or have reason to know will utilize the Restricted Components in the design,
development or production of nuclear, chemical or biological weapons; or (iii) to any person or
entity who has been prohibited from participating in U.S. export transactions by any federal
agency of the U.S. government. You warrant and represent that neither the BXA nor any other
U.S. federal agency has suspended, revoked or denied your export privileges. For additional
information see http://www.microsoft.com/exporting.

© 2000 Microsoft Corporation. All rights reserved. 3

 FAT: General Overview of On-Disk Format—Page 4

6. DISCLAIMER OF WARRANTIES. To the maximum extent permitted by applicable law,

Microsoft and its suppliers provide the Specification (and all intellectual property therein) and any
(if any) support services related to the Specification (“Support Services”) AS IS AND WITH ALL
FAULTS, and hereby disclaim all warranties and conditions, either express, implied or statutory,
including, but not limited to, any (if any) implied warranties or conditions of merchantability, of
fitness for a particular purpose, of lack of viruses, of accuracy or completeness of responses, of
results, and of lack of negligence or lack of workmanlike effort, all with regard to the Specification,
any intellectual property therein and the provision of or failure to provide Support Services.
ALSO, THERE IS NO WARRANTY OR CONDITION OF TITLE, QUIET ENJOYMENT, QUIET
POSSESSION, CORRESPONDENCE TO DESCRIPTION OR NON-INFRINGEMENT, WITH
REGARD TO THE SPECIFICATION AND ANY INTELLECTUAL PROPERTY THEREIN. THE
ENTIRE RISK AS TO THE QUALITY OF OR ARISING OUT OF USE OR PERFORMANCE OF
THE SPECIFICATION, ANY INTELLECTUAL PROPERTY THEREIN, AND SUPPORT
SERVICES, IF ANY, REMAINS WITH YOU.

7. EXCLUSION OF INCIDENTAL, CONSEQUENTIAL AND CERTAIN OTHER DAMAGES.

To the maximum extent permitted by applicable law, in no event shall Microsoft or its suppliers be
liable for any special, incidental, indirect, or consequential damages whatsoever (including, but
not limited to, damages for loss of profits or confidential or other information, for business
interruption, for personal injury, for loss of privacy, for failure to meet any duty including of good
faith or of reasonable care, for negligence, and for any other pecuniary or other loss whatsoever)
arising out of or in any way related to the use of or inability to use the SPECIFICATION, ANY
INTELLECTUAL PROPERTY THEREIN, the provision of or failure to provide Support Services,
or otherwise under or in connection with any provision of this AGREEMENT, even in the event of
the fault, tort (including negligence), strict liability, breach of contract or breach of warranty of
Microsoft or any supplier, and even if Microsoft or any supplier has been advised of the possibility
of such damages.

8. LIMITATION OF LIABILITY AND REMEDIES. Notwithstanding any damages that you

might incur for any reason whatsoever (including, without limitation, all damages referenced
above and all direct or general damages), the entire liability of Microsoft and any of its suppliers
under any provision of this Agreement and your exclusive remedy for all of the foregoing shall be
limited to the greater of the amount actually paid by you for the Specification or U.S.$5.00. The
foregoing limitations, exclusions and disclaimers shall apply to the maximum extent permitted by
applicable law, even if any remedy fails its essential purpose.

9. APPLICABLE LAW. If you acquired this Specification in the United States, this
Agreement is governed by the laws of the State of Washington. If you acquired this Specification
in Canada, unless expressly prohibited by local law, this Agreement is governed by the laws in
force in the Province of Ontario, Canada; and, in respect of any dispute which may arise
hereunder, you consent to the jurisdiction of the federal and provincial courts sitting in Toronto,
Ontario. If this Specification was acquired outside the United States, then local law may apply.

10.QUESTIONS. Should you have any questions concerning this Agreement, or if you desire to
contact Microsoft for any reason, please contact the Microsoft subsidiary serving your country, or
write: Microsoft Sales Information Center/One Microsoft Way/Redmond, WA 98052-6399.

11.ENTIRE AGREEMENT. This Agreement is the entire agreement between you and Microsoft
relating to the Specification and the Support Services (if any) and they supersede all prior or
contemporaneous oral or written communications, proposals and representations with respect to
the Specification or any other subject matter covered by this Agreement. To the extent the terms
of any Microsoft policies or programs for Support Services conflict with the terms of this
Agreement, the terms of this Agreement shall control.

© 2000 Microsoft Corporation. All rights reserved. 4

 FAT: General Overview of On-Disk Format—Page 5

Si vous avez acquis votre produit Microsoft au CANADA, la garantie limitée suivante vous
concerne :

RENONCIATION AUX GARANTIES. Dans toute la mesure permise par la législation en vigueur,
Microsoft et ses fournisseurs fournissent la Specification (et à toute propriété intellectuelle dans
celle-ci) et tous (selon le cas) les services d’assistance liés à la Specification (“Services
d’assistance”) TELS QUELS ET AVEC TOUS LEURS DÉFAUTS, et par les présentes excluent
toute garantie ou condition, expresse ou implicite, légale ou conventionnelle, écrite ou verbale, y
compris, mais sans limitation, toute (selon le cas) garantie ou condition implicite ou légale de
qualité marchande, de conformité à un usage particulier, d’absence de virus, d’exactitude et
d’intégralité des réponses, de résultats, d’efforts techniques et professionnels et d’absence de
négligence, le tout relativement à la Specification, à toute propriété intellectuelle dans celle-ci et à
la prestation ou à la non-prestation des Services d’assistance. DE PLUS, IL N’Y A AUCUNE
GARANTIE ET CONDITION DE TITRE, DE JOUISSANCE PAISIBLE, DE POSSESSION
PAISIBLE, DE SIMILARITÉ À LA DESCRIPTION ET D’ABSENCE DE CONTREFAÇON
RELATIVEMENT À LA SPÉCIFICATION ET À TOUTE PROPRIÉTÉ INTELLECTUELLE DANS
CELLE-CI. VOUS SUPPORTEZ TOUS LES RISQUES DÉCOULANT DE L’UTILISATION ET DE
LA PERFORMANCE DE LA SPÉCIFICATION ET DE TOUTE PROPRIÉTÉ INTELLECTUELLE
DANS CELLE-CI ET CEUX DÉCOULANT DES SERVICES D’ASSISTANCE (S’IL Y A LIEU).

EXCLUSION DES DOMMAGES INDIRECTS, ACCESSOIRES ET AUTRES. Dans toute la

mesure permise par la législation en vigueur, Microsoft et ses fournisseurs ne sont en aucun cas
responsables de tout dommage spécial, indirect, accessoire, moral ou exemplaire quel qu’il soit
(y compris, mais sans limitation, les dommages entraînés par la perte de bénéfices ou la perte
d’information confidentielle ou autre, l’interruption des affaires, les préjudices corporels, la perte
de confidentialité, le défaut de remplir toute obligation y compris les obligations de bonne foi et de
diligence raisonnable, la négligence et toute autre perte pécuniaire ou autre perte de quelque
nature que ce soit) découlant de, ou de toute autre manière lié à, l’utilisation ou l’impossibilité
d’utiliser la Spécification, toute propriété intellectuelle dans celle-ci, la prestation ou la non-
prestation des Services d’assistance ou autrement en vertu de ou relativement à toute disposition
de cette convention, que ce soit en cas de faute, de délit (y compris la négligence), de
responsabilité stricte, de manquement à un contrat ou de manquement à une garantie de
Microsoft ou de l’un de ses fournisseurs, et ce, même si Microsoft ou l’un de ses fournisseurs a
été avisé de la possibilité de tels dommages.

LIMITATION DE RESPONSABILITÉ ET RECOURS. Malgré tout dommage que vous pourriez

encourir pour quelque raison que ce soit (y compris, mais sans limitation, tous les dommages
mentionnés ci-dessus et tous les dommages directs et généraux), la seule responsabilité de
Microsoft et de ses fournisseurs en vertu de toute disposition de cette convention et votre unique
recours en regard de tout ce qui précède sont limités au plus élevé des montants suivants: soit
(a) le montant que vous avez payé pour la Spécification, soit (b) un montant équivalant à cinq
dollars U.S. (5,00 $ U.S.). Les limitations, exclusions et renonciations ci-dessus s’appliquent
dans toute la mesure permise par la législation en vigueur, et ce même si leur application a pour
effet de priver un recours de son essence.

DROITS LIMITÉS DU GOUVERNEMENT AMÉRICAIN

Tout Produit Logiciel fourni au gouvernement américain conformément à des demandes émises
le ou après le 1er décembre 1995 est offert avec les restrictions et droits commerciaux décrits
ailleurs dans la présente convention. Tout Produit Logiciel fourni au gouvernement américain
conformément à des demandes émises avant le 1er décembre 1995 est offert avec des DROITS
LIMITÉS tels que prévus dans le FAR, 48CFR 52.227-14 (juin 1987) ou dans le FAR, 48CFR
252.227-7013 (octobre 1988), tels qu’applicables.
Sauf lorsqu’expressément prohibé par la législation locale, la présente convention est régie par
les lois en vigueur dans la province d’Ontario, Canada. Pour tout différend qui pourrait découler
des présentes, vous acceptez la compétence des tribunaux fédéraux et provinciaux siégeant à
Toronto, Ontario.

© 2000 Microsoft Corporation. All rights reserved. 5

 FAT: General Overview of On-Disk Format—Page 6

Si vous avez des questions concernant cette convention ou si vous désirez communiquer avec
Microsoft pour quelque raison que ce soit, veuillez contacter la succursale Microsoft desservant
votre pays, ou écrire à: Microsoft Sales Information Center, One Microsoft Way, Redmond,
Washington 98052-6399.

© 2000 Microsoft Corporation. All rights reserved. 6

 FAT: General Overview of On-Disk Format—Page 7

Notational Conventions in this Document

Numbers that have the characters “0x” at the beginning of them are hexadecimal (base 16) numbers.

Any numbers that do not have the characters “0x” at the beginning are decimal (base 10) numbers.

The code fragments in this document are written in the ‘C’ programming language. Strict typing and
syntax are not adhered to.

There are several code fragments in this document that freely mix 32-bit and 16-bit data elements. It is
assumed that you are a programmer who understands how to properly type such operations so that
data is not lost due to truncation of 32-bit values to 16-bit values. Also take note that all data types are
UNSIGNED. Do not do FAT computations with signed integer types, because the computations will
be wrong on some FAT volumes.

General Comments (Applicable to FAT File System All Types)

All of the FAT file systems were originally developed for the IBM PC machine architecture. The
importance of this is that FAT file system on disk data structure is all “little endian.” If we look at one
32-bit FAT entry stored on disk as a series of four 8-bit bytes—the first being byte[0] and the last
being byte[4]—here is where the 32 bits numbered 00 through 31 are (00 being the least significant
bit):

byte[3] 3 3 2 2 2 2 2 2
1 0 9 8 7 6 5 4

byte[2] 2 2 2 2 1 1 1 1
3 2 1 0 9 8 7 6

byte[1] 1 1 1 1 1 1 0 0
5 4 3 2 1 0 9 8

byte[0] 0 0 0 0 0 0 0 0
7 6 5 4 3 2 1 0

This is important if your machine is a “big endian” machine, because you will have to translate
between big and little endian as you move data to and from the disk.

A FAT file system volume is composed of four basic regions, which are laid out in this order on the
volume:
0 – Reserved Region
1 – FAT Region
2 – Root Directory Region (doesn’t exist on FAT32 volumes)
3 – File and Directory Data Region

Boot Sector and BPB

The first important data structure on a FAT volume is called the BPB (BIOS Parameter Block), which
is located in the first sector of the volume in the Reserved Region. This sector is sometimes called the
“boot sector” or the “reserved sector” or the “0th sector,” but the important fact is simply that it is the
first sector of the volume.

This is the first thing about the FAT file system that sometimes causes confusion. In MS-DOS version
1.x, there was not a BPB in the boot sector. In this first version of the FAT file system, there were
only two different formats, the one for single-sided and the one for double-sided 360K 5.25-inch

© 2000 Microsoft Corporation. All rights reserved. 7

 FAT: General Overview of On-Disk Format—Page 8

floppy disks. The determination of which type was on the disk was done by looking at the first byte of
the FAT (the low 8 bits of FAT[0]).

This type of media determination was superseded in MS-DOS version 2.x by putting a BPB in the
boot sector, and the old style of media determination (done by looking at the first byte of the FAT)
was no longer supported. All FAT volumes must have a BPB in the boot sector.

This brings us to the second point of confusion relating to FAT volume determination: What exactly
does a BPB look like? The BPB in the boot sector defined for MS-DOS 2.x only allowed for a FAT
volume with strictly less than 65,536 sectors (32 MB worth of 512-byte sectors). This limitation was
due to the fact that the “total sectors” field was only a 16-bit field. This limitation was addressed by
MS-DOS 3.x, where the BPB was modified to include a new 32-bit field for the total sectors value.

The next BPB change occurred with the Microsoft Windows 95 operating system, specifically OEM
Service Release 2 (OSR2), where the FAT32 type was introduced. FAT16 was limited by the
maximum size of the FAT and the maximum valid cluster size to no more than a 2 GB volume if the
disk had 512-byte sectors. FAT32 addressed this limitation on the amount of disk space that one FAT
volume could occupy so that disks larger than 2 GB only had to have one partition defined.

The FAT32 BPB exactly matches the FAT12/FAT16 BPB up to and including the BPB_TotSec32
field. They differ starting at offset 36, depending on whether the media type is FAT12/FAT16 or
FAT32 (see discussion below for determining FAT type). The relevant point here is that the BPB in
the boot sector of a FAT volume should always be one that has all of the new BPB fields for either the
FAT12/FAT16 or FAT32 BPB type. Doing it this way ensures the maximum compatibility of the FAT
volume and ensures that all FAT file system drivers will understand and support the volume properly,
because it always contains all of the currently defined fields.

NOTE: In the following description, all the fields whose names start with BPB_ are part of the BPB.
All the fields whose names start with BS_ are part of the boot sector and not really part of the BPB.
The following shows the start of sector 0 of a FAT volume, which contains the BPB:

© 2000 Microsoft Corporation. All rights reserved. 8

 FAT: General Overview of On-Disk Format—Page 9

Boot Sector and BPB Structure

Name Offset Size Description
(byte) (bytes)
BS_jmpBoot 0 3 Jump instruction to boot code. This field has two allowed forms:
jmpBoot[0] = 0xEB, jmpBoot[1] = 0x??, jmpBoot[2] = 0x90
and
jmpBoot[0] = 0xE9, jmpBoot[1] = 0x??, jmpBoot[2] = 0x??

0x?? indicates that any 8-bit value is allowed in that byte. What this
forms is a three-byte Intel x86 unconditional branch (jump)
instruction that jumps to the start of the operating system bootstrap
code. This code typically occupies the rest of sector 0 of the volume
following the BPB and possibly other sectors. Either of these forms
is acceptable. JmpBoot[0] = 0xEB is the more frequently used
format.
BS_OEMName 3 8 “MSWIN4.1” There are many misconceptions about this field. It is
only a name string. Microsoft operating systems don’t pay any
attention to this field. Some FAT drivers do. This is the reason that
the indicated string, “MSWIN4.1”, is the recommended setting,
because it is the setting least likely to cause compatibility problems.
If you want to put something else in here, that is your option, but
the result may be that some FAT drivers might not recognize the
volume. Typically this is some indication of what system formatted
the volume.
BPB_BytsPerSec 11 2 Count of bytes per sector. This value may take on only the
following values: 512, 1024, 2048 or 4096. If maximum
compatibility with old implementations is desired, only the value
512 should be used. There is a lot of FAT code in the world that is
basically “hard wired” to 512 bytes per sector and doesn’t bother to
check this field to make sure it is 512. Microsoft operating systems
will properly support 1024, 2048, and 4096.

Note: Do not misinterpret these statements about maximum

compatibility. If the media being recorded has a physical sector size
N, you must use N and this must still be less than or equal to 4096.
Maximum compatibility is achieved by only using media with
specific sector sizes.
BPB_SecPerClus 13 1 Number of sectors per allocation unit. This value must be a power
of 2 that is greater than 0. The legal values are 1, 2, 4, 8, 16, 32, 64,
and 128. Note however, that a value should never be used that
results in a “bytes per cluster” value (BPB_BytsPerSec *
BPB_SecPerClus) greater than 32K (32 * 1024). There is a
misconception that values greater than this are OK. Values that
cause a cluster size greater than 32K bytes do not work properly; do
not try to define one. Some versions of some systems allow 64K
bytes per cluster value. Many application setup programs will not
work correctly on such a FAT volume.
BPB_RsvdSecCnt 14 2 Number of reserved sectors in the Reserved region of the volume
starting at the first sector of the volume. This field must not be 0.
For FAT12 and FAT16 volumes, this value should never be
anything other than 1. For FAT32 volumes, this value is typically
32. There is a lot of FAT code in the world “hard wired” to 1
reserved sector for FAT12 and FAT16 volumes and that doesn’t
bother to check this field to make sure it is 1. Microsoft operating
systems will properly support any non-zero value in this field.

© 2000 Microsoft Corporation. All rights reserved. 9

 FAT: General Overview of On-Disk Format—Page 10

BPB_NumFATs 16 1 The count of FAT data structures on the volume. This field should
always contain the value 2 for any FAT volume of any type.
Although any value greater than or equal to 1 is perfectly valid,
many software programs and a few operating systems’ FAT file
system drivers may not function properly if the value is something
other than 2. All Microsoft file system drivers will support a value
other than 2, but it is still highly recommended that no value other
than 2 be used in this field.

The reason the standard value for this field is 2 is to provide redun-
dancy for the FAT data structure so that if a sector goes bad in one
of the FATs, that data is not lost because it is duplicated in the other
FAT. On non-disk-based media, such as FLASH memory cards,
where such redundancy is a useless feature, a value of 1 may be
used to save the space that a second copy of the FAT uses, but
some FAT file system drivers might not recognize such a volume
properly.
BPB_RootEntCnt 17 2 For FAT12 and FAT16 volumes, this field contains the count of 32-
byte directory entries in the root directory. For FAT32 volumes,
this field must be set to 0. For FAT12 and FAT16 volumes, this
value should always specify a count that when multiplied by 32
results in an even multiple of BPB_BytsPerSec. For maximum
compatibility, FAT16 volumes should use the value 512.
BPB_TotSec16 19 2 This field is the old 16-bit total count of sectors on the volume.
This count includes the count of all sectors in all four regions of the
volume. This field can be 0; if it is 0, then BPB_TotSec32 must be
non-zero. For FAT32 volumes, this field must be 0. For FAT12 and
FAT16 volumes, this field contains the sector count, and
BPB_TotSec32 is 0 if the total sector count “fits” (is less than
0x10000).
BPB_Media 21 1 0xF8 is the standard value for “fixed” (non-removable) media. For
removable media, 0xF0 is frequently used. The legal values for this
field are 0xF0, 0xF8, 0xF9, 0xFA, 0xFB, 0xFC, 0xFD, 0xFE, and
0xFF. The only other important point is that whatever value is put
in here must also be put in the low byte of the FAT[0] entry. This
dates back to the old MS-DOS 1.x media determination noted
earlier and is no longer usually used for anything.
BPB_FATSz16 22 2 This field is the FAT12/FAT16 16-bit count of sectors occupied by
ONE FAT. On FAT32 volumes this field must be 0, and
BPB_FATSz32 contains the FAT size count.
BPB_SecPerTrk 24 2 Sectors per track for interrupt 0x13. This field is only relevant for
media that have a geometry (volume is broken down into tracks by
multiple heads and cylinders) and are visible on interrupt 0x13.
This field contains the “sectors per track” geometry value.
BPB_NumHeads 26 2 Number of heads for interrupt 0x13. This field is relevant as
discussed earlier for BPB_SecPerTrk. This field contains the one
based “count of heads”. For example, on a 1.44 MB 3.5-inch floppy
drive this value is 2.
BPB_HiddSec 28 4 Count of hidden sectors preceding the partition that contains this
FAT volume. This field is generally only relevant for media visible
on interrupt 0x13. This field should always be zero on media that
are not partitioned. Exactly what value is appropriate is operating
system specific.
BPB_TotSec32 32 4 This field is the new 32-bit total count of sectors on the volume.
This count includes the count of all sectors in all four regions of the
volume. This field can be 0; if it is 0, then BPB_TotSec16 must be
non-zero. For FAT32 volumes, this field must be non-zero. For
FAT12/FAT16 volumes, this field contains the sector count if
BPB_TotSec16 is 0 (count is greater than or equal to 0x10000).

© 2000 Microsoft Corporation. All rights reserved. 10

 FAT: General Overview of On-Disk Format—Page 11

At this point, the BPB/boot sector for FAT12 and FAT16 differs from the BPB/boot sector for FAT32.
The first table shows the structure for FAT12 and FAT16 starting at offset 36 of the boot sector.

Fat12 and Fat16 Structure Starting at Offset 36

Name Offset Size Description
(byte) (bytes)
BS_DrvNum 36 1 Int 0x13 drive number (e.g. 0x80). This field supports MS-DOS
bootstrap and is set to the INT 0x13 drive number of the media
(0x00 for floppy disks, 0x80 for hard disks).
NOTE: This field is actually operating system specific.
BS_Reserved1 37 1 Reserved (used by Windows NT). Code that formats FAT volumes
should always set this byte to 0.
BS_BootSig 38 1 Extended boot signature (0x29). This is a signature byte that
indicates that the following three fields in the boot sector are
present.
BS_VolID 39 4 Volume serial number. This field, together with BS_VolLab,
supports volume tracking on removable media. These values allow
FAT file system drivers to detect that the wrong disk is inserted in a
removable drive. This ID is usually generated by simply combining
the current date and time into a 32-bit value.
BS_VolLab 43 11 Volume label. This field matches the 11-byte volume label
recorded in the root directory.
NOTE: FAT file system drivers should make sure that they update
this field when the volume label file in the root directory has its
name changed or created. The setting for this field when there is no
volume label is the string “NO NAME ”.
BS_FilSysType 54 8 One of the strings “FAT12 ”, “FAT16 ”, or “FAT ”.
NOTE: Many people think that the string in this field has
something to do with the determination of what type of FAT—
FAT12, FAT16, or FAT32—that the volume has. This is not true.
You will note from its name that this field is not actually part of the
BPB. This string is informational only and is not used by Microsoft
file system drivers to determine FAT typ,e because it is frequently
not set correctly or is not present. See the FAT Type Determination
section of this document. This string should be set based on the
FAT type though, because some non-Microsoft FAT file system
drivers do look at it.

© 2000 Microsoft Corporation. All rights reserved. 11

 FAT: General Overview of On-Disk Format—Page 12

Here is the structure for FAT32 starting at offset 36 of the boot sector.

FAT32 Structure Starting at Offset 36

Name Offset Size Description
(byte) (bytes)
BPB_FATSz32 36 4 This field is only defined for FAT32 media and does not exist on
FAT12 and FAT16 media. This field is the FAT32 32-bit count of
sectors occupied by ONE FAT. BPB_FATSz16 must be 0.
BPB_ExtFlags 40 2 This field is only defined for FAT32 media and does not exist on
FAT12 and FAT16 media.
Bits 0-3 -- Zero-based number of active FAT. Only valid if mirroring
is disabled.
Bits 4-6 -- Reserved.
Bit 7 -- 0 means the FAT is mirrored at runtime into all FATs.
-- 1 means only one FAT is active; it is the one referenced
in bits 0-3.
Bits 8-15 -- Reserved.
BPB_FSVer 42 2 This field is only defined for FAT32 media and does not exist on
FAT12 and FAT16 media. High byte is major revision number.
Low byte is minor revision number. This is the version number of
the FAT32 volume. This supports the ability to extend the FAT32
media type in the future without worrying about old FAT32 drivers
mounting the volume. This document defines the version to 0:0. If
this field is non-zero, back-level Windows versions will not mount
the volume.
NOTE: Disk utilities should respect this field and not operate on
volumes with a higher major or minor version number than that for
which they were designed. FAT32 file system drivers must check
this field and not mount the volume if it does not contain a version
number that was defined at the time the driver was written.
BPB_RootClus 44 4 This field is only defined for FAT32 media and does not exist on
FAT12 and FAT16 media. This is set to the cluster number of the
first cluster of the root directory, usually 2 but not required to be 2.
NOTE: Disk utilities that change the location of the root directory
should make every effort to place the first cluster of the root
directory in the first non-bad cluster on the drive (i.e., in cluster 2,
unless it’s marked bad). This is specified so that disk repair utilities
can easily find the root directory if this field accidentally gets
zeroed.
BPB_FSInfo 48 2 This field is only defined for FAT32 media and does not exist on
FAT12 and FAT16 media. Sector number of FSINFO structure in the
reserved area of the FAT32 volume. Usually 1.
NOTE: There will be a copy of the FSINFO structure in BackupBoot,
but only the copy pointed to by this field will be kept up to date (i.e.,
both the primary and backup boot record will point to the same
FSINFO sector).
BPB_BkBootSec 50 2 This field is only defined for FAT32 media and does not exist on
FAT12 and FAT16 media. If non-zero, indicates the sector number
in the reserved area of the volume of a copy of the boot record.
Usually 6. No value other than 6 is recommended.
BPB_Reserved 52 12 This field is only defined for FAT32 media and does not exist on
FAT12 and FAT16 media. Reserved for future expansion. Code
that formats FAT32 volumes should always set all of the bytes of
this field to 0.
BS_DrvNum 64 1 This field has the same definition as it does for FAT12 and FAT16
media. The only difference for FAT32 media is that the field is at a
different offset in the boot sector.
BS_Reserved1 65 1 This field has the same definition as it does for FAT12 and FAT16
media. The only difference for FAT32 media is that the field is at a
different offset in the boot sector.

© 2000 Microsoft Corporation. All rights reserved. 12

 FAT: General Overview of On-Disk Format—Page 13

BS_BootSig 66 1 This field has the same definition as it does for FAT12 and FAT16
media. The only difference for FAT32 media is that the field is at a
different offset in the boot sector.
BS_VolID 67 4 This field has the same definition as it does for FAT12 and FAT16
media. The only difference for FAT32 media is that the field is at a
different offset in the boot sector.
BS_VolLab 71 11 This field has the same definition as it does for FAT12 and FAT16
media. The only difference for FAT32 media is that the field is at a
different offset in the boot sector.
BS_FilSysType 82 8 Always set to the string ”FAT32 ”. Please see the note for this
field in the FAT12/FAT16 section earlier. This field has nothing to
do with FAT type determination.

There is one other important note about Sector 0 of a FAT volume. If we consider the contents of the
sector as a byte array, it must be true that sector[510] equals 0x55, and sector[511] equals 0xAA.

NOTE: Many FAT documents mistakenly say that this 0xAA55 signature occupies the “last 2 bytes
of the boot sector”. This statement is correct if — and only if — BPB_BytsPerSec is 512. If
BPB_BytsPerSec is greater than 512, the offsets of these signature bytes do not change (although it is
perfectly OK for the last two bytes at the end of the boot sector to also contain this signature).

Check your assumptions about the value in the BPB_TotSec16/32 field. Assume we have a disk or
partition of size in sectors DskSz. If the BPB TotSec field (either BPB_TotSec16 or BPB_TotSec32
— whichever is non-zero) is less than or equal to DskSz, there is nothing whatsoever wrong with the
FAT volume. In fact, it is not at all unusual to have a BPB_TotSec16/32 value that is slightly smaller
than DskSz. It is also perfectly OK for the BPB_TotSec16/32 value to be considerably smaller than
DskSz.

All this means is that disk space is being wasted. It does not by itself mean that the FAT volume is
damaged in some way. However, if BPB_TotSec16/32 is larger than DskSz, the volume is seriously
damaged or malformed because it extends past the end of the media or overlaps data that follows it on
the disk. Treating a volume for which the BPB_TotSec16/32 value is “too large” for the media or
partition as valid can lead to catastrophic data loss.

FAT Data Structure

The next data structure that is important is the FAT itself. What this data structure does is define a
singly linked list of the “extents” (clusters) of a file. Note at this point that a FAT directory or file
container is nothing but a regular file that has a special attribute indicating it is a directory. The only
other special thing about a directory is that the data or contents of the “file” is a series of 32=byte FAT
directory entries (see discussion below). In all other respects, a directory is just like a file. The FAT
maps the data region of the volume by cluster number. The first data cluster is cluster 2.

The first sector of cluster 2 (the data region of the disk) is computed using the BPB fields for the
volume as follows. First, we determine the count of sectors occupied by the root directory:

RootDirSectors = ((BPB_RootEntCnt * 32) + (BPB_BytsPerSec – 1)) / BPB_BytsPerSec;

Note that on a FAT32 volume the BPB_RootEntCnt value is always 0, so on a FAT32 volume
RootDirSectors is always 0. The 32 in the above is the size of one FAT directory entry in bytes.
Note also that this computation rounds up.

The start of the data region, the first sector of cluster 2, is computed as follows:

© 2000 Microsoft Corporation. All rights reserved. 13

 FAT: General Overview of On-Disk Format—Page 14

If(BPB_FATSz16 != 0)
FATSz = BPB_FATSz16;
Else
FATSz = BPB_FATSz32;

FirstDataSector = BPB_ResvdSecCnt + (BPB_NumFATs * FATSz) + RootDirSectors;

NOTE: This sector number is relative to the first sector of the volume that contains the BPB (the
sector that contains the BPB is sector number 0). This does not necessarily map directly onto the
drive, because sector 0 of the volume is not necessarily sector 0 of the drive due to partitioning.

Given any valid data cluster number N, the sector number of the first sector of that cluster (again
relative to sector 0 of the FAT volume) is computed as follows:

FirstSectorofCluster = ((N – 2) * BPB_SecPerClus) + FirstDataSector;

NOTE: Because BPB_SecPerClus is restricted to powers of 2 (1,2,4,8,16,32….), this means that

division and multiplication by BPB_SecPerClus can actually be performed via SHIFT operations on
2s complement architectures that are usually faster instructions than MULT and DIV instructions. On
current Intel X86 processors, this is largely irrelevant though because the MULT and DIV machine
instructions are heavily optimized for multiplication and division by powers of 2.

FAT Type Determination

There is considerable confusion over exactly how this works, which leads to many “off by 1”, “off by
2”, “off by 10”, and “massively off” errors. It is really quite simple how this works. The FAT type—
one of FAT12, FAT16, or FAT32—is determined by the count of clusters on the volume and nothing
else.

Please read everything in this section carefully, all of the words are important. For example, note that
the statement was “count of clusters.” This is not the same thing as “maximum valid cluster number,”
because the first data cluster is 2 and not 0 or 1.

To begin, let’s discuss exactly how the “count of clusters” value is determined. This is all done using
the BPB fields for the volume. First, we determine the count of sectors occupied by the root directory
as noted earlier.

RootDirSectors = ((BPB_RootEntCnt * 32) + (BPB_BytsPerSec – 1)) / BPB_BytsPerSec;

Note that on a FAT32 volume, the BPB_RootEntCnt value is always 0; so on a FAT32 volume,
RootDirSectors is always 0.

Next, we determine the count of sectors in the data region of the volume:

If(BPB_FATSz16 != 0)
FATSz = BPB_FATSz16;
Else
FATSz = BPB_FATSz32;

If(BPB_TotSec16 != 0)
TotSec = BPB_TotSec16;
Else
TotSec = BPB_TotSec32;

DataSec = TotSec – (BPB_ResvdSecCnt + (BPB_NumFATs * FATSz) + RootDirSectors);

© 2000 Microsoft Corporation. All rights reserved. 14

 FAT: General Overview of On-Disk Format—Page 15

Now we determine the count of clusters:

CountofClusters = DataSec / BPB_SecPerClus;

Please note that this computation rounds down.

Now we can determine the FAT type. Please note carefully or you will commit an off-by-one error!

In the following example, when it says <, it does not mean <=. Note also that the numbers are correct.
The first number for FAT12 is 4085; the second number for FAT16 is 65525. These numbers and the
‘<’ signs are not wrong.

If(CountofClusters < 4085) {

/* Volume is FAT12 */
} else if(CountofClusters < 65525) {
/* Volume is FAT16 */
} else {
/* Volume is FAT32 */
}

This is the one and only way that FAT type is determined. There is no such thing as a FAT12 volume
that has more than 4084 clusters. There is no such thing as a FAT16 volume that has less than 4085
clusters or more than 65,524 clusters. There is no such thing as a FAT32 volume that has less than
65,525 clusters. If you try to make a FAT volume that violates this rule, Microsoft operating systems
will not handle them correctly because they will think the volume has a different type of FAT than
what you think it does.

NOTE: As is noted numerous times earlier, the world is full of FAT code that is wrong. There is a lot
of FAT type code that is off by 1 or 2 or 8 or 10 or 16. For this reason, it is highly recommended that
if you are formatting a FAT volume which has maximum compatibility with all existing FAT code,
then you should you avoid making volumes of any type that have close to 4,085 or 65,525 clusters.
Stay at least 16 clusters on each side away from these cut-over cluster counts.

Note also that the CountofClusters value is exactly that—the count of data clusters starting at cluster
2. The maximum valid cluster number for the volume is CountofClusters + 1, and the “count of
clusters including the two reserved clusters” is CountofClusters + 2.

There is one more important computation related to the FAT. Given any valid cluster number N,
where in the FAT(s) is the entry for that cluster number? The only FAT type for which this is complex
is FAT12. For FAT16 and FAT32, the computation is simple:

If(BPB_FATSz16 != 0)
FATSz = BPB_FATSz16;
Else
FATSz = BPB_FATSz32;

If(FATType == FAT16)
FATOffset = N * 2;
Else if (FATType == FAT32)
FATOffset = N * 4;

ThisFATSecNum = BPB_ResvdSecCnt + (FATOffset / BPB_BytsPerSec);

ThisFATEntOffset = REM(FATOffset / BPB_BytsPerSec);

REM(…) is the remainder operator. That means the remainder after division of FATOffset by
BPB_BytsPerSec. ThisFATSecNum is the sector number of the FAT sector that contains the entry for
cluster N in the first FAT. If you want the sector number in the second FAT, you add FATSz to
ThisFATSecNum; for the third FAT, you add 2*FATSz, and so on.

© 2000 Microsoft Corporation. All rights reserved. 15

 FAT: General Overview of On-Disk Format—Page 16

You now read sector number ThisFATSecNum (remember this is a sector number relative to sector 0
of the FAT volume). Assume this is read into an 8-bit byte array named SecBuff. Also assume that the
type WORD is a 16-bit unsigned and that the type DWORD is a 32-bit unsigned.

If(FATType == FAT16)
FAT16ClusEntryVal = *((WORD *) &SecBuff[ThisFATEntOffset]);
Else
FAT32ClusEntryVal = (*((DWORD *) &SecBuff[ThisFATEntOffset])) & 0x0FFFFFFF;

Fetches the contents of that cluster. To set the contents of this same cluster you do the following:
If(FATType == FAT16)
*((WORD *) &SecBuff[ThisFATEntOffset]) = FAT16ClusEntryVal;
Else {
FAT32ClusEntryVal = FAT32ClusEntryVal & 0x0FFFFFFF;
*((DWORD *) &SecBuff[ThisFATEntOffset]) =
(*((DWORD *) &SecBuff[ThisFATEntOffset])) & 0xF0000000;
*((DWORD *) &SecBuff[ThisFATEntOffset]) =
(*((DWORD *) &SecBuff[ThisFATEntOffset])) | FAT32ClusEntryVal;
}

Note how the FAT32 code above works. A FAT32 FAT entry is actually only a 28-bit entry. The high
4 bits of a FAT32 FAT entry are reserved. The only time that the high 4 bits of FAT32 FAT entries
should ever be changed is when the volume is formatted, at which time the whole 32-bit FAT entry
should be zeroed, including the high 4 bits.

A bit more explanation is in order here, because this point about FAT32 FAT entries seems to cause a
great deal of confusion. Basically 32-bit FAT entries are not really 32-bit values; they are only 28-bit
values. For example, all of these 32-bit cluster entry values: 0x10000000, 0xF0000000, and
0x00000000 all indicate that the cluster is FREE, because you ignore the high 4 bits when you read
the cluster entry value. If the 32-bit free cluster value is currently 0x30000000 and you want to mark
this cluster as bad by storing the value 0x0FFFFFF7 in it. Then the 32-bit entry will contain the value
0x3FFFFFF7 when you are done, because you must preserve the high 4 bits when you write in the
0x0FFFFFF7 bad cluster mark.

Take note that because the BPB_BytsPerSec value is always divisible by 2 and 4, you never have to
worry about a FAT16 or FAT32 FAT entry spanning over a sector boundary (this is not true of
FAT12).

The code for FAT12 is more complicated because there are 1.5 bytes (12-bits) per FAT entry.
if (FATType == FAT12)
FATOffset = N + (N / 2);
/* Multiply by 1.5 without using floating point, the divide by 2 rounds DOWN */

ThisFATSecNum = BPB_ResvdSecCnt + (FATOffset / BPB_BytsPerSec);

ThisFATEntOffset = REM(FATOffset / BPB_BytsPerSec);

We now have to check for the sector boundary case:

© 2000 Microsoft Corporation. All rights reserved. 16

 FAT: General Overview of On-Disk Format—Page 17

If(ThisFATEntOffset == (BPB_BytsPerSec – 1)) {

/* This cluster access spans a sector boundary in the FAT */
/* There are a number of strategies to handling this. The */
/* easiest is to always load FAT sectors into memory */
/* in pairs if the volume is FAT12 (if you want to load */
/* FAT sector N, you also load FAT sector N+1 immediately */
/* following it in memory unless sector N is the last FAT */
/* sector). It is assumed that this is the strategy used here */
/* which makes this if test for a sector boundary span */
/* unnecessary. */
}

We now access the FAT entry as a WORD just as we do for FAT16, but if the cluster number is
EVEN, we only want the low 12-bits of the 16-bits we fetch; and if the cluster number is ODD, we
only want the high 12-bits of the 16-bits we fetch.

FAT12ClusEntryVal = *((WORD *) &SecBuff[ThisFATEntOffset]);

If(N & 0x0001)
FAT12ClusEntryVal = FAT12ClusEntryVal >> 4; /* Cluster number is ODD */
Else
FAT12ClusEntryVal = FAT12ClusEntryVal & 0x0FFF; /* Cluster number is EVEN */

Fetches the contents of that cluster. To set the contents of this same cluster you do the following:

If(N & 0x0001) {

FAT12ClusEntryVal = FAT12ClusEntryVal << 4; /* Cluster number is ODD */
*((WORD *) &SecBuff[ThisFATEntOffset]) =
(*((WORD *) &SecBuff[ThisFATEntOffset])) & 0x000F;
} Else {
FAT12ClusEntryVal = FAT12ClusEntryVal & 0x0FFF; /* Cluster number is EVEN */
*((WORD *) &SecBuff[ThisFATEntOffset]) =
(*((WORD *) &SecBuff[ThisFATEntOffset])) & 0xF000;
}
*((WORD *) &SecBuff[ThisFATEntOffset]) =
(*((WORD *) &SecBuff[ThisFATEntOffset])) | FAT12ClusEntryVal;

NOTE: It is assumed that the >> operator shifts a bit value of 0 into the high 4 bits and that the <<
operator shifts a bit value of 0 into the low 4 bits.

The way the data of a file is associated with the file is as follows. In the directory entry, the cluster
number of the first cluster of the file is recorded. The first cluster (extent) of the file is the data
associated with this first cluster number, and the location of that data on the volume is computed from
the cluster number as described earlier (computation of FirstSectorofCluster).

Note that a zero-length file—a file that has no data allocated to it—has a first cluster number of 0
placed in its directory entry. This cluster location in the FAT (see earlier computation of
ThisFATSecNum and ThisFATEntOffset) contains either an EOC mark (End Of Clusterchain) or the
cluster number of the next cluster of the file. The EOC value is FAT type dependant (assume
FATContent is the contents of the cluster entry in the FAT being checked to see whether it is an EOC
mark):

IsEOF = FALSE;
If(FATType == FAT12) {
If(FATContent >= 0x0FF8)
IsEOF = TRUE;
} else if(FATType == FAT16) {
If(FATContent >= 0xFFF8)
IsEOF = TRUE;
} else if (FATType == FAT32) {
If(FATContent >= 0x0FFFFFF8)
IsEOF = TRUE;
}

© 2000 Microsoft Corporation. All rights reserved. 17

 FAT: General Overview of On-Disk Format—Page 18

Note that the cluster number whose cluster entry in the FAT contains the EOC mark is allocated to the
file and is also the last cluster allocated to the file. Microsoft operating system FAT drivers use the
EOC value 0x0FFF for FAT12, 0xFFFF for FAT16, and 0x0FFFFFFF for FAT32 when they set the
contents of a cluster to the EOC mark. There are various disk utilities for Microsoft operating systems
that use a different value, however.

There is also a special “BAD CLUSTER” mark. Any cluster that contains the “BAD CLUSTER”
value in its FAT entry is a cluster that should not be placed on the free list because it is prone to disk
errors. The “BAD CLUSTER” value is 0x0FF7 for FAT12, 0xFFF7 for FAT16, and 0x0FFFFFF7 for
FAT32. The other relevant note here is that these bad clusters are also lost clusters—clusters that
appear to be allocated because they contain a non-zero value but which are not part of any files
allocation chain. Disk repair utilities must recognize lost clusters that contain this special value as bad
clusters and not change the content of the cluster entry.

NOTE: It is not possible for the bad cluster mark to be an allocatable cluster number on FAT12 and
FAT16 volumes, but it is feasible for 0x0FFFFFF7 to be an allocatable cluster number on FAT32
volumes. To avoid possible confusion by disk utilities, no FAT32 volume should ever be configured
such that 0x0FFFFFF7 is an allocatable cluster number.

The list of free clusters in the FAT is nothing more than the list of all clusters that contain the value 0
in their FAT cluster entry. Note that this value must be fetched as described earlier as for any other
FAT entry that is not free. This list of free clusters is not stored anywhere on the volume; it must be
computed when the volume is mounted by scanning the FAT for entries that contain the value 0. On
FAT32 volumes, the BPB_FSInfo sector may contain a valid count of free clusters on the volume. See
the documentation of the FAT32 FSInfo sector.

What are the two reserved clusters at the start of the FAT for? The first reserved cluster, FAT[0],
contains the BPB_Media byte value in its low 8 bits, and all other bits are set to 1. For example, if the
BPB_Media value is 0xF8, for FAT12 FAT[0] = 0x0FF8, for FAT16 FAT[0] = 0xFFF8, and for
FAT32 FAT[0] = 0x0FFFFFF8. The second reserved cluster, FAT[1], is set by FORMAT to the EOC
mark. On FAT12 volumes, it is not used and is simply always contains an EOC mark. For FAT16 and
FAT32, the file system driver may use the high two bits of the FAT[1] entry for dirty volume flags (all
other bits, are always left set to 1). Note that the bit location is different for FAT16 and FAT32,
because they are the high 2 bits of the entry.

For FAT16:
ClnShutBitMask = 0x8000;
HrdErrBitMask = 0x4000;

For FAT32:
ClnShutBitMask = 0x08000000;
HrdErrBitMask = 0x04000000;

Bit ClnShutBitMask – If bit is 1, volume is “clean”.

If bit is 0, volume is “dirty”. This indicates that the file system driver did not
Dismount the volume properly the last time it had the volume mounted. It
would be a good idea to run a Chkdsk/Scandisk disk repair utility on it,
because it may be damaged.
Bit HrdErrBitMask – If this bit is 1, no disk read/write errors were encountered.
If this bit is 0, the file system driver encountered a disk I/O error on the
Volume the last time it was mounted, which is an indicator that some sectors
may have gone bad on the volume. It would be a good idea to run a
Chkdsk/Scandisk disk repair utility that does surface analysis on it to look
for new bad sectors.

© 2000 Microsoft Corporation. All rights reserved. 18

 FAT: General Overview of On-Disk Format—Page 19

Here are two more important notes about the FAT region of a FAT volume:
1. The last sector of the FAT is not necessarily all part of the FAT. The FAT stops at the cluster
number in the last FAT sector that corresponds to the entry for cluster number
CountofClusters + 1 (see the CountofClusters computation earlier), and this entry is not
necessarily at the end of the last FAT sector. FAT code should not make any assumptions
about what the contents of the last FAT sector are after the CountofClusters + 1 entry. FAT
format code should zero the bytes after this entry though.
2. The BPB_FATSz16 (BPB_FATSz32 for FAT32 volumes) value may be bigger than it needs
to be. In other words, there may be totally unused FAT sectors at the end of each FAT in the
FAT region of the volume. For this reason, the last sector of the FAT is always computed
using the CountofClusters + 1 value, never from the BPB_FATSz16/32 value. FAT code
should not make any assumptions about what the contents of these “extra” FAT sectors are.
FAT format code should zero the contents of these extra FAT sectors though.

FAT Volume Initialization

At this point, the careful reader should have one very interesting question. Given that the FAT type
(FAT12, FAT16, or FAT32) is dependant on the number of clusters—and that the sectors available in
the data area of a FAT volume is dependant on the size of the FAT—when handed an unformatted
volume that does not yet have a BPB, how do you determine all this and compute the proper values to
put in BPB_SecPerClus and either BPB_FATSz16 or BPB_FATSz32? The way Microsoft operating
systems do this is with a fixed value, several tables, and a clever piece of arithmetic.

Microsoft operating systems only do FAT12 on floppy disks. Because there is a limited number of
floppy formats that all have a fixed size, this is done with a simple table:

“If it is a floppy of this type, then the BPB looks like this.”

There is no dynamic computation for FAT12. For the FAT12 formats, all the computation for
BPB_SecPerClus and BPB_FATSz16 was worked out by hand on a piece of paper and recorded in the
table (being careful of course that the resultant cluster count was always less than 4085). If your media
is larger than 4 MB, do not bother with FAT12. Use smaller BPB_SecPerClus values so that the
volume will be FAT16.

The rest of this section is totally specific to drives that have 512 bytes per sector. You cannot use these
tables, or the clever arithmetic, with drives that have a different sector size. The “fixed value” is
simply a volume size that is the “FAT16 to FAT32 cutover value”. Any volume size smaller than this
is FAT16 and any volume of this size or larger is FAT32. For Windows, this value is 512 MB. Any
FAT volume smaller than 512 MB is FAT16, and any FAT volume of 512 MB or larger is FAT32.

Please don’t draw an incorrect conclusion here.

There are many FAT16 volumes out there that are larger than 512 MB. There are various ways to
force the format to be FAT16 rather than the default of FAT32, and there is a great deal of code that
implements different limits. All we are talking about here is the default cutover value for MS-DOS
and Windows on volumes that have not yet been formatted. There are two tables—one is for FAT16
and the other is for FAT32. An entry in these tables is selected based on the size of the volume in 512
byte sectors (the value that will go in BPB_TotSec16 or BPB_TotSec32), and the value that this table
sets is the BPB_SecPerClus value.

© 2000 Microsoft Corporation. All rights reserved. 19

 FAT: General Overview of On-Disk Format—Page 20

struct DSKSZTOSECPERCLUS {
DWORD DiskSize;
BYTE SecPerClusVal;
};

/*
*This is the table for FAT16 drives. NOTE that this table includes
* entries for disk sizes larger than 512 MB even though typically
* only the entries for disks < 512 MB in size are used.
* The way this table is accessed is to look for the first entry
* in the table for which the disk size is less than or equal
* to the DiskSize field in that table entry. For this table to
* work properly BPB_RsvdSecCnt must be 1, BPB_NumFATs
* must be 2, and BPB_RootEntCnt must be 512. Any of these values
* being different may require the first table entries DiskSize value
* to be changed otherwise the cluster count may be to low for FAT16.
*/
DSKSZTOSECPERCLUS DskTableFAT16 [] = {
{ 8400, 0}, /* disks up to 4.1 MB, the 0 value for SecPerClusVal trips an error */
{ 32680, 2}, /* disks up to 16 MB, 1k cluster */
{ 262144, 4}, /* disks up to 128 MB, 2k cluster */
{ 524288, 8}, /* disks up to 256 MB, 4k cluster */
{ 1048576, 16}, /* disks up to 512 MB, 8k cluster */
/* The entries after this point are not used unless FAT16 is forced */
{ 2097152, 32}, /* disks up to 1 GB, 16k cluster */
{ 4194304, 64}, /* disks up to 2 GB, 32k cluster */
{ 0xFFFFFFFF, 0} /* any disk greater than 2GB, 0 value for SecPerClusVal trips an error */
};

/*
* This is the table for FAT32 drives. NOTE that this table includes
* entries for disk sizes smaller than 512 MB even though typically
* only the entries for disks >= 512 MB in size are used.
* The way this table is accessed is to look for the first entry
* in the table for which the disk size is less than or equal
* to the DiskSize field in that table entry. For this table to
* work properly BPB_RsvdSecCnt must be 32, and BPB_NumFATs
* must be 2. Any of these values being different may require the first
* table entries DiskSize value to be changed otherwise the cluster count
* may be to low for FAT32.
*/
DSKSZTOSECPERCLUS DskTableFAT32 [] = {
{ 66600, 0}, /* disks up to 32.5 MB, the 0 value for SecPerClusVal trips an error */
{ 532480, 1}, /* disks up to 260 MB, .5k cluster */
{ 16777216, 8}, /* disks up to 8 GB, 4k cluster */
{ 33554432, 16}, /* disks up to 16 GB, 8k cluster */
{ 67108864, 32}, /* disks up to 32 GB, 16k cluster */
{ 0xFFFFFFFF, 64}/* disks greater than 32GB, 32k cluster */
};

So given a disk size and a FAT type of FAT16 or FAT32, we now have a BPB_SecPerClus value. The
only thing we have left is do is to compute how many sectors the FAT takes up so that we can set
BPB_FATSz16 or BPB_FATSz32. Note that at this point we assume that BPB_RootEntCnt,
BPB_RsvdSecCnt, and BPB_NumFATs are appropriately set. We also assume that DskSize is the size
of the volume that we are either going to put in BPB_TotSec32 or BPB_TotSec16.

© 2000 Microsoft Corporation. All rights reserved. 20

 FAT: General Overview of On-Disk Format—Page 21

RootDirSectors = ((BPB_RootEntCnt * 32) + (BPB_BytsPerSec – 1)) / BPB_BytsPerSec;

TmpVal1 = DskSize – (BPB_ResvdSecCnt + RootDirSectors);
TmpVal2 = (256 * BPB_SecPerClus) + BPB_NumFATs;
If(FATType == FAT32)
TmpVal2 = TmpVal2 / 2;
FATSz = (TMPVal1 + (TmpVal2 – 1)) / TmpVal2;
If(FATType == FAT32) {
BPB_FATSz16 = 0;
BPB_FATSz32 = FATSz;
} else {
BPB_FATSz16 = LOWORD(FATSz);
/* there is no BPB_FATSz32 in a FAT16 BPB */
}

Do not spend too much time trying to figure out why this math works. The basis for the computation
is complicated; the important point is that this is how Microsoft operating systems do it, and it works.
Note, however, that this math does not work perfectly. It will occasionally set a FATSz that is up to
2 sectors too large for FAT16, and occasionally up to 8 sectors too large for FAT32. It will never
compute a FATSz value that is too small, however. Because it is OK to have a FATSz that is too
large, at the expense of wasting a few sectors, the fact that this computation is surprisingly simple
more than makes up for it being off in a safe way in some cases.

FAT32 FSInfo Sector Structure and Backup Boot Sector

On a FAT32 volume, the FAT can be a large data structure, unlike on FAT16 where it is limited to a
maximum of 128K worth of sectors and FAT12 where it is limited to a maximum of 6K worth of
sectors. For this reason, a provision is made to store the “last known” free cluster count on the FAT32
volume so that it does not have to be computed as soon as an API call is made to ask how much free
space there is on the volume (like at the end of a directory listing). The FSInfo sector number is the
value in the BPB_FSInfo field; for Microsoft operating systems it is always set to 1. Here is the
structure of the FSInfo sector:

FAT32 FSInfo Sector Structure and Backup Boot Sector

Name Offset Size Description
(byte) (bytes)
FSI_LeadSig 0 4 Value 0x41615252. This lead signature is used to validate that this
is in fact an FSInfo sector.
FSI_Reserved1 4 480 This field is currently reserved for future expansion. FAT32 format
code should always initialize all bytes of this field to 0. Bytes in
this field must currently never be used.
FSI_StrucSig 484 4 Value 0x61417272. Another signature that is more localized in the
sector to the location of the fields that are used.
FSI_Free_Count 488 4 Contains the last known free cluster count on the volume. If the
value is 0xFFFFFFFF, then the free count is unknown and must be
computed. Any other value can be used, but is not necessarily
correct. It should be range checked at least to make sure it is <=
volume cluster count.
FSI_Nxt_Free 492 4 This is a hint for the FAT driver. It indicates the cluster number at
which the driver should start looking for free clusters. Because a
FAT32 FAT is large, it can be rather time consuming if there are a
lot of allocated clusters at the start of the FAT and the driver starts
looking for a free cluster starting at cluster 2. Typically this value is
set to the last cluster number that the driver allocated. If the value is
0xFFFFFFFF, then there is no hint and the driver should start
looking at cluster 2. Any other value can be used, but should be
checked first to make sure it is a valid cluster number for the
volume.
FSI_Reserved2 496 12 This field is currently reserved for future expansion. FAT32 format
code should always initialize all bytes of this field to 0. Bytes in
this field must currently never be used.

© 2000 Microsoft Corporation. All rights reserved. 21

 FAT: General Overview of On-Disk Format—Page 22

FSI_TrailSig 508 4 Value 0xAA550000. This trail signature is used to validate that this
is in fact an FSInfo sector. Note that the high 2 bytes of this
value—which go into the bytes at offsets 510 and 511—match the
signature bytes used at the same offsets in sector 0.

Another feature on FAT32 volumes that is not present on FAT16/FAT12 is the BPB_BkBootSec field.
FAT16/FAT12 volumes can be totally lost if the contents of sector 0 of the volume are overwritten or
sector 0 goes bad and cannot be read. This is a “single point of failure” for FAT16 and FAT12
volumes. The BPB_BkBootSec field reduces the severity of this problem for FAT32 volumes, because
starting at that sector number on the volume—6—there is a backup copy of the boot sector
information including the volume’s BPB.

In the case where the sector 0 information has been accidentally overwritten, all a disk repair utility
has to do is restore the boot sector(s) from the backup copy. In the case where sector 0 goes bad, this
allows the volume to be mounted so that the user can access data before replacing the disk.

This second case—sector 0 goes bad—is the reason why no value other than 6 should ever be placed
in the BPB_BkBootSec field. If sector 0 is unreadable, various operating systems are “hard wired” to
check for backup boot sector(s) starting at sector 6 of the FAT32 volume. Note that starting at the
BPB_BkBootSec sector is a complete boot record. The Microsoft FAT32 “boot sector” is actually
three 512-byte sectors long. There is a copy of all three of these sectors starting at the
BPB_BkBootSec sector. A copy of the FSInfo sector is also there, even though the BPB_FSInfo field
in this backup boot sector is set to the same value as is stored in the sector 0 BPB.

NOTE: All 3 of these sectors have the 0xAA55 signature in sector offsets 510 and 511, just like the
first boot sector does (see the earlier discussion at the end of the BPB structure description).

FAT Directory Structure

We will first talk about short directory entries and ignore long directory entries for the moment.

A FAT directory is nothing but a “file” composed of a linear list of 32-byte structures. The only
special directory, which must always be present, is the root directory. For FAT12 and FAT16 media,
the root directory is located in a fixed location on the disk immediately following the last FAT and is
of a fixed size in sectors computed from the BPB_RootEntCnt value (see computations for
RootDirSectors earlier in this document). For FAT12 and FAT16 media, the first sector of the root
directory is sector number relative to the first sector of the FAT volume:

FirstRootDirSecNum = BPB_ResvdSecCnt + (BPB_NumFATs * BPB_FATSz16);

For FAT32, the root directory can be of variable size and is a cluster chain, just like any other
directory is. The first cluster of the root directory on a FAT32 volume is stored in BPB_RootClus.
Unlike other directories, the root directory itself on any FAT type does not have any date or time
stamps, does not have a file name (other than the implied file name “\”), and does not contain “.” and
“..” files as the first two directory entries in the directory. The only other special aspect of the root
directory is that it is the only directory on the FAT volume for which it is valid to have a file that has
only the ATTR_VOLUME_ID attribute bit set (see below).

© 2000 Microsoft Corporation. All rights reserved. 22

 FAT: General Overview of On-Disk Format—Page 23

FAT 32 Byte Directory Entry Structure

Name Offset Size Description
(byte) (bytes)
DIR_Name 0 11 Short name.
DIR_Attr 11 1 File attributes:
ATTR_READ_ONLY 0x01
ATTR_HIDDEN 0x02
ATTR_SYSTEM 0x04
ATTR_VOLUME_ID 0x08
ATTR_DIRECTORY 0x10
ATTR_ARCHIVE 0x20
ATTR_LONG_NAME ATTR_READ_ONLY |
ATTR_HIDDEN |
ATTR_SYSTEM |
ATTR_VOLUME_ID
The upper two bits of the attribute byte are reserved and should
always be set to 0 when a file is created and never modified or
looked at after that.
DIR_NTRes 12 1 Reserved for use by Windows NT. Set value to 0 when a file is
created and never modify or look at it after that.
DIR_CrtTimeTenth 13 1 Millisecond stamp at file creation time. This field actually
contains a count of tenths of a second. The granularity of the
seconds part of DIR_CrtTime is 2 seconds so this field is a
count of tenths of a second and its valid value range is 0-199
inclusive.
DIR_CrtTime 14 2 Time file was created.
DIR_CrtDate 16 2 Date file was created.
DIR_LstAccDate 18 2 Last access date. Note that there is no last access time, only a
date. This is the date of last read or write. In the case of a write,
this should be set to the same date as DIR_WrtDate.
DIR_FstClusHI 20 2 High word of this entry’s first cluster number (always 0 for a
FAT12 or FAT16 volume).
DIR_WrtTime 22 2 Time of last write. Note that file creation is considered a write.
DIR_WrtDate 24 2 Date of last write. Note that file creation is considered a write.
DIR_FstClusLO 26 2 Low word of this entry’s first cluster number.
DIR_FileSize 28 4 32-bit DWORD holding this file’s size in bytes.

DIR_Name[0]
Special notes about the first byte (DIR_Name[0]) of a FAT directory entry:

• If DIR_Name[0] == 0xE5, then the directory entry is free (there is no file or directory name in this
entry).

• If DIR_Name[0] == 0x00, then the directory entry is free (same as for 0xE5), and there are no
allocated directory entries after this one (all of the DIR_Name[0] bytes in all of the entries after
this one are also set to 0).

The special 0 value, rather than the 0xE5 value, indicates to FAT file system driver code that the
rest of the entries in this directory do not need to be examined because they are all free.

• If DIR_Name[0] == 0x05, then the actual file name character for this byte is 0xE5. 0xE5 is
actually a valid KANJI lead byte value for the character set used in Japan. The special 0x05 value
is used so that this special file name case for Japan can be handled properly and not cause FAT file
system code to think that the entry is free.

© 2000 Microsoft Corporation. All rights reserved. 23

 FAT: General Overview of On-Disk Format—Page 24

The DIR_Name field is actually broken into two parts+ the 8-character main part of the name, and the
3-character extension. These two parts are “trailing space padded” with bytes of 0x20.

DIR_Name[0] may not equal 0x20. There is an implied ‘.’ character between the main part of the
name and the extension part of the name that is not present in DIR_Name. Lower case characters are
not allowed in DIR_Name (what these characters are is country specific).

The following characters are not legal in any bytes of DIR_Name:

• Values less than 0x20 except for the special case of 0x05 in DIR_Name[0] described above.
• 0x22, 0x2A, 0x2B, 0x2C, 0x2E, 0x2F, 0x3A, 0x3B, 0x3C, 0x3D, 0x3E, 0x3F, 0x5B, 0x5C, 0x5D,
and 0x7C.

Here are some examples of how a user-entered name maps into DIR_Name:

“foo.bar” -> “FOO BAR”

“FOO.BAR” -> “FOO BAR”
“Foo.Bar” -> “FOO BAR”
“foo” -> “FOO “
“foo.” -> “FOO “
“PICKLE.A” -> “PICKLE A “
“prettybg.big” -> “PRETTYBGBIG”
“.big” -> illegal, DIR_Name[0] cannot be 0x20

In FAT directories all names are unique. Look at the first three examples earlier. Those different
names all refer to the same file, and there can only be one file with DIR_Name set to “FOO BAR”
in any directory.

DIR_Attr specifies attributes of the file:

ATTR_READ_ONLY Indicates that writes to the file should fail.

ATTR_HIDDEN Indicates that normal directory listings should not show this file.
ATTR_SYSTEM Indicates that this is an operating system file.
ATTR_VOLUME_ID There should only be one “file” on the volume that has this attribute
set, and that file must be in the root directory. This name of this file is
actually the label for the volume. DIR_FstClusHI and
DIR_FstClusLO must always be 0 for the volume label (no data
clusters are allocated to the volume label file).
ATTR_DIRECTORY Indicates that this file is actually a container for other files.
ATTR_ARCHIVE This attribute supports backup utilities. This bit is set by the FAT file
system driver when a file is created, renamed, or written to. Backup
utilities may use this attribute to indicate which files on the volume
have been modified since the last time that a backup was performed.

Note that the ATTR_LONG_NAME attribute bit combination indicates that the “file” is actually part
of the long name entry for some other file. See the next section for more information on this attribute
combination.

When a directory is created, a file with the ATTR_DIRECTORY bit set in its DIR_Attr field, you set
its DIR_FileSize to 0. DIR_FileSize is not used and is always 0 on a file with the
ATTR_DIRECTORY attribute (directories are sized by simply following their cluster chains to the
EOC mark). One cluster is allocated to the directory (unless it is the root directory on a FAT16/FAT12
volume), and you set DIR_FstClusLO and DIR_FstClusHI to that cluster number and place an EOC
mark in that clusters entry in the FAT. Next, you initialize all bytes of that cluster to 0. If the directory
is the root directory, you are done (there are no dot or dotdot entries in the root directory). If the
directory is not the root directory, you need to create two special entries in the first two 32-byte

© 2000 Microsoft Corporation. All rights reserved. 24

 FAT: General Overview of On-Disk Format—Page 25

directory entries of the directory (the first two 32 byte entries in the data region of the cluster you just
allocated).

The first directory entry has DIR_Name set to:

“. ”

The second has DIR_Name set to:

“.. ”

These are called the dot and dotdot entries. The DIR_FileSize field on both entries is set to 0, and all
of the date and time fields in both of these entries are set to the same values as they were in the
directory entry for the directory that you just created. You now set DIR_FstClusLO and
DIR_FstClusHI for the dot entry (the first entry) to the same values you put in those fields for the
directories directory entry (the cluster number of the cluster that contains the dot and dotdot entries).

Finally, you set DIR_FstClusLO and DIR_FstClusHI for the dotdot entry (the second entry) to the
first cluster number of the directory in which you just created the directory (value is 0 if this directory
is the root directory even for FAT32 volumes).

Here is the summary for the dot and dotdot entries:

• The dot entry is a directory that points to itself.
• The dotdot entry points to the starting cluster of the parent of this directory (which is 0 if this
directories parent is the root directory).

Date and Time Formats

Many FAT file systems do not support Date/Time other than DIR_WrtTime and DIR_WrtDate. For
this reason, DIR_CrtTimeMil, DIR_CrtTime, DIR_CrtDate, and DIR_LstAccDate are actually
optional fields. DIR_WrtTime and DIR_WrtDate must be supported, however. If the other date and
time fields are not supported, they should be set to 0 on file create and ignored on other file
operations.

Date Format. A FAT directory entry date stamp is a 16-bit field that is basically a date relative to the
MS-DOS epoch of 01/01/1980. Here is the format (bit 0 is the LSB of the 16-bit word, bit 15 is the
MSB of the 16-bit word):

Bits 0–4: Day of month, valid value range 1-31 inclusive.

Bits 5–8: Month of year, 1 = January, valid value range 1–12 inclusive.
Bits 9–15: Count of years from 1980, valid value range 0–127 inclusive (1980–2107).

Time Format. A FAT directory entry time stamp is a 16-bit field that has a granularity of 2 seconds.
Here is the format (bit 0 is the LSB of the 16-bit word, bit 15 is the MSB of the 16-bit word).

Bits 0–4: 2-second count, valid value range 0–29 inclusive (0 – 58 seconds).
Bits 5–10: Minutes, valid value range 0–59 inclusive.
Bits 11–15: Hours, valid value range 0–23 inclusive.

The valid time range is from Midnight 00:00:00 to 23:59:58.

FAT Long Directory Entries

In adding long directory entries to the FAT file system it was crucial that their addition to the FAT file
system's existing design:

© 2000 Microsoft Corporation. All rights reserved. 25

 FAT: General Overview of On-Disk Format—Page 26

• Be essentially transparent on earlier versions of MS-DOS. The primary goal being that existing
MS-DOS APIs on previous versions of MS-DOS/Windows do not easily "find" long directory
entries. The only MS-DOS APIs that can "find" long directory entries are the FCB-based-find
APIs when used with a full meta-character matching pattern (i.e. *.*) and full attribute matching
bits (i.e. matching attributes are FFh). On post-Windows 95 versions of MS-DOS/Windows, no
MS-DOS API can accidentally "find" a single long directory entry.
• Be located in close physical proximity, on the media, to the short directory entries they are
associated with. As will be evident, long directory entries are immediately contiguous to the short
directory entry they are associated with and their existence imposes an unnoticeable performance
impact on the file system.
• If detected by disk maintenance utilities, they do not jeopardize the integrity of existing file data.
Disk maintenance utilities typically do not use MS-DOS APIs to access on-media file-system-
specific data structures. Rather they read physical or logical sector information from the disk and
judge for themselves what the directory entries contain. Based on the heuristics employed in the
utilities, the utility may take various steps to "repair" what it perceives to be "damaged" file-
system-specific data structures. Long directory entries were added to the FAT file system in such
a way as to not cause the loss of file data if a disk containing long directory entries was "repaired"
by a pre-Windows 95-compatible disk utility on a previous version of MS-DOS/Windows.

In order to meet the goals of locality-of-access and transparency, the long directory entry is defined as
a short directory entry with a special attribute. As described previously, a long directory entry is just a
regular directory entry in which the attribute field has a value of:

ATTR_LONG_NAME ATTR_READ_ONLY |
ATTR_HIDDEN |
ATTR_SYSTEM |
ATTR_VOLUME_ID

A mask for determining whether an entry is a long-name sub-component should also be defined:

ATTR_LONG_NAME_MASK ATTR_READ_ONLY |
ATTR_HIDDEN |
ATTR_SYSTEM |
ATTR_VOLUME_ID |
ATTR_DIRECTORY |
ATTR_ARCHIVE

When such a directory entry is encountered it is given special treatment by the file system. It is
treated as part of a set of directory entries that are associated with a single short directory entry. Each
long directory entry has the following structure:

FAT Long Directory Entry Structure

Name Offset Size Description
(byte) (bytes)
LDIR_Ord 0 1 The order of this entry in the sequence of long dir entries
associated with the short dir entry at the end of the long dir set.

If masked with 0x40 (LAST_LONG_ENTRY), this indicates the

entry is the last long dir entry in a set of long dir entries. All valid
sets of long dir entries must begin with an entry having this
mask.
LDIR_Name1 1 10 Characters 1-5 of the long-name sub-component in this dir entry.
LDIR_Attr 11 1 Attributes - must be ATTR_LONG_NAME

© 2000 Microsoft Corporation. All rights reserved. 26

 FAT: General Overview of On-Disk Format—Page 27

LDIR_Type 12 1 If zero, indicates a directory entry that is a sub-component of a

long name. NOTE: Other values reserved for future extensions.

Non-zero implies other dirent types.

LDIR_Chksum 13 1 Checksum of name in the short dir entry at the end of the long dir
set.
LDIR_Name2 14 12 Characters 6-11 of the long-name sub-component in this dir
entry.
LDIR_FstClusLO 26 2 Must be ZERO. This is an artifact of the FAT "first cluster" and
must be zero for compatibility with existing disk utilities. It's
meaningless in the context of a long dir entry.
LDIR_Name3 28 4 Characters 12-13 of the long-name sub-component in this dir
entry.

Organization and Association of Short & Long Directory Entries

A set of long entries is always associated with a short entry that they always immediately precede.
Long entries are paired with short entries for one reason: only short directory entries are visible to
previous versions of MS-DOS/Windows. Without a short entry to accompany it, a long directory
entry would be completely invisible on previous versions of MS-DOS/Windows. A long entry never
legally exists all by itself. If long entries are found without being paired with a valid short entry, they
are termed orphans. The following figure depicts a set of n long directory entries associated with it's
single short entry.

Long entries always immediately precede and are physically contiguous with, the short entry they are
associated with. The file system makes a few other checks to ensure that a set of long entries is
actually associated with a short entry.

Sequence Of Long Directory Entries

Entry Ordinal
Nth Long entry LAST_LONG_ENTRY (0x40) | N
… Additional Long Entries …
1st Long entry 1
Short Entry Associated With Preceding Long Entries (not applicable)

First, every member of a set of long entries is uniquely numbered and the last member of the set is or'd
with a flag indicating that it is, in fact, the last member of the set. The LDIR_Ord field is used to
make this determination. The first member of a set has an LDIR_Ord value of one. The nth long
member of the set has a value of (n OR LAST_LONG_ENTRY). Note that the LDIR_Ord field
cannot have values of 0xE5 or 0x00. These values have always been used by the file system to
indicate a "free" directory entry, or the "last" directory entry in a cluster. Values for LDIR_Ord do not
take on these two values over their range. Values for LDIR_Ord must run from 1 to (n OR
LAST_LONG_ENTRY). If they do not, the long entries are "damaged" and are treated as orphans by
the file system.

Second, an 8-bit checksum is computed on the name contained in the short directory entry at the time
the short and long directory entries are created. All 11 characters of the name in the short entry are
used in the checksum calculation. The check sum is placed in every long entry. If any of the check
sums in the set of long entries do not agree with the computed checksum of the name contained in the
short entry, then the long entries are treated as orphans. This can occur if a disk containing long and
short entries is taken to a previous version of MS-DOS/Windows and only the short name of a file or
directory with a long entries is renamed.

The algorithm, implemented in C, for computing the checksum is:

© 2000 Microsoft Corporation. All rights reserved. 27

 FAT: General Overview of On-Disk Format—Page 28

//-----------------------------------------------------------------------------
// ChkSum()
// Returns an unsigned byte checksum computed on an unsigned byte
// array. The array must be 11 bytes long and is assumed to contain
// a name stored in the format of a MS-DOS directory entry.
// Passed: pFcbName Pointer to an unsigned byte array assumed to be
// 11 bytes long.
// Returns: Sum An 8-bit unsigned checksum of the array pointed
// to by pFcbName.
//------------------------------------------------------------------------------
unsigned char ChkSum (unsigned char *pFcbName)
{
short FcbNameLen;
unsigned char Sum;

Sum = 0;
for (FcbNameLen=11; FcbNameLen!=0; FcbNameLen--) {
// NOTE: The operation is an unsigned char rotate right
Sum = ((Sum & 1) ? 0x80 : 0) + (Sum >> 1) + *pFcbName++;
}
return (Sum);
}

As a consequence of this pairing, the short directory entry serves as the structure that contains fields
like: last access date, creation time, creation date, first cluster, and size. It also holds a name that is
visible on previous versions of MS-DOS/Windows. The long directory entries are free to contain new
information and need not replicate information already available in the short entry. Principally, the
long entries contain the long name of a file. The name contained in a short entry which is associated
with a set of long entries is termed the alias name, or simply alias, of the file.

Storage of a Long-Name Within Long Directory Entries

A long name can consist of more characters than can fit in a single long directory entry. When this
occurs the name is stored in more than one long entry. In any event, the name fields themselves
within the long entries are disjoint. The following example is provided to illustrate how a long name
is stored across several long directory entries. Names are also NUL terminated and padded with
0xFFFF characters in order to detect corruption of long name fields by errant disk utilities. A name
that fits exactly in a n long directory entries (i.e. is an integer multiple of 13) is not NUL terminated
and not padded with 0xFFFFs.

Suppose a file is created with the name: "The quick brown.fox". The following example illustrates
how the name is packed into long and short directory entries. Most fields in the directory entries are
also filled in as well.

2nd long entry w n f o chk- x

42h . 0Fh 00h
sum
(and last)
0000h FFFFh FFFFh FFFFh FFFFh 0000h FFFFh FFFFh

T h e q chk- u
1st long entry 01h 0Fh 00h
sum
i c k b 0000h r o

Created
Short entry T H E Q U I ~ 1 F O X 20h NT Rsvd Time
Created Last Last Last First
Access 0000h Modified Modified File Size
Date Date Time Date Cluster

© 2000 Microsoft Corporation. All rights reserved. 28

 FAT: General Overview of On-Disk Format—Page 29

The heuristics used to "auto-generate" a short name from a long name are explained in a later section.

Name Limits and Character Sets

Short Directory Entries

Short names are limited to 8 characters followed by an optional period (.) and extension of up to 3
characters. The total path length of a short name cannot exceed 80 characters (64 char path + 3 drive
letter + 12 for 8.3 name + NUL) including the trailing NUL. The characters may be any combination
of letters, digits, or characters with code point values greater than 127. The following special
characters are also allowed:

$ % ' - _ @ ~ ` ! ( ) { } ^ # &

Names are stored in a short directory entry in the OEM code page that the system is configured for at
the time the directory entry is created. Short directory entries remain in OEM for compatibility with
previous versions of MS-DOS/Windows. OEM characters are single 8-bit characters or can be DBCS
character pairs for certain code pages.

Short names passed to the file system are always converted to upper case and their original case value
is lost. One problem that is generally true of most OEM code pages is that they map lower to upper
case extended characters in a non-unique fashion. That is, they map multiple extended characters to a
single upper case character. This creates problems because it does not preserve the information that
the extended character provides. This mapping also prevents the creation of some file names that
would normally differ, but because of the mapping to upper case they become the same file name.

Long Directory Entries

Long names are limited to 255 characters, not including the trailing NUL. The total path length of a
long name cannot exceed 260 characters, including the trailing NUL. The characters may be any
combination of those defined for short names with the addition of the period (.) character used
multiple times within the long name. A space is also a valid character in a long name as it always has
been for a short name. However, in short names it typically is not used. The following six special
characters are now allowed in a long name. They are not legal in a short name.

+ , ; = [ ]

Embedded spaces within a long name are allowed. Leading and trailing spaces in a long name are
ignored.

Leading and embedded periods are allowed in a name and are stored in the long name. Trailing
periods are ignored.

Long names are stored in long directory entries in UNICODE. UNICODE characters are 16-bit
characters. It is not be possible to store UNICODE in short directory entries since the names stored
there are 8-bit characters or DBCS characters.

Long names passed to the file system are not converted to upper case and their original case value is
preserved. UNICODE solves the case mapping problem prevalent in some OEM code pages by
always providing a translation for lower case characters to a single, unique upper case character.

© 2000 Microsoft Corporation. All rights reserved. 29

 FAT: General Overview of On-Disk Format—Page 30

Name Matching In Short & Long Names

The names contained in the set of all short directory entries are termed the "short name space". The
names contained in the set of all long directory entries are termed the "long name space". Together,
they form a single unified name space in which no duplicate names can exist. That is: any name
within a specific directory, whether it is a short name or a long name, can occur only once in the name
space. Furthermore, although the case of a name is preserved in a long name, no two names can have
the same name although the names on the media actually differ by case. That is names like "foobar"
cannot be created if there is already a short entry with a name of "FOOBAR" or a long name with a
name of "FooBar".

All types of search operations within the file system (i.e. find, open, create, delete, rename) are case-
insensitive. An open of "FOOBAR" will open either "FooBar" or "foobar" if one or the other exists.
A find using "FOOBAR" as a pattern will find the same files mentioned. The same rules are also true
for extended characters that are accented.

A short name search operation checks only the names of the short directory entries for a match. A
long name search operation checks both the long and short directory entries. As the file system
traverses a directory, it caches the long-name sub-components contained in long directory entries. As
soon as a short directory entry is encountered that is associated with the cached long name, the long
name search operation will check the cached long name first and then the short name for a match.

When a character on the media, whether it is stored in the OEM character set or in UNICODE, cannot
be translated into the appropriate character in the OEM or ANSI code page, it is always "translated" to
the "_" (underscore) character as it is returned to the user – it is NOT modified on the disk. This
character is the same in all OEM code pages and ANSI.

Naming Conventions and Long Names

An API allows the caller to specify the long name to be assigned to a file or directory. They do not
allow the caller to independently specify the short name. The reason for this prohibition is that the
short and long names are considered to be a single unified name space. As should be obvious the file
system's name space does not support duplicate names. In other words, a long name for a file may not
contain the same name, ignoring case, as the short name in a different file. This restriction is intended
to prevent confusion among users, and applications, regarding the proper name of a file or directory.
To make this restriction transparent, whenever a long name is created and the no matching long name
exists, the short name is automatically generated from the long name in such a way that it does not
collide with an existing short name.

The technique chosen to auto-generate short names from long names is modeled after Windows NT.
Auto-generated short names are composed of the basis-name and an optional numeric-tail.

The Basis-Name Generation Algorithm

The basis-name generation algorithm is outlined below. This is a sample algorithm and serves to
illustrate how short names can be auto-generated from long names. An implementation should follow
this basic sequence of steps.

1. The UNICODE name passed to the file system is converted to upper case.

2. The upper cased UNICODE name is converted to OEM.

if (the uppercased UNICODE glyph does not exist as an OEM glyph in the OEM code page)
or (the OEM glyph is invalid in an 8.3 name)
{

© 2000 Microsoft Corporation. All rights reserved. 30

 FAT: General Overview of On-Disk Format—Page 31

Replace the glyph to an OEM '_' (underscore) character.

Set a "lossy conversion" flag.
}

3. Strip all leading and embedded spaces from the long name.

4. Strip all leading periods from the long name.

5. While (not at end of the long name)

and (char is not a period)
and (total chars copied < 8)
{
Copy characters into primary portion of the basis name
}

6. Insert a dot at the end of the primary components of the basis-name iff the basis name has an
extension after the last period in the name.

7. Scan for the last embedded period in the long name.

If (the last embedded period was found)
{
While (not at end of the long name)
and (total chars copied < 3)
{
Copy characters into extension portion of the basis name
}
}

Proceed to numeric-tail generation.

The Numeric-Tail Generation Algorithm

If (a "lossy conversion" was not flagged)
and (the long name fits within the 8.3 naming conventions)
and (the basis-name does not collide with any existing short name)
{
The short name is only the basis-name without the numeric tail.
}
else
{
Insert a numeric-tail "~n" to the end of the primary name such that the value of the "~n" is
chosen so that the
name thus formed does not collide with any existing short name and that the primary name does
not exceed eight characters in length.
}

The "~n" string can range from "~1" to "~999999". The number "n" is chosen so that it is the next
number in a sequence of files with similar basis-names. For example, assume the following short
names existed: LETTER~1.DOC and LETTER~2.DOC. As expected, the next auto-generated name
of name of this type would be LETTER~3.DOC. Assume the following short names existed:
LETTER~1.DOC, LETTER~3.DOC. Again, the next auto-generated name of name of this type
would be LETTER~2.DOC. However, one absolutely cannot count on this behavior. In a directory
with a very large mix of names of this type, the selection algorithm is optimized for speed and may
select another "n" based on the characteristics of short names that end in "~n" and have similar leading
name patterns.

© 2000 Microsoft Corporation. All rights reserved. 31

 FAT: General Overview of On-Disk Format—Page 32

Effect of Long Directory Entries on Down Level Versions of FAT

The support of long names is most important on the hard disk, however it will be supported on
removable media as well. The implementation provides support for long names without breaking
compatibility with the existing FAT format. A disk can be read by a down level system without any
compatibility problems. An existing disk does not go through a conversion process before it can start
using long names. All of the current files remain unmodified. The long name directory entries are
added when a long name is created. The addition of a long name to an existing file may require the
8.3 directory entry to be moved if the required adjacent directory entries are not available.

The long name entries are as hidden as hidden or system files are on a down level system. This is
enough to keep the casual user from causing problems. The user can copy the files off using the 8.3
name, and put new files on without any side effects

The interesting part of this is what happens when the disk is taken to a down level FAT system and the
directory is changed. This can affect the long name entries since the down level system ignores these
long names and will not ensure they are properly associated with the 8.3 names.

A down level system will only see the long name entries when searching for a label. On a down level
system, the volume label will be incorrectly reported if the true volume label does not come before all
of the long name entries in the root directory. This is because the long name entries also have the
volume label bit set. This is unfortunate, but is not a critical problem.

If an attempt is made to remove the volume label, one of the long name directory entries may be
deleted. This would be a rare occurrence. It is easily detected on an aware system. The long name
entry will no longer be a valid file entry, since one or more of the long entries is marked as deleted. If
the deleted entry is reused, then the attribute byte will not have the proper value for a long name entry.

If a file is renamed on a down level system, then only the short name will be renamed. The long name
will not be affected. Since the long and short names must be kept consistent across the name space, it
is desirable to have the long name become invalid as a result of this rename. The checksum of the 8.3
name that is kept in the long name directory provides the ability to detect this type of change. This
checksum will be checked to validate the long name before it is used. Rename will cause problems
only if the renamed 8.3 file name happens to have the same checksum. The checksum algorithm
chosen has a relatively flat distribution across the short name space.

This rename of the 8.3 name must also not conflict with any of the long names. Otherwise a down
level system could create a short name in one file that matches a long name, when case is ignored, in a
different file. To prevent this, the automatic creation of an 8.3 name from a long name, that has an 8.3
format, will directly map the long name to the 8.3 name by converting the characters to upper case.

If the file is deleted, then the long name is simply orphaned. If a new file is created, the long name
may be incorrectly associated with the new file name. As in the case of a rename the checksum of the
8.3 name will help prevent this incorrect association.

Validating The Contents of a Directory

These guidelines are provided so that disk maintenance utilities can verify individual directory entries
for 'correctness' while maintaining compatibility with future enhancements to the directory structure.

1. DO NOT look at the content of directory entry fields marked 'reserved' and assume that, if they
are any value other than zero, that they are 'bad'.
2. DO NOT reset the content of directory entry fields marked reserved to zero when they contain
non-zero values (under the assumption that they are "bad"). Directory entry fields are designated

© 2000 Microsoft Corporation. All rights reserved. 32

 FAT: General Overview of On-Disk Format—Page 33

reserved, rather than must-be-zero. They should be ignored by your application.. These fields are
intended for future extensions of the file system. By ignoring them an utility can continue to run
on future versions of the operating system.
3. DO use the A_LONG attribute first when determining whether a directory entry is a long
directory entry or a short directory entry. The following algorithm is the correct algorithm for
making this determination:

if (((LDIR_attr & ATTR_LONG_NAME_MASK) == ATTR_LONG_NAME) && (LDIR_Ord != 0xE5))

{
/* Found an active long name sub-component. */
}

4. DO use bits 4 and 3 of a short entry together when determining what type of short directory entry
is being inspected. The following algorithm is the correct algorithm for making this
determination:

if (((LDIR_attr & ATTR_LONG_NAME_MASK) != ATTR_LONG_NAME) && (LDIR_Ord != 0xE5))

{
if ((DIR_Attr & (ATTR_DIRECTORY | ATTR_VOLUME_ID)) == 0x00)
/* Found a file. */
else if ((DIR_Attr & (ATTR_DIRECTORY | ATTR_VOLUME_ID)) == ATTR_DIRECTORY)
/* Found a directory. */
else if ((DIR_Attr & (ATTR_DIRECTORY | ATTR_VOLUME_ID)) == ATTR_VOLUME_ID)
/* Found a volume label. */
else
/* Found an invalid directory entry. */
}

5. DO NOT assume that a non-zero value in the "type" field indicates a bad directory entry. Do not
force the "type" field to zero.
6. Use the "checksum" field as a value to validate the directory entry. The "first cluster" field is
currently being set to zero, though this might change in future.

Other Notes Relating to FAT Directories

• Long File Name directory entries are identical on all FAT types. See the preceeding sections for
details.

• DIR_FileSize is a 32-bit field. For FAT32 volumes, your FAT file system driver must not allow a
cluster chain to be created that is longer than 0x100000000 bytes, and the last byte of the last
cluster in a chain that long cannot be allocated to the file. This must be done so that no file has a
file size > 0xFFFFFFFF bytes. This is a fundamental limit of all FAT file systems. The maximum
allowed file size on a FAT volume is 0xFFFFFFFF (4,294,967,295) bytes.

• Similarly, a FAT file system driver must not allow a directory (a file that is actually a container for
other files) to be larger than 65,536 * 32 (2,097,152) bytes.

NOTE: This limit does not apply to the number of files in the directory. This limit is on the size of
the directory itself and has nothing to do with the content of the directory. There are two reasons
for this limit:

1. Because FAT directories are not sorted or indexed, it is a bad idea to create huge directories;
otherwise, operations like creating a new entry (which requires every allocated directory entry
to be checked to verify that the name doesn’t already exist in the directory) become very slow.

© 2000 Microsoft Corporation. All rights reserved. 33

 FAT: General Overview of On-Disk Format—Page 34

2. There are many FAT file system drivers and disk utilities, including Microsoft’s, that expect to
be able to count the entries in a directory using a 16-bit WORD variable. For this reason,
directories cannot have more than 16-bits worth of entries.

© 2000 Microsoft Corporation. All rights reserved. 34