About Kernel Documentation Linux Kernel Contact Linux Resources Linux Blog

Documentation / filesystems / vfat.txt

Custom Search

Based on kernel version 3.15.4. Page generated on 2014-07-07 09:03 EST.

2	----------------------------------------------------------------------
3	To use the vfat filesystem, use the filesystem type 'vfat'.  i.e.
4	  mount -t vfat /dev/fd0 /mnt
6	No special partition formatter is required.  mkdosfs will work fine
7	if you want to format from within Linux.
10	----------------------------------------------------------------------
11	uid=###       -- Set the owner of all files on this filesystem.
12			 The default is the uid of current process.
14	gid=###       -- Set the group of all files on this filesystem.
15			 The default is the gid of current process.
17	umask=###     -- The permission mask (for files and directories, see umask(1)).
18	                 The default is the umask of current process.
20	dmask=###     -- The permission mask for the directory.
21	                 The default is the umask of current process.
23	fmask=###     -- The permission mask for files.
24	                 The default is the umask of current process.
26	allow_utime=### -- This option controls the permission check of mtime/atime.
28	                  20 - If current process is in group of file's group ID,
29	                       you can change timestamp.
30	                   2 - Other users can change timestamp.
32	                 The default is set from `dmask' option. (If the directory is
33	                 writable, utime(2) is also allowed. I.e. ~dmask & 022)
35	                 Normally utime(2) checks current process is owner of
36	                 the file, or it has CAP_FOWNER capability.  But FAT
37	                 filesystem doesn't have uid/gid on disk, so normal
38	                 check is too unflexible. With this option you can
39	                 relax it.
41	codepage=###  -- Sets the codepage number for converting to shortname
42			 characters on FAT filesystem.
43			 By default, FAT_DEFAULT_CODEPAGE setting is used.
45	iocharset=<name> -- Character set to use for converting between the
46			 encoding is used for user visible filename and 16 bit
47			 Unicode characters. Long filenames are stored on disk
48			 in Unicode format, but Unix for the most part doesn't
49			 know how to deal with Unicode.
50			 By default, FAT_DEFAULT_IOCHARSET setting is used.
52			 There is also an option of doing UTF-8 translations
53			 with the utf8 option.
55			 NOTE: "iocharset=utf8" is not recommended. If unsure,
56			 you should consider the following option instead.
58	utf8=<bool>   -- UTF-8 is the filesystem safe version of Unicode that
59			 is used by the console.  It can be enabled for the
60			 filesystem with this option. If 'uni_xlate' gets set,
61			 UTF-8 gets disabled.
63	uni_xlate=<bool> -- Translate unhandled Unicode characters to special
64			 escaped sequences.  This would let you backup and
65			 restore filenames that are created with any Unicode
66			 characters.  Until Linux supports Unicode for real,
67			 this gives you an alternative.  Without this option,
68			 a '?' is used when no translation is possible.  The
69			 escape character is ':' because it is otherwise
70			 illegal on the vfat filesystem.  The escape sequence
71			 that gets used is ':' and the four digits of hexadecimal
72			 unicode.
74	nonumtail=<bool> -- When creating 8.3 aliases, normally the alias will
75	                 end in '~1' or tilde followed by some number.  If this
76	                 option is set, then if the filename is 
77	                 "longfilename.txt" and "longfile.txt" does not
78	                 currently exist in the directory, 'longfile.txt' will
79	                 be the short alias instead of 'longfi~1.txt'. 
81	usefree       -- Use the "free clusters" value stored on FSINFO. It'll
82	                 be used to determine number of free clusters without
83	                 scanning disk. But it's not used by default, because
84	                 recent Windows don't update it correctly in some
85	                 case. If you are sure the "free clusters" on FSINFO is
86	                 correct, by this option you can avoid scanning disk.
88	quiet         -- Stops printing certain warning messages.
90	check=s|r|n   -- Case sensitivity checking setting.
91	                 s: strict, case sensitive
92	                 r: relaxed, case insensitive
93	                 n: normal, default setting, currently case insensitive
95	nocase        -- This was deprecated for vfat. Use shortname=win95 instead.
97	shortname=lower|win95|winnt|mixed
98		      -- Shortname display/create setting.
99			 lower: convert to lowercase for display,
100				emulate the Windows 95 rule for create.
101			 win95: emulate the Windows 95 rule for display/create.
102			 winnt: emulate the Windows NT rule for display/create.
103			 mixed: emulate the Windows NT rule for display,
104				emulate the Windows 95 rule for create.
105			 Default setting is `mixed'.
107	tz=UTC        -- Interpret timestamps as UTC rather than local time.
108	                 This option disables the conversion of timestamps
109	                 between local time (as used by Windows on FAT) and UTC
110	                 (which Linux uses internally).  This is particularly
111	                 useful when mounting devices (like digital cameras)
112	                 that are set to UTC in order to avoid the pitfalls of
113	                 local time.
114	time_offset=minutes
115		      -- Set offset for conversion of timestamps from local time
116			 used by FAT to UTC. I.e. <minutes> minutes will be subtracted
117			 from each timestamp to convert it to UTC used internally by
118			 Linux. This is useful when time zone set in sys_tz is
119			 not the time zone used by the filesystem. Note that this
120			 option still does not provide correct time stamps in all
121			 cases in presence of DST - time stamps in a different DST
122			 setting will be off by one hour.
124	showexec      -- If set, the execute permission bits of the file will be
125			 allowed only if the extension part of the name is .EXE,
126			 .COM, or .BAT. Not set by default.
128	debug         -- Can be set, but unused by the current implementation.
130	sys_immutable -- If set, ATTR_SYS attribute on FAT is handled as
131			 IMMUTABLE flag on Linux. Not set by default.
133	flush         -- If set, the filesystem will try to flush to disk more
134			 early than normal. Not set by default.
136	rodir	      -- FAT has the ATTR_RO (read-only) attribute. On Windows,
137			 the ATTR_RO of the directory will just be ignored,
138			 and is used only by applications as a flag (e.g. it's set
139			 for the customized folder).
141			 If you want to use ATTR_RO as read-only flag even for
142			 the directory, set this option.
144	errors=panic|continue|remount-ro
145		      -- specify FAT behavior on critical errors: panic, continue
146			 without doing anything or remount the partition in
147			 read-only mode (default behavior).
149	discard       -- If set, issues discard/TRIM commands to the block
150			 device when blocks are freed. This is useful for SSD devices
151			 and sparse/thinly-provisoned LUNs.
153	nfs=stale_rw|nostale_ro
154			Enable this only if you want to export the FAT filesystem
155			over NFS.
157			stale_rw: This option maintains an index (cache) of directory
158			inodes by i_logstart which is used by the nfs-related code to
159			improve look-ups. Full file operations (read/write) over NFS is
160			supported but with cache eviction at NFS server, this could
161			result in ESTALE issues.
163			nostale_ro: This option bases the inode number and filehandle
164			on the on-disk location of a file in the MS-DOS directory entry.
165			This ensures that ESTALE will not be returned after a file is
166			evicted from the inode cache. However, it means that operations
167			such as rename, create and unlink could cause filehandles that
168			previously pointed at one file to point at a different file,
169			potentially causing data corruption. For this reason, this
170			option also mounts the filesystem readonly.
172			To maintain backward compatibility, '-o nfs' is also accepted,
173			defaulting to stale_rw
176	<bool>: 0,1,yes,no,true,false
178	TODO
179	----------------------------------------------------------------------
180	* Need to get rid of the raw scanning stuff.  Instead, always use
181	  a get next directory entry approach.  The only thing left that uses
182	  raw scanning is the directory renaming code.
186	----------------------------------------------------------------------
187	* vfat_valid_longname does not properly checked reserved names.
188	* When a volume name is the same as a directory name in the root
189	  directory of the filesystem, the directory name sometimes shows
190	  up as an empty file.
191	* autoconv option does not work correctly.
194	----------------------------------------------------------------------
195	If you have trouble with the VFAT filesystem, mail bug reports to
196	chaffee@bmrc.cs.berkeley.edu.  Please specify the filename
197	and the operation that gave you trouble.
200	----------------------------------------------------------------------
201	If you plan to make any modifications to the vfat filesystem, please
202	get the test suite that comes with the vfat distribution at
204	  http://web.archive.org/web/*/http://bmrc.berkeley.edu/
205	  people/chaffee/vfat.html
207	This tests quite a few parts of the vfat filesystem and additional
208	tests for new features or untested features would be appreciated.
211	----------------------------------------------------------------------
212	(This documentation was provided by Galen C. Hunt <gchunt@cs.rochester.edu>
213	 and lightly annotated by Gordon Chaffee).
215	This document presents a very rough, technical overview of my
216	knowledge of the extended FAT file system used in Windows NT 3.5 and
217	Windows 95.  I don't guarantee that any of the following is correct,
218	but it appears to be so.
220	The extended FAT file system is almost identical to the FAT
221	file system used in DOS versions up to and including 6.223410239847
222	:-).  The significant change has been the addition of long file names.
223	These names support up to 255 characters including spaces and lower
224	case characters as opposed to the traditional 8.3 short names.
226	Here is the description of the traditional FAT entry in the current
227	Windows 95 filesystem:
229	        struct directory { // Short 8.3 names 
230	                unsigned char name[8];          // file name 
231	                unsigned char ext[3];           // file extension 
232	                unsigned char attr;             // attribute byte 
233			unsigned char lcase;		// Case for base and extension
234			unsigned char ctime_ms;		// Creation time, milliseconds
235			unsigned char ctime[2];		// Creation time
236			unsigned char cdate[2];		// Creation date
237			unsigned char adate[2];		// Last access date
238			unsigned char reserved[2];	// reserved values (ignored) 
239	                unsigned char time[2];          // time stamp 
240	                unsigned char date[2];          // date stamp 
241	                unsigned char start[2];         // starting cluster number 
242	                unsigned char size[4];          // size of the file 
243	        };
245	The lcase field specifies if the base and/or the extension of an 8.3
246	name should be capitalized.  This field does not seem to be used by
247	Windows 95 but it is used by Windows NT.  The case of filenames is not
248	completely compatible from Windows NT to Windows 95.  It is not completely
249	compatible in the reverse direction, however.  Filenames that fit in
250	the 8.3 namespace and are written on Windows NT to be lowercase will
251	show up as uppercase on Windows 95.
253	Note that the "start" and "size" values are actually little
254	endian integer values.  The descriptions of the fields in this
255	structure are public knowledge and can be found elsewhere.
257	With the extended FAT system, Microsoft has inserted extra
258	directory entries for any files with extended names.  (Any name which
259	legally fits within the old 8.3 encoding scheme does not have extra
260	entries.)  I call these extra entries slots.  Basically, a slot is a
261	specially formatted directory entry which holds up to 13 characters of
262	a file's extended name.  Think of slots as additional labeling for the
263	directory entry of the file to which they correspond.  Microsoft
264	prefers to refer to the 8.3 entry for a file as its alias and the
265	extended slot directory entries as the file name. 
267	The C structure for a slot directory entry follows:
269	        struct slot { // Up to 13 characters of a long name 
270	                unsigned char id;               // sequence number for slot 
271	                unsigned char name0_4[10];      // first 5 characters in name 
272	                unsigned char attr;             // attribute byte
273	                unsigned char reserved;         // always 0 
274	                unsigned char alias_checksum;   // checksum for 8.3 alias 
275	                unsigned char name5_10[12];     // 6 more characters in name
276	                unsigned char start[2];         // starting cluster number
277	                unsigned char name11_12[4];     // last 2 characters in name
278	        };
280	If the layout of the slots looks a little odd, it's only
281	because of Microsoft's efforts to maintain compatibility with old
282	software.  The slots must be disguised to prevent old software from
283	panicking.  To this end, a number of measures are taken:
285	        1) The attribute byte for a slot directory entry is always set
286	           to 0x0f.  This corresponds to an old directory entry with
287	           attributes of "hidden", "system", "read-only", and "volume
288	           label".  Most old software will ignore any directory
289	           entries with the "volume label" bit set.  Real volume label
290	           entries don't have the other three bits set.
292	        2) The starting cluster is always set to 0, an impossible
293	           value for a DOS file.
295	Because the extended FAT system is backward compatible, it is
296	possible for old software to modify directory entries.  Measures must
297	be taken to ensure the validity of slots.  An extended FAT system can
298	verify that a slot does in fact belong to an 8.3 directory entry by
299	the following:
301	        1) Positioning.  Slots for a file always immediately proceed
302	           their corresponding 8.3 directory entry.  In addition, each
303	           slot has an id which marks its order in the extended file
304	           name.  Here is a very abbreviated view of an 8.3 directory
305	           entry and its corresponding long name slots for the file
306	           "My Big File.Extension which is long":
308	                <proceeding files...>
309	                <slot #3, id = 0x43, characters = "h is long">
310	                <slot #2, id = 0x02, characters = "xtension whic">
311	                <slot #1, id = 0x01, characters = "My Big File.E">
312	                <directory entry, name = "MYBIGFIL.EXT">
314	           Note that the slots are stored from last to first.  Slots
315	           are numbered from 1 to N.  The Nth slot is or'ed with 0x40
316	           to mark it as the last one.
318	        2) Checksum.  Each slot has an "alias_checksum" value.  The
319	           checksum is calculated from the 8.3 name using the
320	           following algorithm:
322	                for (sum = i = 0; i < 11; i++) {
323	                        sum = (((sum&1)<<7)|((sum&0xfe)>>1)) + name[i]
324	                }
326		3) If there is free space in the final slot, a Unicode NULL (0x0000) 
327		   is stored after the final character.  After that, all unused 
328		   characters in the final slot are set to Unicode 0xFFFF.
330	Finally, note that the extended name is stored in Unicode.  Each Unicode
331	character takes two bytes.
Hide Line Numbers
About Kernel Documentation Linux Kernel Contact Linux Resources Linux Blog

Information is copyright its respective author. All material is available from the Linux Kernel Source distributed under a GPL License. This page is provided as a free service by mjmwired.net.