mkfs(1m)
_________________________________________________________________
mkfs, newfs
create a file system
_________________________________________________________________
SYNTAX
/etc/mkfs special [-m freespace] [-r regionsize] [-i
inodedensity]
[-s dataelementlog] [-x indexelementlog] [-S
dirdataelementlog]
[-X dirindexelementlog] [-e firstanniversarysize]
[-E secondanniversarysize] [ {proto, blocks[:inodes]} [gap
blockspercyl]]
DESCRIPTION
Mkfs creates an empty file system on a logical or physical disk.
The argument <special> must be a block-special or character-
special device node, such as those nodes found in /dev/dsk or
/dev/rdsk. The file system will span the entire disk; to create
a file system of a particular size, first create a logical disk
of that size with diskman(1m) and then run mkfs on that disk.
Most invocations of mkfs will not need to alter the defaults, so
no option arguments need to be specified:
# /etc/mkfs /dev/rdsk/<special>
Newfs is identical to mkfs and is retained for Berkeley
compatibility.
Arguments are:
special
This is the name of the disk upon which a file system is to
be created. Special must be the pathname of a writable
character-special or block-special file.
-m freespace: The minimum percentage of free space the file
system must have. If the file system's free space drops
below this level, only a superuser can allocate more space.
The value for free space must be an integer in the range 0
to 99, inclusive. The default value is 10%.
-r regionsize: Determines how many blocks each Disk
Allocation Region (DAR) in the file system will occupy
(including the bitmap, inode table, and data blocks). This
number must be an integer greater than or equal to 1000; the
default value is based on the size of the file system. The
DG/UX 4.00 Page 1
Licensed material--property of copyright holder(s)
mkfs(1m)
last DAR created may be smaller than all others due to the
target logical disk being an uneven multiple of DAR size.
-i inodedensity: Determines how many inode slots (potential
files) the file system will have. The value specified is
the ratio of usable data bytes in the logical disk to the
number of inodes; the default is 3500. Any non-negative
integer may be specified, but the actual density will be
rounded up to the next integral multiple of 4 inodes per
DAR. The maximum possible number of inodes occurs when
every usable block of the DAR is occupied by inode slots.
-s dataelementlog: Determines the default data element size
of files to be created in the new file system. The value
specified is the element size in disk blocks, expressed as a
base 2 logarthm. This number must be an integer from 0 to
31, inclusive. The default value is 4 (meaning data
elements of 16 blocks).
-x indexelementlog: Determines the default index element
size of files to be created in the new file system. The
value specified is the element size in disk blocks,
expressed as a base 2 logarithm. This number must be an
integer from 0 to 15, inclusive; the default value is 0
(meaning index elements of 1 block).
-S dirdataelementlog: Determines the default data element
size of directories to be created in the new file system.
The value specified is the element size in disk blocks,
expressed as a base 2 logarithm. This number must be an
integer from 0 to 31, inclusive; the default value is 4
(meaning data elements of 16 blocks).
-X dirindexelementlog: Determines the default index element
size of directories to be created in the new file system.
The value specified is the element size in disk blocks,
expressed as a base 2 logarithm. This number must be an
integer from 0 to 15, inclusive; the default value is 0
(meaning index elements of 1 block).
-e firstanniversarysize: Determines the maximum number of
blocks a file can allocate in its initial disk allocation
region before subsequent allocation requests are redirected
to a different region. This number must be a positive
integer; the default is determined based on the size of the
disk allocation region.
-E secondanniversarysize: Determines the maximum number of
blocks a file can allocate in any noninitial disk allocation
region before subsequent allocation requests are redirected
to a different region. This number must be a positive
DG/UX 4.00 Page 2
Licensed material--property of copyright holder(s)
mkfs(1m)
integer greater than first_anniversary_size; the default is
determined based on the size of the disk allocation region.
proto
If the argument following special is a non-numeric name of a
file that can be opened, it is taken as the pathname of a
prototype file.
blocks
If the argument following special is numeric, it is
interpreted as a decimal number specifying the size of the
file system in 512-byte disk blocks. However, DG/UX file
systems must occupy the entire logical disk, so if this
number is not equal to the disk size, mkfs will fail.
inodes
If this argument is specified, it is taken as the number of
inodes to be created in the file system. However, the
specified number will be rounded up so that each DAR is
given an equal number (which is itself a multiple of 4) of
inodes. The same purpose can be achieved through use of the
-i option.
gap If this argument is specified, it is completely ignored.
Under System V it is used to allow for characteristics of
the target physical disk, a purpose that is irrelevant under
the DG/UX system.
blockspercyl
If this argument is specified, it is completely ignored.
Under System V it is used to allow for characteristics of
the target physical disk, a purpose that is irrelevant under
the DG/UX system.
Prototype File Format
The prototype file format is as follows. The file contains
tokens separated by spaces or new lines. The first token is the
name of the bootstrap program; this is completely ignored since
mkfs does not need to install bootstraps.
The second token is the size of the file system in disk blocks;
it is subject to exactly the same constraints as the blocks
argument.
The third token is the number of inodes to be created in the file
system; it is subject to exactly the same constraints as the
inodes argument.
The next set of tokens comprise the specification for the root
DG/UX 4.00 Page 3
Licensed material--property of copyright holder(s)
mkfs(1m)
directory: the mode, the user id, the group id and the initial
contents. The syntax of the contents field depends on the file
mode. The mode token for a file is a six-character string. The
first character specifies the file type using the same rules as
ls(1). The second character is either "u" or "-" to specify
setuid or not. The third character is either "g" or "-" to
specify setgid or not. The rest of the mode is a 3 digit octal
number in the same manner as ls(1). Two decimal number tokens
follow the mode; they specify the user and group ids of the
file's owner.
If the file is an ordinary file, the next token is a pathname
from which the contents and size are copied. If the file is a
block-special or character special file, two decimal tokens
follow which give the file's major and minor device numbers. If
the file is a directory, mkfs makes the entries specified. This
specification may be recursive; each directory is terminated with
the token "$".
DIAGNOSTICS
Mkfs will have no output except for diagnostic output in the case
of errors. Mkfs will return an exit status of 0 if and only if
the specified file system was successfully created. Otherwise,
mkfs will return 1.
SEE ALSO
fsck(1m), tunefs(1m), diskman(1m), fs(4)
DG/UX 4.00 Page 4
Licensed material--property of copyright holder(s)