2.11BSD/src/sbin/mkfs/mkfs.8

.\" Copyright (c) 1980 Regents of the University of California.
.\" All rights reserved.  The Berkeley software License Agreement
.\" specifies the terms and conditions for redistribution.
.\"
.\"	@(#)mkfs.8	2.2 (2.11BSD) 1996/11/16
.\"
.TH MKFS 8 "November 16, 1996"
.UC 2
.SH NAME
mkfs \- construct a file system
.SH SYNOPSIS
.B /sbin/mkfs
[
.B \-i bytes
]
[
.B \-s size
]
[
.B \-m gap
]
[
.B \-n modulus
]
.B special
.SH DESCRIPTION
.B N.B.:
file systems are normally created with the
.IR newfs (8)
command.
.PP
.I Mkfs
constructs a file system
by writing on the special file
.IR special .
The size of the filesystem in logical blocks is specified by the 
\fB\-s\fP
.I size
option.
Logical blocks are 1K (2 sectors) under 2.11BSD.
.PP
.in +0.5i
\fBNOTE:\fP The 
.BR newfs (8)
program's \fB\-s\fP option is in units of \fBsectors\fP.    
.BR Newfs (8)
converts this to filesystem (logical) blocks for
.BR mkfs (8) .
.br
.in -0.5i
.PP
The number of inodes is calculated based on the argument
.I bytes
to the 
.B \-i
option.  The default is 4096.  If more inodes are desired in a filesystem
(there is an absolute maximum of 65500) then a lower value for \fIbytes\fP
should be used, perhaps 3072 or even 2048.
.PP
The flags 
.B \-m
.I gap
and
.B \-n
.I modulus
determine the block interleaving
of the freelist that will be constructed,
where \fIgap\fP is the distance between successive 1024-byte blocks,
and \fImodulus\fP is the number of blocks before the pattern repeats,
typically one cylinder.
The optimal values for
these parameters vary with the speed and geometry of the disk,
as well as the speed of the processor.  
.BR Newfs (8)
will calculate the correct values in almost all cases from the disklabel.
.SH "SEE ALSO"
fs(5),
dir(5),
disklabel(8),
fsck(8),
mkproto(8)
newfs(8)
.SH BUGS
The
.I lost+found
directory is created but the boot block is left uninitialized (see
.BR disklabel (8).)