xref: /netbsd-src/share/man/man4/options.4 (revision 27527e67bbdf8d9ec84fd58803048ed6d181ece2)
1.\"	$NetBSD: options.4,v 1.310 2006/02/04 18:50:39 wiz Exp $
2.\"
3.\" Copyright (c) 1996
4.\" 	Perry E. Metzger.  All rights reserved.
5.\"
6.\" Redistribution and use in source and binary forms, with or without
7.\" modification, are permitted provided that the following conditions
8.\" are met:
9.\" 1. Redistributions of source code must retain the above copyright
10.\"    notice, this list of conditions and the following disclaimer.
11.\" 2. Redistributions in binary form must reproduce the above copyright
12.\"    notice, this list of conditions and the following disclaimer in the
13.\"    documentation and/or other materials provided with the distribution.
14.\" 3. All advertising materials mentioning features or use of this software
15.\"    must display the following acknowledgment:
16.\"	This product includes software developed for the NetBSD Project
17.\"	by Perry E. Metzger.
18.\" 4. The name of the author may not be used to endorse or promote products
19.\"    derived from this software without specific prior written permission.
20.\"
21.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
22.\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
23.\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
24.\" IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
25.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
26.\" NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
27.\" DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
28.\" THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
29.\" (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
30.\" THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
31.\"
32.\"
33.Dd February 3, 2006
34.Os
35.Dt OPTIONS 4
36.Sh NAME
37.Nm options
38.Nd Miscellaneous kernel configuration options
39.Sh SYNOPSIS
40.Cd cinclude ...
41.Cd config ...
42.Cd [no] file-system ...
43.Cd ident ...
44.Cd include ...
45.Cd [no] makeoptions ...
46.Cd maxusers ...
47.Cd [no] options ...
48.Cd [no] pseudo-device ...
49.Sh DESCRIPTION
50This manual page describes a number of miscellaneous kernel
51configuration options that may be specified in a kernel config
52file.
53See
54.Xr config 1
55for information on how to configure and build kernels.
56.Pp
57The
58.Ar no
59form removes a previously specified option.
60.Ss Keywords
61The following keywords are recognized in a kernel configuration file:
62.Bl -ohang
63.It Sy cinclude Xo
64.Qq Ar filename
65.Xc
66Conditionally includes another kernel configuration file whose name is
67.Ar filename ,
68which may be double-quoted and may be an explicit path or relative to
69the kernel source directory.
70Failure to open the named file is ignored.
71.It Sy config Xo
72.Ar exec_name
73.No root on
74.Ar rootdev
75.Op type Ar fstype
76.Op dumps on Ar dumpdev
77.Xc
78Defines a configuration whose kernel executable is named
79.Ar exec_name ,
80normally
81.Dq netbsd ,
82with its root file system of type
83.Ar fstype
84on the device
85.Ar rootdev ,
86and optionally specifying the location of kernel core dumps on the device
87.Ar dumpdev .
88.Ar dev
89or
90.Ar dumpdev
91and
92.Ar fstype
93may be specified as
94.Dq \&? ,
95which is a wild card.
96The root
97.Ar fstype
98and
99.Ar dumpdev
100are optional and assumed to be wild carded if they are not specified.
101.It Ar device_instance Xo
102at
103.Ar attachment
104.Op Ar locators value Op ...
105.Op flags Ar value
106.Xc
107Define an instance of the device driver
108.Ar device_instance
109that attaches to the bus or device named
110.Ar attachment .
111An
112.Ar attachment
113may require additional information on where the device can be found, such
114as an address, channel, function, offset, and/or slot, referred to as
115.Ar locators ,
116whose
117.Ar value
118often may be a wild card,
119.Dq \&? .
120Some device drivers have one or more
121.Ar flags
122that can be adjusted to affect the way they operate.
123.It Sy file-system Xo
124.Ar fs_name Op , Ar fs_name Op ...
125.Xc
126Include support for the file-system
127.Ar fs_name .
128.It Sy ident Xo
129.Qq Ar string
130.Xc
131Sets the kernel identification string to
132.Ar string .
133.It Sy include Xo
134.Qq Ar filename
135.Xc
136Functions the same as
137.Ar cinclude ,
138except failure to open
139.Ar filename
140produces a fatal error.
141.It Sy options Xo
142.Ar option_name Op , Ar option_name=value Op ...
143.Xc
144Specifies (or sets) the option, or comma-separated list of options,
145.Ar option_name .
146Some options expect to be assigned a value, which may be an integer,
147a double-quoted word, a bare word, or an empty string
148.Pq Qq .
149Note that those are eventually handled by the C compiler, so the rules
150of that language apply.
151.Pp
152.Em Note :
153Options that are not defined by device definition files are passed to
154the compile process as
155.Fl D
156flags to the C compiler.
157.It Sy makeoptions Xo
158.Ar name=value
159.Xc
160Defines a
161.Xr make 1
162macro
163.Ar name
164with the value
165.Ar value
166in the kernel Makefile.
167.It Sy maxusers Xo
168.Ar integer
169.Xc
170Set the maxusers variable in the kernel.
171.It Sy no Xo
172.Ar keyword Ar name Op Ar arguments Op ...
173.Xc
174For the
175.Xr config 1
176.Ar keywords
177file-system, makeoptions, options, and pseudo-device,
178.Ar no
179removes the file-system, makeoption, options, or pseudo-device,
180.Ar name .
181This is useful when a kernel configuration file includes another which
182has undesired options.
183.Pp
184For example, a local configuration file that wanted the kitchen sink, but
185not COMPAT_09 or bridging, might be:
186.Bd -literal -offset indent
187include "arch/i386/conf/GENERIC"
188no options COMPAT_09
189no pseudo-device bridge
190.Ed
191.It Sy pseudo-device Xo
192.Ar name
193.Op Ar N
194.Xc
195Includes support for the pseudo-device
196.Ar name .
197Some pseudo-devices can have multiple or
198.Ar N
199instances.
200.El
201.Ss Compatibility Options
202.Bl -ohang
203.It Cd options COMPAT_09
204Enable binary compatibility with
205.Nx 0.9 .
206This enables support for
20716-bit user, group, and process IDs (following revisions support
20832-bit identifiers).
209It also allows the use of the deprecated
210.Xr getdomainname 3 ,
211.Xr setdomainname 3 ,
212and
213.Xr uname 3
214syscalls.
215This option also allows using numeric file system identifiers rather
216than strings.
217Post
218.Nx 0.9
219versions use string identifiers.
220.It Cd options COMPAT_10
221Enable binary compatibility with
222.Nx 1.0 .
223This option allows the use of the file system name of
224.Dq ufs
225as an alias for
226.Dq ffs .
227The name
228.Dq ffs
229should be used post 1.0 in
230.Pa /etc/fstab
231and other files.
232It also adds old syscalls for the
233.At V
234shared memory interface.
235This was changed post 1.0 to work on 64-bit architectures.
236This option also enables
237.Dq sgtty
238compatibility, without which programs using the old interface produce
239an
240.Dq inappropriate ioctl
241error, and
242.Pa /dev/io
243only works when this option is set in the kernel,
244see
245.Xr io 4
246on ports that support it.
247.It Cd options COMPAT_11
248Enable binary compatibility with
249.Nx 1.1 .
250This allows binaries running on the i386 port to gain direct access to
251the io ports by opening
252.Pa /dev/io
253read/write.
254This functionality was replaced by
255.Xr i386_iopl 2
256post 1.1.
257On the
258.Tn Atari
259port, the location of the disk label was moved after 1.1.
260When the
261.Em COMPAT_11
262option is set, the kernel will read (pre) 1.1 style disk labels as a
263last resort.
264When a disklabel is re-written, the old style label will be replaced
265with a post 1.1 style label.
266This also enables the
267.Em EXEC_ELF_NOTELESS
268option.
269.It Cd options COMPAT_12
270Enable binary compatibility with
271.Nx 1.2 .
272This allows the use of old syscalls for
273.Fn reboot
274and
275.Fn swapon .
276The syscall numbers were changed post 1.2 to add functionality to the
277.Xr reboot 2
278syscall, and the new
279.Xr swapctl 2
280interface was introduced.
281This also enables the
282.Em EXEC_ELF_NOTELESS
283option.
284.It Cd options COMPAT_13
285Enable binary compatibility with
286.Nx 1.3 .
287This allows the use of old syscalls for
288.Fn sigaltstack ,
289and also enables the old
290.Xr swapctl 2
291command
292.Dv SWAP_STATS
293(now called
294.Dv SWAP_OSTATS ) ,
295which does not include the
296.Fa se_path
297member of
298.Va struct swapent .
299.It Cd options COMPAT_14
300Enable binary compatibility with
301.Nx 1.4 .
302This allows some old
303.Xr ioctl 2
304on
305.Xr wscons 4
306to be performed, and allows the
307.Dv NFSSVC_BIOD
308mode of the
309.Xr nfssvc 2
310system call to be used for compatibility with the deprecated nfsiod program.
311.It Cd options COMPAT_16
312Enable binary compatibility with
313.Nx 1.6 .
314This allows the use of old signal trampoline code which has been deprecated
315with the addition of
316.Xr siginfo 2 .
317.It Cd options COMPAT_20
318Enable binary compatibility with
319.Nx 2.0 .
320This allows the use of old syscalls for
321.Fn statfs
322and
323.Fn fstatfs ,
324which has been deprecated with the addition of the
325.Xr statvfs 2
326family of system calls.
327.It Cd options COMPAT_43
328Enables compatibility with
329.Bx 4.3 .
330This adds an old syscall for
331.Xr lseek 2 .
332It also adds the ioctls for
333.Dv TIOCGETP
334and
335.Dv TIOCSETP .
336The return values for
337.Xr getpid 2 ,
338.Xr getgid 2 ,
339and
340.Xr getuid 2
341syscalls are modified as well, to return the parent's PID and
342UID as well as the current process's.
343It also enables the deprecated
344.Dv NTTYDISC
345terminal line discipline.
346It also provides backwards compatibility with
347.Dq old
348SIOC[GS]IF{ADDR,DSTADDR,BRDADDR,NETMASK} interface ioctls, including
349binary compatibility with code written before the introduction of the
350sa_len field in sockaddrs.
351It also enables
352support for some older pre
353.Bx 4.4
354socket calls.
355.It Cd options COMPAT_BSDPTY
356This option is currenty on by default and enables the pty multiplexor
357.Xr ptm 4
358and
359.Xr ptmx 4
360to find and use ptys named
361.Pa /dev/ptyXX
362(master) and
363.Pa /dev/ttyXX
364(slave).
365Eventually this option will become optional as ptyfs based pseudo-ttys become
366the default, see
367.Xr mount_ptyfs 8 .
368.It Cd options COMPAT_SVR4
369On those architectures that support it, this enables binary
370compatibility with
371.At V.4
372applications built for the same architecture.
373This currently includes the i386, m68k, and sparc ports.
374.It Cd options COMPAT_LINUX
375On those architectures that support it, this enables binary
376compatibility with Linux ELF and
377.Xr a.out 5
378applications built for the same architecture.
379This currently includes the alpha, arm, i386, m68k, mips, powerpc, and
380x86_64 ports.
381.It Cd options COMPAT_SUNOS
382On those architectures that support it, this enables binary
383compatibility with
384.Tn SunOS 4.1
385applications built for the same architecture.
386This currently includes the sparc, sparc64 and most or all m68k ports.
387Note that the sparc64 requires the
388.Em COMPAT_NETBSD32
389option for 64-bit kernels, in addition to this option.
390.It Cd options COMPAT_ULTRIX
391On those architectures that support it, this enables binary
392compatibility with
393.Tn ULTRIX
394applications built for the same architecture.
395This currently is limited to the pmax.
396The functionality of this option is unknown.
397.It Cd options COMPAT_DARWIN
398On those architectures that support it, this enables binary compatibility with
399.Tn Darwin
400applications built for the same architecture.
401This feature is highly experimental, it requires COMPAT_MACH and
402EXEC_MACHO and it is currently limited to i386 and powerpc ports of
403.Nx .
404.It Cd options COMPAT_FREEBSD
405On those architectures that support it, this enables binary
406compatibility with
407.Fx
408applications built for the same architecture.
409At the moment this is limited to the i386 port.
410.It Cd options COMPAT_HPUX
411On those architectures that support it, this enables binary
412compatibility with
413.Tn HP/UX
414applications built for the same architecture.
415This is limited to the hp300 port, and has some known bugs.
416A limited set of programs do work.
417.It Cd options COMPAT_IBCS2
418On those architectures that support it, this enables binary
419compatibility with iBCS2 or SVR3 applications built for the same architecture.
420This is currently limited to the i386 and vax ports.
421.It Cd options COMPAT_IRIX
422On those architectures that support it, this enables binary
423compatibility with IRIX o32 binaries built for the same architecture.
424This feature is experimental, and it is currently limited to
425the sgimips port.
426.It Cd options COMPAT_MACH
427On those architectures that support it, this enables the emulation of
428Mach kernel traps for binaries built for the same architecture.
429This feature is highly experimental and it is currently
430limited to the i386 and powerpc ports of
431.Nx .
432.It Cd options COMPAT_OSF1
433On those architectures that support it, this enables binary
434compatibility with
435.Tn Digital
436.Ux
437.Po
438formerly
439.Tn OSF/1
440.Pc
441applications built for the same architecture.
442This is currently limited to the alpha port.
443.It Cd options COMPAT_NOMID
444Enable compatibility with
445.Xr a.out 5
446executables that lack a machine ID.
447This includes
448.Nx 0.8 Ns 's
449ZMAGIC format, and 386BSD and BSDI's
450QMAGIC, NMAGIC, and OMAGIC
451.Xr a.out 5
452formats.
453.It Cd options COMPAT_NETBSD32
454On those architectures that support it, this enables binary
455compatibility with 32-bit applications built for the same architecture.
456This is currently limited to the amd64 and sparc64 ports, and only
457applicable for 64-bit kernels.
458.It Cd options COMPAT_SVR4_32
459On those architectures that support it, this enables binary
460compatibility with 32-bit SVR4 applications built for the same architecture.
461This is currently limited to the sparc64 port, and only applicable for
46264-bit kernels.
463.It Cd options COMPAT_AOUT_M68K
464On m68k architectures which have switched to ELF,
465this enables binary compatibility with
466.Nx Ns Tn /m68k
467.Xr a.out 5
468executables on
469.Nx Ns Tn /m68k
470ELF kernels.
471This handles alignment incompatibility of m68k ABI between
472a.out and ELF which causes the structure padding differences.
473Currently only some system calls which use
474.Va struct stat
475are adjusted and some binaries which use
476.Xr sysctl 3
477to retrieve network details would not work properly.
478.It Cd options EXEC_MACHO
479On those architectures that support it, this adds support for running
480Mach-O executables.
481This is currently limited to the i386 and powerpc ports of
482.Nx .
483.It Cd options EXEC_ELF_NOTELESS
484Run unidentified ELF binaries as
485.Nx
486binaries.
487This might be needed for very old
488.Nx
489ELF binaries on some archs.
490These old binaries didn't contain an appropriate
491.Li .note.netbsd.ident
492section, and thus can't be identified by the kernel as
493.Nx
494binaries otherwise.
495Beware - if this option is on, the kernel would run
496.Em any
497unknown ELF binaries as if they were
498.Nx
499binaries.
500.It Cd options P1003_1B_SEMAPHORE
501Includes kernel support for the standard C library
502.Pq libc
503functions that implement semaphores as specified in
504.St -p1003.1-96 .
505.El
506.Ss Debugging Options
507.Bl -ohang
508.It Cd options DDB
509Compiles in a kernel debugger for diagnosing kernel problems.
510See
511.Xr ddb 4
512for details.
513.Em NOTE :
514not available on all architectures.
515.It Cd options DDB_FROMCONSOLE=integer
516If set to non-zero, DDB may be entered by sending a break on a serial
517console or by a special key sequence on a graphics console.
518A value of "0" ignores console breaks or key sequences.
519If not explicitly specified, the default value is "1".
520Note that this sets the value of the
521.Em ddb.fromconsole
522.Xr sysctl 3
523variable which may be changed at run time -- see
524.Xr sysctl 8
525for details.
526.It Cd options DDB_HISTORY_SIZE=integer
527If this is non-zero, enable history editing in the kernel debugger
528and set the size of the history to this value.
529.It Cd options DDB_ONPANIC
530The default if not specified is
531.Dq 1
532- just enter into DDB.
533If set to
534.Dq 2
535the kernel will
536attempt to print out a stack trace before entering into DDB.
537If set to
538.Dq 0
539the kernel will attempt to print out a stack trace
540and reboot the system.
541If set to
542.Dq -1
543then neither a stack trace is printed or DDB entered -
544it is as if DDB were not compiled into the kernel.
545Note that this sets the value of the
546.Em ddb.onpanic
547.Xr sysctl 3
548variable which may be changed at run time -- see
549.Xr sysctl 8
550for details.
551.It Cd options DDB_COMMANDONENTER=string
552This option specify commands which will be executed on each entry to DDB.
553This sets the default value of the
554.Em ddb.commandonenter
555.Xr sysctl 3
556variable which may be changed at run time.
557.It Cd options DDB_BREAK_CHAR=integer
558This option overrides using break to enter the kernel debugger
559on the serial console.
560The value given is the ascii value to be used instead.
561This is currently only supported by the com driver.
562.It Cd options KGDB
563Compiles in a remote kernel debugger stub for diagnosing kernel problems
564using the
565.Dq remote target
566feature of gdb.
567See
568.Xr gdb 1
569for details.
570.Em NOTE :
571not available on all architectures.
572.It Cd options KGDB_DEV
573Device number
574.Po
575as a
576.Dv dev_t
577.Pc
578of kgdb device.
579.It Cd options KGDB_DEVADDR
580Memory address of kgdb device.
581.It Cd options KGDB_DEVMODE
582Permissions of kgdb device.
583.It Cd options KGDB_DEVNAME
584Device name of kgdb device.
585.It Cd options KGDB_DEVRATE
586Baud rate of kgdb device.
587.It Cd makeoptions DEBUG="-g"
588The
589.Fl g
590flag causes
591.Pa netbsd.gdb
592to be built in addition to
593.Pa netbsd .
594.Pa netbsd.gdb
595is useful for debugging kernel crash dumps with gdb.
596See
597.Xr gdb 1
598for details.
599This also turns on
600.Em options DEBUG
601(which see).
602.It Cd options DEBUG
603Turns on miscellaneous kernel debugging.
604Since options are turned into preprocessor defines (see above),
605.Em options DEBUG
606is equivalent to doing a
607.Em #define DEBUG
608throughout the kernel.
609Much of the kernel has
610.Em #ifdef DEBUG
611conditionalized debugging code.
612Note that many parts of the kernel (typically device drivers) include their own
613.Em #ifdef XXX_DEBUG
614conditionals instead.
615This option also turns on certain other options,
616which may decrease system performance.
617.It Cd options DIAGNOSTIC
618Adds code to the kernel that does internal consistency checks.
619This code will cause the kernel to panic if corruption of internal data
620structures is detected.
621These checks can decrease performance up to 15%.
622.It Cd options KSTACK_CHECK_MAGIC
623Check kernel stack usage and panic if stack overflow is detected.
624This check is performance sensitive because it scans stack on each context
625switch.
626.It Cd options KTRACE
627Add hooks for the system call tracing facility, which allows users to
628watch the system call invocation behavior of processes.
629See
630.Xr ktrace 1
631for details.
632.It Cd options MSGBUFSIZE=integer
633This option sets the size of the kernel message buffer.
634This buffer holds the kernel output of
635.Fn printf
636when not (yet) read by
637.Xr syslogd 8 .
638This is particularly useful when the system has crashed and you wish to lookup
639the kernel output from just before the crash.
640Also, since the autoconfig output becomes more and more verbose,
641it sometimes happens that the message buffer overflows before
642.Xr syslogd 8
643was able to read it.
644Note that not all systems are capable of obtaining a variable sized message
645buffer.
646There are also some systems on which memory contents are not preserved
647across reboots.
648.It Cd options MALLOCLOG
649Enables an event log for
650.Xr malloc 9 .
651Useful for tracking down
652.Dq Data modified on freelist
653and
654.Dq multiple free
655problems.
656.It Cd options MALLOCLOGSIZE=integer
657Defines the number of entries in the malloc log.
658Default is 100000 entries.
659.It Cd options SYSTRACE
660Add hooks for the system call policy facility.
661See
662.Xr systrace 1
663for details.
664.It Cd options UVMHIST
665Enables the UVM history logs, which create in-memory traces of
666various UVM activities.
667These logs can be displayed be calling
668.Fn uvmhist_dump
669or
670.Fn uvm_hist
671with appropriate arguments from DDB.
672See the kernel source file sys/uvm/uvm_stat.c for details.
673.It Cd options UVMHIST_PRINT
674Prints the UVM history logs on the system console as entries are added.
675Note that the output is
676.Em extremely
677voluminous, so this option is really only useful for debugging
678the very earliest parts of kernel initialization.
679.El
680.Ss File Systems
681.Bl -ohang
682.It Cd file-system FFS
683Includes code implementing the Berkeley Fast File System
684.Em ( FFS ) .
685Most machines need this if they are not running diskless.
686.It Cd file-system EXT2FS
687Includes code implementing the Second Extended File System
688.Em ( EXT2FS ) ,
689revision 0 and revision 1 with the
690.Em filetype
691and
692.Em sparse_super
693options.
694This is the most commonly used file system on the Linux operating system,
695and is provided here for compatibility.
696Some of the specific features of
697.Em EXT2FS
698like the "behavior on errors" are not implemented.
699This file system can't be used with UID or GID greater than 65535.
700See
701.Xr mount_ext2fs 8
702for details.
703.It Cd file-system LFS
704.Bq Em EXPERIMENTAL
705Include the Log-structured File System
706.Em ( LFS ) .
707See
708.Xr mount_lfs 8
709and
710.Xr newfs_lfs 8
711for details.
712.It Cd file-system MFS
713Include the Memory File System
714.Em ( MFS ) .
715This file system stores files in swappable memory, and produces
716notable performance improvements when it is used as the file store
717for
718.Pa /tmp
719and similar file systems.
720See
721.Xr mount_mfs 8
722for details.
723.It Cd file-system NFS
724Include the client side of the Network File System
725.Pq Tn NFS
726remote file sharing protocol.
727Although the bulk of the code implementing
728.Tn NFS
729is kernel based, several user level daemons are needed for it to work.
730See
731.Xr mount_nfs 8
732for details.
733.It Cd file-system CD9660
734Includes code for the
735.Tn ISO
7369660 + Rock Ridge file system, which is the standard file system on many
737.Tn CD-ROM
738discs.
739Useful primarily if you have a
740.Tn CD-ROM
741drive.
742See
743.Xr mount_cd9660 8
744for details.
745.It Cd file-system MSDOSFS
746Includes the
747.Tn MS-DOS
748FAT file system, which is reportedly still used
749by unfortunate people who have not heard about
750.Nx .
751Also implements the
752.Tn Windows 95
753extensions to the same, which permit the use of longer, mixed case
754file names.
755See
756.Xr mount_msdos 8
757and
758.Xr fsck_msdos 8
759for details.
760.It Cd file-system NTFS
761.Bq Em EXPERIMENTAL
762Includes code for the
763.Tn Microsoft Windows NT
764file system.
765See
766.Xr mount_ntfs 8
767for details.
768.It Cd file-system FDESC
769Includes code for a file system, conventionally mounted on
770.Pa /dev/fd ,
771which permits access to the per-process file descriptor space via
772special files in the file system.
773See
774.Xr mount_fdesc 8
775for details.
776Note that this facility is redundant, and thus unneeded on most
777.Nx
778systems, since the
779.Xr fd 4
780pseudo-device driver already provides identical functionality.
781On most
782.Nx
783systems, instances of
784.Xr fd 4
785are mknoded under
786.Pa /dev/fd/
787and on
788.Pa /dev/stdin ,
789.Pa /dev/stdout ,
790and
791.Pa /dev/stderr .
792.It Cd file-system KERNFS
793Includes code which permits the mounting of a special file system
794(normally mounted on
795.Pa /kern )
796in which files representing various kernel variables and parameters
797may be found.
798See
799.Xr mount_kernfs 8
800for details.
801.It Cd file-system NULLFS
802Includes code for a loopback file system.
803This permits portions of the file hierarchy to be re-mounted in other places.
804The code really exists to provide an example of a stackable file system layer.
805See
806.Xr mount_null 8
807for details.
808.It Cd file-system OVERLAY
809Includes code for a file system filter.
810This permits the overlay file system to intercept all access to an underlying
811file system.
812This file system is intended to serve as an example of a stacking file
813system which has a need to interpose itself between an underlying file
814system and all other access.
815See
816.Xr mount_overlay 8
817for details.
818.It Cd file-system PORTAL
819.Bq Em EXPERIMENTAL
820Includes the portal file system.
821This permits interesting tricks like opening
822.Tn TCP
823sockets by opening files in the file system.
824The portal file system is conventionally mounted on
825.Pa /p
826and is partially implemented by a special daemon.
827See
828.Xr mount_portal 8
829for details.
830.It Cd file-system PROCFS
831Includes code for a special file system (conventionally mounted on
832.Pa /proc )
833in which the process space becomes visible in the file system.
834Among
835other things, the memory spaces of processes running on the system are
836visible as files, and signals may be sent to processes by writing to
837.Pa ctl
838files in the procfs namespace.
839See
840.Xr mount_procfs 8
841for details.
842.It Cd file-system UDF
843.Bq Em EXPERIMENTAL
844Includes code for the UDF file system commonly found on CD and DVD
845media but also more and more on USB sticks.
846Useful primarily if you have a CD or a DVD drive, be it a read-only
847or a rewritable device.
848Currently only supports read-access.
849See
850.Xr mount_udf 8
851for details.
852.It Cd file-system UMAPFS
853Includes a loopback file system in which user and group IDs may be
854remapped -- this can be useful when mounting alien file systems with
855different UIDs and GIDs than the local system.
856See
857.Xr mount_umap 8
858for details.
859.It Cd file-system UNION
860.Bq Em EXPERIMENTAL
861Includes code for the union file system, which permits directories to
862be mounted on top of each other in such a way that both file systems
863remain visible -- this permits tricks like allowing writing (and the
864deleting of files) on a read-only file system like a
865.Tn CD-ROM
866by mounting a local writable file system on top of the read-only file system.
867See
868.Xr mount_union 8
869for details.
870.It Cd file-system CODA
871.Bq Em EXPERIMENTAL
872Includes code for the Coda file system.
873Coda is a distributed file system like NFS and AFS.
874It is freely available, like NFS, but it functions much like AFS in being a
875.Dq stateful
876file system.
877Both Coda and AFS cache files on your local machine to improve performance.
878Then Coda goes a step further than AFS by letting you access the cached
879files when there is no available network, viz. disconnected laptops and
880network outages.
881In Coda, both the client and server are outside the kernel which makes
882them easier to experiment with.
883Coda is available for several UNIX and non-UNIX platforms.
884See http://www.coda.cs.cmu.edu for more details.
885.Em NOTE :
886You also need to enable the pseudo-device, vcoda, for the Coda
887file system to work.
888.It Cd file-system SMBFS
889.Bq Em EXPERIMENTAL
890Includes code for the SMB/CIFS file system.
891See
892.Xr mount_smbfs 8
893for details.
894.Em NOTE :
895You also need to enable the pseudo-device, nsmb, for the SMB
896file system to work.
897.It Cd file-system PTYFS
898.Bq Em EXPERIMENTAL
899Includes code for a special file system (normally mounted on
900.Pa /dev/pts )
901in which pseudo-terminal slave devices become visible in the file system.
902See
903.Xr mount_ptyfs 8
904for details.
905.It Cd file-system TMPFS
906.Bq Em EXPERIMENTAL
907Includes code for the efficient memory file system, normally used over
908.Pa /tmp .
909See
910.Xr mount_tmpfs 8
911for details.
912.El
913.Ss File System Options
914.Bl -ohang
915.It Cd options ROOTFS_MAGICLINKS
916Enables the expansion of special strings
917.Po
918beginning with
919.Dq @
920.Pc
921when traversing symbolic links on the root file system.
922See
923.Xr symlink 7
924for a list of supported strings.
925Note that this option only controls the enabling of this feature when
926the root file system is first mounted by the kernel at boot-up.
927This feature can still be manipulated with the
928.Xr mount 8
929command regardless of the setting of this option.
930.It Cd options NFSSERVER
931Include the server side of the
932.Em NFS
933(Network File System) remote file sharing protocol.
934Although the bulk of the code implementing
935.Em NFS
936is kernel based, several user level daemons are needed for it to
937work.
938See
939.Xr mountd 8
940and
941.Xr nfsd 8
942for details.
943.It Cd options QUOTA
944Enables kernel support for file system quotas.
945See
946.Xr quotaon 8 ,
947.Xr edquota 8 ,
948and
949.Xr quota 1
950for details.
951Note that quotas only work on
952.Dq ffs
953file systems, although
954.Xr rpc.rquotad 8
955permits them to be accessed over
956.Em NFS .
957.It Cd options FFS_EI
958Enable ``Endian-Independent'' FFS support.
959This allows a system to mount an FFS file system created for another
960architecture, at a small performance cost for all FFS file systems.
961See also
962.Xr newfs 8 ,
963.Xr fsck_ffs 8 ,
964.Xr dumpfs 8
965for file system byte order status and manipulation.
966.It Cd options FFS_NO_SNAPSHOT
967Disable the
968.Dq file system snapshot
969support in FFS file systems.
970Maybe useful for install media kernels, small memory systems and
971embedded systems which don't require the snapshot support.
972.It Cd options NVNODE=integer
973This option sets the size of the cache used by the name-to-inode translation
974routines, (a.k.a. the
975.Fn namei
976cache, though called by many other names in the kernel source).
977By default, this cache has
978.Dv NPROC
979(set as 20 + 16 * MAXUSERS) * (80 + NPROC / 8) entries.
980A reasonable way to derive a value of
981.Dv NVNODE ,
982should you notice a large number of namei cache misses with a tool such as
983.Xr systat 1 ,
984is to examine your system's current computed value with
985.Xr sysctl 8 ,
986(which calls this parameter "kern.maxvnodes") and to increase this value
987until either the namei cache hit rate improves or it is determined that
988your system does not benefit substantially from an increase in the size of
989the namei cache.
990.It Cd options NAMECACHE_ENTER_REVERSE
991Causes the namei cache to always enter a reverse mapping (vnode -\*[Gt] name)
992as well as a normal one.
993Normally, this is already done for directory vnodes, to speed up the getcwd
994operation.
995This option will cause longer hash chains in the reverse cache, and thus
996slow down getcwd somewhat.
997However, it does make vnode -\*[Gt] path translations possible in some cases.
998For now, only useful if strict /proc/#/maps emulation for Linux binaries is
999required.
1000.It Cd options EXT2FS_SYSTEM_FLAGS
1001This option changes the behavior of the APPEND and IMMUTABLE flags
1002for a file on an
1003.Em EXT2FS
1004file system.
1005Without this option, the superuser or owner of the file can
1006set and clear them.
1007With this option, only the superuser can set them, and
1008they can't be cleared if the securelevel is greater than 0.
1009See also
1010.Xr chflags 1 .
1011.It Cd options NFS_BOOT_BOOTP
1012Enable use of the BOOTP protocol (RFCs 951 and 1048) to get configuration
1013information if NFS is used to mount the root file system.
1014See
1015.Xr diskless 8
1016for details.
1017.It Cd options NFS_BOOT_DHCP
1018Same as
1019.Dq NFS_BOOT_BOOTP ,
1020but use the DHCP extensions to the
1021BOOTP protocol (RFC 1541).
1022.It Cd options NFS_BOOT_BOOTP_REQFILE
1023Specifies the string sent in the bp_file field of the BOOTP / DHCP
1024request packet.
1025.It Cd options NFS_BOOT_BOOTPARAM
1026Enable use of the BOOTPARAM protocol, consisting of RARP and
1027BOOTPARAM RPC, to get configuration information if NFS
1028is used to mount the root file system.
1029See
1030.Xr diskless 8
1031for details.
1032.It Cd options NFS_BOOT_RWSIZE=value
1033Set the initial NFS read and write sizes for diskless-boot requests.
1034The normal default is 8Kbytes.
1035This option provides a way to lower the value (e.g., to 1024 bytes)
1036as a workaround for buggy network interface cards or boot PROMs.
1037Once booted, the read and write request sizes can be increased by
1038remounting the file system.
1039See
1040.Xr mount_nfs 8
1041for details.
1042.It Cd options NFS_V2_ONLY
1043Reduce the size of the NFS client code by omitting code that's only required
1044for NFSv3 and NQNFS support, leaving only that code required to use NFSv2
1045servers.
1046.It Cd options SOFTDEP
1047Enable kernel support for soft-dependencies in FFS filesystems.
1048Softdep improves write performance by aggregating and properly
1049ordering disk metadata writes, achieving near-asynchronous write
1050performance while maintaining the filesystem consistency of
1051synchronous writes.
1052Soft-dependencies are enabled on a per-mount basis.
1053See
1054.Xr mount 8
1055for details.
1056.It Cd options UFS_DIRHASH
1057Increase lookup performance by maintaining in-core hash tables
1058for large directories.
1059.El
1060.Ss Buffer queue strategy options
1061The following options enable alternative buffer queue strategies.
1062.Bl -ohang
1063.It Cd options BUFQ_READPRIO
1064Enable experimental buffer queue strategy for disk I/O.
1065In the default strategy, outstanding disk requests are ordered by
1066sector number and sent to the disk, regardless of whether the
1067operation is a read or write; this option gives priority to issuing
1068read requests over write requests.
1069Although requests may therefore be issued out of sector-order, causing
1070more seeks and thus lower overall throughput, interactive system
1071responsiveness under heavy disk I/O load may be improved, as processes
1072blocking on disk reads are serviced sooner (file writes typically
1073don't cause applications to block).
1074The performance effect varies greatly depending on the hardware, drive
1075firmware, filesystem configuration, workload, and desired performance
1076trade-off.
1077Systems using drive write-cache (most modern IDE disks, by default)
1078are unlikely to benefit and may well suffer; such disks acknowledge
1079writes very quickly, and optimise them internally according to
1080physical layout.
1081Giving these disks as many requests to work with as possible (the
1082standard strategy) will typically produce the best results, especially
1083if the drive has a large cache; the drive will silently complete
1084writes from cache as it seeks for reads.
1085Disks that support a large number of concurrent tagged requests (SCSI
1086disks and many hardware RAID controllers) expose this internal
1087scheduling with tagged responses, and don't block for reads; such
1088disks may not see a noticable difference with either strategy.
1089However, if IDE disks are run with write-cache disabled for safety,
1090writes are not acknowledged until actually completed, and only one
1091request can be outstanding; a large number of small writes in one
1092locality can keep the disk busy, starving reads elsewhere on the disk.
1093Such systems are likely to see the most benefit from this option.
1094Finally, the performance interaction of this option with ffs soft
1095dependencies can be subtle, as that mechanism can drastically alter
1096the workload for filesystem metadata writes.
1097.It Cd options BUFQ_PRIOCSCAN
1098Enable another experimental buffer queue strategy for disk I/O,
1099Per-priority cyclical scan.
1100.It Cd options NEW_BUFQ_STRATEGY
1101Synonym of
1102.Em BUFQ_READPRIO .
1103.El
1104.Ss Miscellaneous Options
1105.Bl -ohang
1106.It Cd options LKM
1107Enable loadable kernel modules.
1108See
1109.Xr lkm 4
1110for details.
1111.Em NOTE :
1112not available on all architectures.
1113.It Cd options MEMORY_DISK_DYNAMIC
1114This options makes the
1115.Xr md 4
1116.Tn RAM
1117disk size dynamically sized.
1118It is incompatible with
1119.Xr mdsetimage 8 .
1120.It Cd options MEMORY_DISK_HOOKS
1121This option allows for some machine dependent functions to be called when
1122the
1123.Xr md 4
1124.Tn RAM
1125disk driver is configured.
1126This can result in automatically loading a
1127.Tn RAM
1128disk from floppy on open (among other things).
1129.It Cd options MEMORY_DISK_IS_ROOT
1130Forces the
1131.Xr md 4
1132.Tn RAM
1133disk to be the root device.
1134This can only be overridden when
1135the kernel is booted in the 'ask-for-root' mode.
1136.It Cd options MEMORY_DISK_ROOT_SIZE=integer
1137Allocates the given number of 512 byte blocks as memory for the
1138.Xr md 4
1139.Tn RAM
1140disk, to be populated with
1141.Xr mdsetimage 8 .
1142.It Cd options MEMORY_DISK_SERVER=0
1143Do not include the interface to a userland memory disk server process.
1144Per default, this options is set to 1, including the support code.
1145Useful for install media kernels.
1146.It Cd options VND_COMPRESSION
1147Enables the
1148.Xr vnd 4
1149driver to also handle compressed images.
1150See
1151.Xr vndcompress 1 ,
1152.Xr vnd 4
1153and
1154.Xr vnconfig 8
1155for more information.
1156.It Cd options MALLOC_NOINLINE
1157Time critical fixed size memory allocation is performed with
1158.Fn MALLOC
1159and
1160.Fn FREE .
1161Normally these expand to inline code, but with
1162.Em MALLOC_NOINLINE
1163these call the normal
1164.Fn malloc
1165and
1166.Fn free
1167functions.
1168Useful for install media kernels, small memory systems and embedded systems.
1169.It Cd options HZ=integer
1170On ports that support it, set the system clock frequency (see
1171.Xr hz 9 )
1172to the supplied value.
1173Handle with care.
1174.It Cd options NTP
1175Turns on in-kernel precision timekeeping support used by software
1176implementing
1177.Em NTP
1178(Network Time Protocol, RFC 1305).
1179The
1180.Em NTP
1181option adds an in-kernel Phase-Locked Loop (PLL) for normal
1182.Em NTP
1183operation, and a Frequency-Locked Loop (FLL) for intermittently-connected
1184operation.
1185.Xr ntpd 8
1186will employ a user-level PLL when kernel support is unavailable,
1187but the in-kernel version has lower latency and more precision, and
1188so typically keeps much better time.
1189The interface to the kernel
1190.Em NTP
1191support is provided by the
1192.Xr ntp_adjtime 2
1193and
1194.Xr ntp_gettime 2
1195system calls, which are intended for use by
1196.Xr ntpd 8
1197and are enabled by the option.
1198On systems with sub-microsecond resolution timers, or where (HZ / 100000)
1199is not an integer, the
1200.Em NTP
1201option also enables extended-precision arithmetic to keep track of
1202fractional clock ticks at NTP time-format precision.
1203.It Cd options PPS_SYNC
1204This option enables a kernel serial line discipline for receiving time
1205phase signals from an external reference clock such as a radio clock.
1206(The
1207.Em NTP
1208option (which see) must be on if the
1209.Em PPS_SYNC
1210option is used.)
1211Some reference clocks generate a Pulse Per Second (PPS) signal in
1212phase with their time source.
1213The
1214.Em PPS
1215line discipline receives this signal on either the data leads
1216or the DCD control lead of a serial port.
1217.Em NTP
1218uses the PPS signal to discipline the local clock oscillator to a high
1219degree of precision (typically less than 50 microseconds in time and
12200.1 ppm in accuracy).
1221.Em PPS
1222can also generate a serial output pulse when the system receives a PPS
1223interrupt.
1224This can be used to measure the system interrupt latency and thus calibrate
1225.Em NTP
1226to account for it.
1227Using
1228.Em PPS
1229usually requires a
1230gadget box
1231to convert from TTL to RS-232 signal levels.
1232The gadget box and PPS are described in more detail in the HTML documentation
1233for
1234.Xr ntpd 8
1235in
1236.Pa /usr/share/doc/html/ntp .
1237.It Cd options NO_TSC_TIME
1238Don't use TSC microtime, even if available (i386 only).
1239Improves time behavior under VMware.
1240.It Cd options SETUIDSCRIPTS
1241Allows scripts with the setuid bit set to execute as the effective
1242user rather than the real user, just like binary executables.
1243.Pp
1244.Em NOTE :
1245Using this option will also enable
1246.Em options FDSCRIPTS
1247.It Cd options FDSCRIPTS
1248Allows execution of scripts with the execute bit set, but not the
1249read bit, by opening the file and passing the file descriptor to
1250the shell, rather than the filename.
1251.Pp
1252.Em NOTE :
1253Execute only (non-readable) scripts will have
1254.Va argv[0]
1255set to
1256.Pa /dev/fd/* .
1257What this option allows as far as security is
1258concerned, is the ability to safely ensure that the correct script
1259is run by the interpreter, as it is passed as an already open file.
1260.It Cd options PUCCN
1261Enables treating serial ports found on PCI boards
1262.Xr puc 4
1263as potential console devices.
1264The method for choosing such a console device is port dependent.
1265.It Cd options RTC_OFFSET=integer
1266The kernel (and typically the hardware battery backed-up clock on
1267those machines that have one) keeps time in
1268.Em UTC
1269(Universal Coordinated Time, once known as
1270.Em GMT ,
1271or Greenwich Mean Time)
1272and not in the time of the local time zone.
1273The
1274.Em RTC_OFFSET
1275option is used on some ports (such as the i386) to tell the kernel
1276that the hardware clock is offset from
1277.Em UTC
1278by the specified number of minutes.
1279This is typically used when a machine boots several operating
1280systems and one of them wants the hardware clock to run in the
1281local time zone and not in
1282.Em UTC ,
1283e.g.
1284.Em RTC_OFFSET=300
1285means
1286the hardware clock is set to US Eastern Time (300 minutes behind
1287.Em UTC ) ,
1288and not
1289.Em UTC .
1290(Note:
1291.Em RTC_OFFSET
1292is used to initialize a kernel variable named
1293.Va rtc_offset
1294which is the source actually used to determine the clock offset, and
1295which may be accessed via the kern.rtc_offset sysctl variable.
1296See
1297.Xr sysctl 8
1298and
1299.Xr sysctl 3
1300for details.
1301Since the kernel clock is initialized from the hardware clock very
1302early in the boot process, it is not possible to meaningfully change
1303.Va rtc_offset
1304in system initialization scripts.
1305Changing this value currently may only be done at kernel compile
1306time or by patching the kernel and rebooting).
1307.Pp
1308.Em NOTE :
1309Unfortunately, in many cases where the hardware clock
1310is kept in local time, it is adjusted for Daylight Savings
1311Time; this means that attempting to use
1312.Em RTC_OFFSET
1313to let
1314.Nx
1315coexist with such an operating system, like Windows,
1316would necessitate changing
1317.Em RTC_OFFSET
1318twice a year.
1319As such, this solution is imperfect.
1320.It Cd options KMEMSTATS
1321The kernel memory allocator,
1322.Xr malloc 9 ,
1323will keep statistics on its performance if this option is enabled.
1324Unfortunately, this option therefore essentially disables the
1325.Fn MALLOC
1326and
1327.Fn FREE
1328forms of the memory allocator, which are used to enhance the performance
1329of certain critical sections of code in the kernel.
1330This option therefore can lead to a significant decrease in the
1331performance of certain code in the kernel if enabled.
1332Examples of such code include the
1333.Fn namei
1334routine, the
1335.Xr ccd 4
1336driver,
1337and much of the networking code.
1338.It Cd options MAXUPRC=integer
1339Sets the soft
1340.Dv RLIMIT_NPROC
1341resource limit, which specifies the maximum number of simultaneous
1342processes a user is permitted to run, for process 0;
1343this value is inherited by its child processes.
1344It defaults to
1345.Em CHILD_MAX ,
1346which is currently defined to be 160.
1347Setting
1348.Em MAXUPRC
1349to a value less than
1350.Em CHILD_MAX
1351is not permitted, as this would result in a violation of the semantics of
1352.St -p1003.1-90 .
1353.It Cd options NOFILE=integer
1354Sets the soft
1355.Dv RLIMIT_NOFILE
1356resource limit, which specifies the maximum number of open
1357file descriptors for each process;
1358this value is inherited by its child processes.
1359It defaults to
1360.Em OPEN_MAX ,
1361which is currently defined to be 64.
1362.It Cd options MAXFILES=integer
1363Sets the default value of the
1364.Em kern.maxfiles
1365sysctl variable, which indicates the maximum number of files that may
1366be open in the system.
1367.It Cd options DEFCORENAME=string
1368Sets the default value of the
1369.Em kern.defcorename
1370sysctl variable, otherwise it is set to
1371.Nm %n.core .
1372See
1373.Xr sysctl 8
1374and
1375.Xr sysctl 3
1376for details.
1377.It Cd options RASOPS_CLIPPING
1378Enables clipping within the
1379.Nm rasops
1380raster-console output system.
1381.Em NOTE :
1382only available on architectures that use
1383.Nm rasops
1384for console output.
1385.It Cd options RASOPS_SMALL
1386Removes optimized character writing code from the
1387.Nm rasops
1388raster-console output system.
1389.Em NOTE :
1390only available on architectures that use
1391.Nm rasops
1392for console output.
1393.It Cd options INCLUDE_CONFIG_FILE
1394Embeds the kernel config file used to define the kernel in the kernel
1395binary itself.
1396The embedded data also includes any files directly included by the config
1397file itself, e.g.
1398.Pa GENERIC.local
1399or
1400.Pa std.$MACHINE .
1401The embedded config file can be extracted from the resulting kernel with
1402.Xr config 1
1403.Fl x ,
1404or by the following command:
1405.Bd -literal -offset indent
1406strings netbsd | sed -n 's/^_CFG_//p' | unvis
1407.Ed
1408.It Cd options INCLUDE_JUST_CONFIG
1409Similar to the above option, but includes just the actual config file,
1410not any included files.
1411.It Cd options PIPE_SOCKETPAIR
1412Use slower, but smaller socketpair(2)-based pipe implementation instead
1413of default faster, but bigger one.
1414Primarily useful for installation kernels.
1415.It Cd options USERCONF
1416Compiles in the in-kernel device configuration manager.
1417See
1418.Xr userconf 4
1419for details.
1420.It Cd options PERFCTRS
1421Compiles in kernel support for CPU performance-monitoring counters.
1422See
1423.Xr pmc 1
1424for details.
1425.Em NOTE :
1426not available on all architectures.
1427.It Cd options XSERVER
1428Compiles in kernel support for
1429.Tn X11
1430on architectures that still use (or can use) the legacy
1431.Em pccons
1432console drivers rather than
1433.Xr wscons 4 .
1434These include bebox, i386, shark.
1435.It Cd options XSERVER_DDB
1436A supplement to XSERVER that adds support for entering
1437.Xr ddb 4
1438while in
1439.Tn X11 .
1440.El
1441.Ss Networking Options
1442.Bl -ohang
1443.It Cd options GATEWAY
1444Enables
1445.Em IPFORWARDING
1446(which see)
1447and (on most ports) increases the size of
1448.Em NMBCLUSTERS
1449(which see).
1450In general,
1451.Em GATEWAY
1452is used to indicate that a system should act as a router, and
1453.Em IPFORWARDING
1454is not invoked directly.
1455(Note that
1456.Em GATEWAY
1457has no impact on protocols other than
1458.Tn IP ,
1459such as
1460.Tn CLNP
1461or
1462.Tn XNS ) .
1463.Em GATEWAY
1464option also compiles IPv4 fast forwarding code into the kernel.
1465.It Cd options ICMPPRINTFS
1466The
1467.Em ICMPPRINTFS
1468option will enable debugging information to be printed about
1469the
1470.Xr icmp 4
1471protocol.
1472.It Cd options IPFORWARDING=value
1473If
1474.Em value
1475is 1 this enables IP routing behavior.
1476If
1477.Em value
1478is 0 (the default), it disables it.
1479The
1480.Em GATEWAY
1481option sets this to 1 automatically.
1482With this option enabled, the machine will forward IP datagrams destined
1483for other machines between its interfaces.
1484Note that even without this option, the kernel will
1485still forward some packets (such as source routed packets) -- removing
1486.Em GATEWAY
1487and
1488.Em IPFORWARDING
1489is insufficient to stop all routing through a bastion host on a
1490firewall -- source routing is controlled independently.
1491To turn off source routing, use
1492.Em options IPFORWSRCRT=0
1493(which see).
1494Note that IP forwarding may be turned on and off independently of the
1495setting of the
1496.Em IPFORWARDING
1497option through the use of the
1498.Em net.inet.ip.forwarding
1499sysctl variable.
1500If
1501.Em net.inet.ip.forwarding
1502is 1, IP forwarding is on.
1503See
1504.Xr sysctl 8
1505and
1506.Xr sysctl 3
1507for details.
1508.It Cd options IPFORWSRCRT=value
1509If
1510.Em value
1511is set to zero, source routing of IP datagrams is turned off.
1512If
1513.Em value
1514is set to one (the default) or the option is absent, source routed IP
1515datagrams are forwarded by the machine.
1516Note that source routing of IP packets may be turned on and off
1517independently of the setting of the
1518.Em IPFORWSRCRT
1519option through the use of the
1520.Em net.inet.ip.forwsrcrt
1521sysctl variable.
1522If
1523.Em net.inet.ip.forwsrcrt
1524is 1, forwarding of source routed IP datagrams is on.
1525See
1526.Xr sysctl 8
1527and
1528.Xr sysctl 3
1529for details.
1530.It Cd options IFA_STATS
1531Tells the kernel to maintain per-address statistics on bytes sent
1532and received over (currently) internet and appletalk addresses.
1533.\"This can be a fairly expensive operation, so you probably want to
1534.\"keep this disabled.
1535The option is not recommended as it degrades system stability.
1536.It Cd options IFQ_MAXLEN=value
1537Increases the allowed size of the network interface packet queues.
1538The default queue size is 50 packets, and you do not normally need
1539to increase it.
1540.It Cd options MROUTING
1541Includes support for IP multicast routers.
1542You certainly want
1543.Em INET
1544with this.
1545Multicast routing is controlled by the
1546.Xr mrouted 8
1547daemon.
1548See also option
1549.Cd PIM .
1550.It Cd options PIM
1551Includes support for Protocol Independent Multicast (PIM) routing.
1552You need
1553.Em MROUTING
1554and
1555.Em INET
1556with this.
1557Software using this can be found e.g. in
1558.Pa pkgsrc/net/xorp .
1559.It Cd options INET
1560Includes support for the
1561.Tn TCP/IP
1562protocol stack.
1563You almost certainly want this.
1564See
1565.Xr inet 4
1566for details.
1567.It Cd options INET6
1568Includes support for the
1569.Tn IPv6
1570protocol stack.
1571See
1572.Xr inet6 4
1573for details.
1574Unlike
1575.Em INET ,
1576.Em INET6
1577enables multicast routing code as well.
1578This option requires
1579.Em INET
1580at this moment, but it should not.
1581.It Cd options ND6_DEBUG
1582The option sets the default value of net.inet6.icmp6.nd6_debug to 1,
1583for debugging IPv6 neighbor discovery protocol handling.
1584See
1585.Xr sysctl 3
1586for details.
1587.It Cd options IPSEC
1588Includes support for the
1589.Tn IPsec
1590protocol.
1591See
1592.Xr ipsec 4
1593for details.
1594.It Cd options IPSEC_DEBUG
1595Enables debugging code in
1596.Tn IPsec
1597stack.
1598See
1599.Xr ipsec 4
1600for details.
1601.It Cd options IPSEC_ESP
1602Includes support for
1603.Tn IPsec
1604.Tn ESP
1605protocol.
1606See
1607.Xr ipsec 4
1608for details.
1609.It Cd options IPSEC_NAT_T
1610Includes support for
1611.Tn IPsec
1612Network Address Translator traversal (NAT-T), as described in RFCs 3947
1613and 3948.
1614This feature might be patent-encumbered in some countries.
1615.It Cd options ALTQ
1616Enabled ALTQ (Alternate Queueing).
1617For simple rate-limiting, use
1618.Xr tbrconfig 8
1619to set up the interface transmission rate.
1620To use queueing disciplines, their appropriate kernel options should also
1621be defined (documented below).
1622Queueing disciplines are managed by
1623.Xr altqd 8 .
1624See
1625.Xr altq 9
1626for details.
1627.It Cd options ALTQ_HFSC
1628Include support for ALTQ-implemented HFSC (Hierarchical Fair Service Curve)
1629module.
1630HFSC supports both link-sharing and guaranteed real-time services.
1631HFSC employs a service curve based QoS model, and its unique feature
1632is an ability to decouple delay and bandwidth allocation.
1633Requires
1634.Em ALTQ_RED
1635to use the RED queueing discipline on HFSC classes, or
1636.Em ALTQ_RIO
1637to use the RIO queueing discipline on HFSC classes.
1638This option assumes
1639.Em ALTQ .
1640.It Cd options ALTQ_PRIQ
1641Include support for ALTQ-implemented PRIQ (Priority Queueing).
1642PRIQ implements a simple priority-based queueing discipline.
1643A higher priority class is always served first.
1644Requires
1645.Em ALTQ_RED
1646to use the RED queueing discipline on HFSC classes, or
1647.Em ALTQ_RIO
1648to use the RIO queueing discipline on HFSC classes.
1649This option assumes
1650.Em ALTQ .
1651.It Cd options ALTQ_WFQ
1652Include support for ALTQ-implemented WFQ (Weighted Fair Queueing).
1653WFQ implements a weighted-round robin scheduler for a set of queues.
1654A weight can be assigned to each queue to give a different proportion
1655of the link capacity.
1656A hash function is used to map a flow to one of a set of queues.
1657This option assumes
1658.Em ALTQ .
1659.It Cd options ALTQ_FIFOQ
1660Include support for ALTQ-implemented FIFO queueing.
1661FIFOQ is a simple drop-tail FIFO (First In, First Out) queueing discipline.
1662This option assumes
1663.Em ALTQ .
1664.It Cd options ALTQ_RIO
1665Include support for ALTQ-implemented RIO (RED with In/Out).
1666The original RIO has 2 sets of RED parameters; one for in-profile
1667packets and the other for out-of-profile packets.
1668At the ingress of the network, profile meters tag packets as IN or
1669OUT based on contracted profiles for customers.
1670Inside the network, IN packets receive preferential treatment by
1671the RIO dropper.
1672ALTQ/RIO has 3 drop precedence levels defined for the Assured Forwarding
1673PHB of DiffServ (RFC 2597).
1674This option assumes
1675.Em ALTQ .
1676.It Cd options ALTQ_BLUE
1677Include support for ALTQ-implemented Blue buffer management.
1678Blue is another active buffer management mechanism.
1679This option assumes
1680.Em ALTQ .
1681.It Cd options ALTQ_FLOWVALVE
1682Include support for ALTQ-implemented Flowvalve.
1683Flowvalve is a simple implementation of a RED penalty box that identifies
1684and punishes misbehaving flows.
1685This option requires
1686.Em ALTQ_RED
1687and assumes
1688.Em ALTQ .
1689.It Cd options ALTQ_CDNR
1690Include support for ALTQ-implemented CDNR (diffserv traffic conditioner)
1691packet marking/manipulation.
1692Traffic conditioners are components to meter, mark, or drop incoming
1693packets according to some rules.
1694As opposed to queueing disciplines, traffic conditioners handle incoming
1695packets at an input interface.
1696This option assumes
1697.Em ALTQ .
1698.It Cd options ALTQ_NOPCC
1699Disables use of processor cycle counter to measure time in ALTQ.
1700This option should be defined for a non-Pentium i386 CPU which does not
1701have TSC, SMP (per-CPU counters are not in sync), or power management
1702which affects processor cycle counter.
1703This option assumes
1704.Em ALTQ .
1705.It Cd options ALTQ_IPSEC
1706Include support for IPsec in IPv4 ALTQ.
1707This option assumes
1708.Em ALTQ .
1709.It Cd options ALTQ_JOBS
1710Undocumented at this time.
1711.It Cd options ALTQ_AFMAP
1712Include support for an undocumented ALTQ feature that is used to map an IP
1713flow to an ATM VC (Virtual Circuit).
1714This option assumes
1715.Em ALTQ .
1716.It Cd options ALTQ_LOCALQ
1717Include support for ALTQ-implemented local queues.
1718Its practical use is undefined.
1719Assumes
1720.Em ALTQ .
1721.It Cd options SUBNETSARELOCAL
1722Sets default value for net.inet.ip.subnetsarelocal variable, which
1723controls whether non-directly-connected subnets of connected networks
1724are considered "local" for purposes of choosing the MSS for a TCP
1725connection.
1726This is mostly present for historic reasons and completely irrelevant if
1727you enable Path MTU discovery.
1728.It Cd options HOSTZEROBROADCAST
1729Sets default value for net.inet.ip.hostzerobroadcast variable, which
1730controls whether the zeroth host address of each connected subnet is
1731also considered a broadcast address.
1732Default value is "1", for compatibility with old systems; if this is
1733set to zero on all hosts on a subnet, you should be able to fit an extra
1734host per subnet on the
1735".0" address.
1736.It Cd options MCLSHIFT=value
1737This option is the base-2 logarithm of the size of mbuf clusters.
1738The
1739.Bx
1740networking stack keeps network packets in a linked
1741list, or chain, of kernel buffer objects called mbufs.
1742The system provides larger mbuf clusters as an optimization for
1743large packets, instead of using long chains for large packets.
1744The mbuf cluster size,
1745or
1746.Em MCLBYTES ,
1747must be a power of two, and is computed as two raised to the power
1748.Em MCLSHIFT .
1749On systems with Ethernet network adaptors,
1750.Em MCLSHIFT
1751is often set to 11, giving 2048-byte mbuf clusters, large enough to
1752hold a 1500-byte
1753.Tn Ethernet
1754frame in a single cluster.
1755Systems with network interfaces supporting larger frame sizes like
1756.Tn ATM ,
1757.Tn FDDI ,
1758or
1759.Tn HIPPI
1760may perform better with
1761.Em MCLSHIFT
1762set to 12 or 13, giving mbuf cluster sizes of 4096 and 8192 bytes,
1763respectively.
1764.It Cd options NS
1765Include support for the
1766.Tn Xerox
1767.Tn XNS
1768protocol stack.
1769See
1770.Xr ns 4
1771for details.
1772.It Cd options ISO,TPIP
1773Include support for the ubiquitous
1774.Tn OSI
1775protocol stack.
1776See
1777.Xr iso 4
1778for details.
1779This option assumes
1780.Em INET .
1781.It Cd options EON
1782Include support for tunneling
1783.Tn OSI
1784protocols over
1785.Tn IP .
1786Known to be broken, or at least very fragile, and undocumented.
1787.It Cd options CCITT,LLC,HDLC
1788Include support for the
1789.Tn CCITT
1790(nee
1791.Tn ITU-TSS )
1792.Tn X.25
1793protocol stack.
1794The state of this code is currently unknown, and probably contains bugs.
1795This option assumes
1796.Em INET .
1797.It Cd options NETATALK
1798Include support for the
1799.Tn AppleTalk
1800protocol stack.
1801The kernel provides provision for the
1802.Em Datagram Delivery Protocol
1803(DDP), providing SOCK_DGRAM support and
1804.Tn AppleTalk
1805routing.
1806This stack is used by the
1807.Em NETATALK
1808package, which adds support for
1809.Tn AppleTalk
1810server services via user libraries and applications.
1811.It Cd options IPNOPRIVPORTS
1812Normally, only root can bind a socket descriptor to a so-called
1813.Dq privileged
1814.Tn TCP
1815port, that is, a port number in the range 0-1023.
1816This option eliminates those checks from the kernel.
1817This can be useful if there is a desire to allow daemons without
1818privileges to bind those ports, e.g., on firewalls.
1819The security tradeoffs in doing this are subtle.
1820This option should only be used by experts.
1821.It Cd options TCP_COMPAT_42
1822.Tn TCP
1823bug compatibility with
1824.Bx 4.2 .
1825In
1826.Bx 4.2 ,
1827.Tn TCP
1828sequence numbers were 32-bit signed values.
1829Modern implementations of TCP use unsigned values.
1830This option clamps the initial sequence number to start in
1831the range 2^31 rather than the full unsigned range of 2^32.
1832Also, under
1833.Bx 4.2 ,
1834keepalive packets must contain at least one byte or else
1835the remote end would not respond.
1836.It Cd options TCP_DEBUG
1837Record the last
1838.Em TCP_NDEBUG
1839TCP packets with SO_DEBUG set, and decode to the console if
1840.Em tcpconsdebug
1841is set.
1842.It Cd options TCP_NDEBUG
1843Number of packets to record for
1844.Em TCP_DEBUG .
1845Defaults to 100.
1846.It Cd options TCP_SENDSPACE=value
1847.It Cd options TCP_RECVSPACE=value
1848These options set the max TCP window size to other sizes than the default.
1849The TCP window sizes can be altered via
1850.Xr sysctl 8
1851as well.
1852.It Cd options TCP_INIT_WIN=value
1853This option sets the initial TCP window size for non-local connections,
1854which is used when the transmission starts.
1855The default size is 1, but if the machine should act more aggressively,
1856the initial size can be set to some other value.
1857The initial TCP window size can be set via
1858.Xr sysctl 8
1859as well.
1860.It Cd options PFIL_HOOKS
1861This option turns on the packet filter interface hooks.
1862See
1863.Xr pfil 9
1864for details.
1865This option assumes
1866.Em INET .
1867.It Cd options IPFILTER_LOG
1868This option, in conjunction with
1869.Em pseudo-device ipfilter ,
1870enables logging of IP packets using ip-filter.
1871.It Cd options IPFILTER_DEFAULT_BLOCK
1872This option sets the default policy of ip-filter.
1873If it is set, ip-filter will block packets by default.
1874.It Cd options BRIDGE_IPF
1875This option causes
1876.Em bridge
1877devices to use the IP and/or IPv6 filtering hooks, forming
1878a link-layer filter that uses protocol-layer rules.
1879This option assumes the presence of
1880.Em pseudo-device ipfilter .
1881.It Cd options MBUFTRACE
1882This option can help track down mbuf leaks.
1883When enabled, mbufs are tagged with the devices and protocols using them,
1884which slightly decreases network performance.
1885This additional information can be viewed with
1886.Xr netstat 1 :
1887.Dl Ic netstat Fl mssv
1888Not all devices or protocols support this option.
1889.El
1890.Ss Sysctl Related Options
1891.Bl -ohang
1892.It Cd options SYSCTL_DISALLOW_CREATE
1893Disallows the creation or deletion of nodes from the sysctl tree, as
1894well as the assigning of descriptions to nodes that lack them, by any
1895process.
1896These operations are still available to kernel sub-systems, including
1897loadable kernel modules.
1898.It Cd options SYSCTL_DISALLOW_KWRITE
1899Prevents processes from adding nodes to the sysctl tree that make
1900existing kernel memory areas writeable.
1901Sections of kernel memory can still be read and new nodes that own
1902their own data may still be writeable.
1903.It Cd options SYSCTL_DEBUG_SETUP
1904Causes the SYSCTL_SETUP routines to print a brief message when they
1905are invoked.
1906This is merely meant as an aid in determining the order in which
1907sections of the tree are created.
1908.It Cd options SYSCTL_DEBUG_CREATE
1909Prints a message each time
1910.Fn sysctl_create ,
1911the function that adds nodes to the tree, is called.
1912.It Cd options SYSCTL_INCLUDE_DESCR
1913Causes the kernel to include short, human readable descriptions for
1914nodes in the sysctl tree.
1915The descriptions can be retrieved programmatically (see
1916.Xr sysctl 3 ) ,
1917or by the sysctl binary itself (see
1918.Xr sysctl 8 ) .
1919The descriptions are meant to give an indication of the purpose and/or
1920effects of a given node's value, not replace the documentation for the
1921given subsystem as a whole.
1922.El
1923.Ss System V IPC Options
1924.Bl -ohang
1925.It Cd options SYSVMSG
1926Includes support for
1927.At V
1928style message queues.
1929See
1930.Xr msgctl 2 ,
1931.Xr msgget 2 ,
1932.Xr msgrcv 2 ,
1933.Xr msgsnd 2 .
1934.It Cd options SYSVSEM
1935Includes support for
1936.At V
1937style semaphores.
1938See
1939.Xr semctl 2 ,
1940.Xr semget 2 ,
1941.Xr semop 2 .
1942.It Cd options SEMMNI=value
1943Sets the number of
1944.At V
1945style semaphore identifiers.
1946The GENERIC config file for your port will have the default.
1947.It Cd options SEMMNS=value
1948Sets the number of
1949.At V
1950style semaphores in the system.
1951The GENERIC config file for your port will have the default.
1952.It Cd options SEMUME=value
1953Sets the maximum number of undo entries per process for
1954.At V
1955style semaphores.
1956The GENERIC config file for your port will have the default.
1957.It Cd options SEMMNU=value
1958Sets the number of undo structures in the system for
1959.At V
1960style semaphores.
1961The GENERIC config file for your port will have the default.
1962.It Cd options SYSVSHM
1963Includes support for
1964.At V
1965style shared memory.
1966See
1967.Xr shmat 2 ,
1968.Xr shmctl 2 ,
1969.Xr shmdt 2 ,
1970.Xr shmget 2 .
1971.It Cd options SHMMAXPGS=value
1972Sets the maximum number of
1973.At V
1974style shared memory pages that are available through the
1975.Xr shmget 2
1976system call.
1977Default value is 1024 on most ports.
1978See
1979.Pa /usr/include/machine/vmparam.h
1980for the default.
1981.El
1982.Ss VM Related Options
1983.Bl -ohang
1984.It Cd options NMBCLUSTERS=value
1985The number of mbuf clusters the kernel supports.
1986Mbuf clusters are MCLBYTES in size (usually 2k).
1987This is used to compute the size of the kernel VM map
1988.Em mb_map ,
1989which maps mbuf clusters.
1990Default on most ports is 1024 (2048 with
1991.Dq options GATEWAY
1992).
1993See
1994.Pa /usr/include/machine/param.h
1995for exact default information.
1996Increase this value if you get
1997.Dq mclpool limit reached
1998messages.
1999.It Cd options NKMEMPAGES=value
2000.It Cd options NKMEMPAGES_MIN=value
2001.It Cd options NKMEMPAGES_MAX=value
2002Size of kernel VM map
2003.Em kmem_map ,
2004in PAGE_SIZE-sized chunks (the VM page size; this value may be read
2005from the
2006.Xr sysctl 8
2007variable
2008.Em hw.pagesize
2009).
2010This VM map is used to map the kernel malloc arena.
2011The kernel attempts to auto-size this map based on the amount of
2012physical memory in the system.
2013Platform-specific code may place bounds on this computed size,
2014which may be viewed with the
2015.Xr sysctl 8
2016variable
2017.Em vm.nkmempages .
2018See
2019.Pa /usr/include/machine/param.h
2020for the default upper and lower bounds.
2021The related options
2022.Sq NKMEMPAGES_MIN
2023and
2024.Sq NKMEMPAGES_MAX
2025allow the bounds to be overridden in the kernel configuration file.
2026These options are provided in the event the computed value is
2027insufficient resulting in an
2028.Dq out of space in kmem_map
2029panic.
2030.It Cd options SB_MAX=value
2031Sets the max size in bytes that a socket buffer is allowed to occupy.
2032The default is 256k, but sometimes it needs to be increased, for example
2033when using large TCP windows.
2034This option can be changed via
2035.Xr sysctl 8
2036as well.
2037.It Cd options SOMAXKVA=value
2038Sets the maximum size of kernel virtual memory that the socket buffers
2039are allowed to use.
2040The default is 16MB, but in situations where for example large TCP
2041windows are used this value must also be increased.
2042This option can be changed via
2043.Xr sysctl 8
2044as well.
2045.It Cd options BUFCACHE=value
2046Size of the buffer cache as a percentage of total available
2047.Tn RAM .
2048Ignored if BUFPAGES is also specified.
2049.It Cd options NBUF=value
2050Sets the number of buffer headers available, i.e., the number of
2051open files that may have a buffer cache entry.
2052Each buffer header
2053requires MAXBSIZE (machine dependent, but usually 65536) bytes.
2054The default value is machine dependent, but is usually equal to the
2055value of BUFPAGES.
2056If an architecture dependent VM_MAX_KERNEL_BUF constant is defined
2057then NBUF may be reduced at run time so that the storage allocated
2058for buffer headers doesn't exceed that limit.
2059.It Cd options BUFPAGES=value
2060These options set the number of pages available for the buffer cache.
2061Their default value is a machine dependent value, often calculated as
2062between 5% and 10% of total available
2063.Tn RAM .
2064.It Cd options MAXTSIZ=bytes
2065Sets the maximum size limit of a process' text segment.
2066See
2067.Pa /usr/include/machine/vmparam.h
2068for the port-specific default.
2069.It Cd options DFLDSIZ=bytes
2070Sets the default size limit of a process' data segment, the value that
2071will be returned as the soft limit for
2072.Dv RLIMIT_DATA
2073(as returned by
2074.Xr getrlimit 2 ) .
2075See
2076.Pa /usr/include/machine/vmparam.h
2077for the port-specific default.
2078.It Cd options MAXDSIZ=bytes
2079Sets the maximum size limit of a process' data segment, the value that
2080will be returned as the hard limit for
2081.Dv RLIMIT_DATA
2082(as returned by
2083.Xr getrlimit 2 ) .
2084See
2085.Pa /usr/include/machine/vmparam.h
2086for the port-specific default.
2087.It Cd options DFLSSIZ=bytes
2088Sets the default size limit of a process' stack segment, the value that
2089will be returned as the soft limit for
2090.Dv RLIMIT_STACK
2091(as returned by
2092.Xr getrlimit 2 ) .
2093See
2094.Pa /usr/include/machine/vmparam.h
2095for the port-specific default.
2096.It Cd options MAXSSIZ=bytes
2097Sets the maximum size limit of a process' stack segment, the value that
2098will be returned as the hard limit for
2099.Dv RLIMIT_STACK
2100(as returned by
2101.Xr getrlimit 2 ) .
2102See
2103.Pa /usr/include/machine/vmparam.h
2104for the port-specific default.
2105.It Cd options DUMP_ON_PANIC=integer
2106Defaults to one.
2107If set to zero, the kernel will not dump to the dump device when
2108it panics, though dumps can still be forced via
2109.Xr ddb 4
2110with the
2111.Dq sync
2112command.
2113Note that this sets the value of the
2114.Em kern.dump_on_panic
2115.Xr sysctl 3
2116variable which may be changed at run time -- see
2117.Xr sysctl 8
2118for details.
2119.It Cd options USE_TOPDOWN_VM
2120User space memory allocations (as made by
2121.Xr mmap 2 )
2122will be arranged in a
2123.Dq top down
2124fashion instead of the traditional
2125.Dq upwards from MAXDSIZ \+ vm_daddr
2126method.
2127This includes the placement of
2128.Xr ld.so 1 .
2129Arranging memory in this manner allows either (or both of) the heap or
2130.Xr mmap 2
2131allocated space to grow larger than traditionally possible.
2132This option is not available on all ports, but is instead expected to be
2133offered on a port-by-port basis, after which some ports will commit to
2134using it by default.
2135See the files
2136.Pa /usr/include/uvm/uvm_param.h
2137for some implementation details, and
2138.Pa /usr/include/machine/vmparam.h
2139for port specific details including availability.
2140.It Cd options VMSWAP
2141Enable paging device/file support.
2142This option is on by default.
2143.El
2144.Ss Security Options
2145.Bl -ohang
2146.It Cd options INSECURE
2147Hardwires the kernel security level at \-1.
2148This means that the system
2149always runs in secure level \-1 mode, even when running multiuser.
2150See the manual page for
2151.Xr init 8
2152for details on the implications of this.
2153The kernel secure level may manipulated by the superuser by altering the
2154.Em kern.securelevel
2155.Xr sysctl 3
2156variable (the secure level may only be lowered by a call from process ID 1,
2157i.e.,
2158.Xr init 8 ) .
2159See also
2160.Xr sysctl 8
2161and
2162.Xr sysctl 3 .
2163.It Cd options VERIFIED_EXEC
2164Enables Veriexec, the in-kernel file integrity subsystem.
2165.It Cd options VERIFIED_EXEC_FP_MD5
2166Enables support for MD5 hashes in Veriexec.
2167.It Cd options VERIFIED_EXEC_FP_SHA1
2168Enables support for SHA1 hashes in Veriexec.
2169.It Cd options VERIFIED_EXEC_FP_RMD160
2170Enables support for RMD160 hashes in Veriexec.
2171.It Cd options VERIFIED_EXEC_FP_SHA256
2172Enables support for SHA256 hashes in Veriexec.
2173.It Cd options VERIFIED_EXEC_FP_SHA384
2174Enables support for SHA384 hashes in Veriexec.
2175.It Cd options VERIFIED_EXEC_FP_SHA512
2176Enables support for SHA512 hashes in Veriexec.
2177.El
2178.Ss amiga-specific Options
2179.Bl -ohang
2180.It Cd options BB060STUPIDROM
2181When the bootloader (which passes
2182.Tn AmigaOS
2183.Tn ROM
2184information) claims we have a 68060
2185.Tn CPU
2186without
2187.Tn FPU ,
2188go look into the Processor Configuration Register (PCR) to find out.
2189You need this with
2190.Tn Amiga
2191.Tn ROM Ns s
2192up to (at least) V40.xxx (OS3.1),
2193when you boot via the bootblocks and don't have a DraCo.
2194.It Cd options IOBZCLOCK=frequency
2195The IOBlix boards come with two different serial master clocks: older ones
2196use 24 MHz, newer ones use 22.1184 MHz.
2197The driver normally assumes the latter.
2198If your board uses 24 MHz, you can recompile your kernel with
2199options IOBZCLOCK=24000000
2200or patch the kernel variable
2201.Tn iobzclock
2202to the same value.
2203.It Cd options LIMITMEM=value
2204If there, limit the part of the first memory bank used by
2205.Nx
2206to value megabytes.
2207Default is unlimited.
2208.It Cd options NKPTADD=addvalue
2209.It Cd options NKPTADDSHIFT=shiftvalue
2210The
2211.Tn CPU
2212specific
2213.Tn MMU
2214table for the kernel is pre-allocated at kernel startup time.
2215Part of it is scaled with
2216.Va maxproc ,
2217to have enough room to hold the user program
2218.Tn MMU
2219tables; the second part is a fixed amount for the kernel itself.
2220.Pp
2221The third part accounts for the size of the file buffer cache.
2222Its size is either
2223.Dv NKPTADD
2224pages (if defined) or memory size in bytes divided by two to
2225the power of
2226.Dv NKPTADDSHIFT .
2227The default is undefined
2228.Dv NKPTADD
2229and
2230.Dv NKPTADDSHIFT=24 ,
2231allowing for 16 buffers per megabyte of main memory (while
2232a GENERIC kernel allocates about half of that).
2233When you get "can't get KPT page" panics, you should increase
2234.Dv NKPTADD
2235(if defined), or decrease
2236.Dv NKPTADDSHIFT
2237by one.
2238.It Cd options P5PPC68KBOARD
2239Add special support for Phase5 mixed 68k+PPC boards.
2240Currently, this only affects rebooting from
2241.Nx
2242and is only needed on 68040+PPC, not on
224368060+PPC; without this, affected machines will hang after
2244.Nx
2245has shut
2246down and will only restart after a keyboard reset or a power cycle.
2247.El
2248.Ss arm32-specific Options
2249.Bl -ohang
2250.It Cd options FRENCH_KBD
2251Include translation for French keyboards when using
2252.Em pccons
2253on a Shark.
2254.It Cd options FINNISH_KBD
2255Include translation for Finnish keyboards when using
2256.Em pccons
2257on a Shark.
2258.It Cd options GERMAN_KBD
2259Include translation for German keyboards when using
2260.Em pccons
2261on a Shark.
2262.It Cd options NORWEGIAN_KBD
2263Include translation for French keyboards when using
2264.Em pccons
2265on a Shark.
2266.El
2267.Ss atari-specific Options
2268.Bl -ohang
2269.It Cd options DISKLABEL_AHDI
2270Include support for AHDI (native Atari) disklabels.
2271.It Cd options DISKLABEL_NBDA
2272Include support for
2273.Nx Ns Tn /atari
2274labels.
2275If you don't set this option, it will be set automatically.
2276.Nx Ns Tn /atari
2277will not work without it.
2278.It Cd options FALCON_SCSI
2279Include support for the 5380-SCSI configuration as found on the Falcon.
2280.It Cd options RELOC_KERNEL
2281If set, the kernel will relocate itself to TT-RAM, if possible.
2282This will give you a slightly faster system.
2283.Em Beware
2284that on some TT030 systems,
2285the system will frequently dump with MMU-faults with this option enabled.
2286.It Cd options SERCONSOLE
2287Allow the modem1-port to act as the system-console.
2288A carrier should be active on modem1 during system boot to active
2289the console functionality.
2290.It Cd options TT_SCSI
2291Include support for the 5380-SCSI configuration as found on the TT030
2292and Hades.
2293.El
2294.Ss i386-specific Options
2295.Bl -ohang
2296.It Cd options I386_CPU,I486_CPU,I586_CPU,I686_CPU
2297Include support for a particular class of
2298.Tn CPU
2299.Po
2300.Tn i386 ,
2301.Tn i486 ,
2302.Tn Pentium ,
2303or
2304.Tn Pentium Pro
2305.Pc .
2306If the appropriate class for your
2307.Tn CPU
2308is not configured, the kernel will use the highest class available
2309that will work.
2310In general, using the correct
2311.Tn CPU
2312class will result in the best performance.
2313At least one of these options must be present.
2314.It Cd options ENHANCED_SPEEDSTEP
2315Include support for the
2316.Tn Enhanced SpeedStep Technology
2317present in newer
2318.Tn CPU Ns s .
2319.It Cd options EST_FREQ_USERWRITE
2320Allow any user to change the frequency of an
2321.Tn Enhanced SpeedStep Technology
2322capable
2323.Tn CPU .
2324.It Cd options POWERNOW_K7
2325Include support for the AMD
2326.Tn PowerNow! Technology
2327present in AMD Athlon Mobile processors.
2328.It Cd options CPURESET_DELAY=value
2329Specifies the time (in millisecond) to wait before doing a hardware reset
2330in the last phase of a reboot.
2331This gives the user a chance to see error messages from the shutdown
2332operations (like NFS unmounts, buffer cache flush, etc ...).
2333Setting this to 0 will disable the delay.
2334Default is 2 seconds.
2335.It Cd options MATH_EMULATE
2336Include the floating point emulator.
2337This is useful only for
2338.Tn CPU Ns s
2339that lack an
2340internal Floating Point Unit
2341.Pq Tn FPU
2342or co-processor.
2343.It Cd options VM86
2344Include support for virtual 8086 mode, used by
2345.Tn DOS
2346emulators and X servers to run BIOS code, e.g., for some VESA routines.
2347.It Cd options USER_LDT
2348Include i386-specific system calls for modifying the local descriptor table,
2349used by Windows emulators.
2350.It Cd options REALBASEMEM=integer
2351Overrides the base memory size passed in from the boot block.
2352(Value given in kilobytes.)
2353Use this option only if the boot block reports the size incorrectly.
2354(Note that some
2355.Tn BIOS Ns es
2356put the extended
2357.Tn BIOS
2358data area at the top of base memory, and therefore report a smaller
2359base memory size to prevent programs overwriting it.
2360This is correct behavior, and you should not use the
2361.Em REALBASEMEM
2362option to access this memory).
2363.It Cd options REALEXTMEM=integer
2364Overrides the extended memory size passed in from the boot block.
2365(Value given in kilobytes.
2366Extended memory does not include the first megabyte.)
2367Use this option only if the boot block reports the size incorrectly.
2368.It Cd options FRENCH_KBD,FINNISH_KBD,GERMAN_KBD,NORWEGIAN_KBD
2369Select a non-US keyboard layout for the
2370.Em pccons
2371console driver.
2372.It Cd options CYRIX_CACHE_WORKS
2373Relevant only to the Cyrix 486DLC CPU.
2374This option is used to turn on the cache in hold-flush mode.
2375It is not turned on by default because it is known to have problems in
2376certain motherboard implementations.
2377.It Cd options CYRIX_CACHE_REALLY_WORKS
2378Relevant only to the Cyrix 486DLC CPU.
2379This option is used to turn on the cache in write-back mode.
2380It is not turned on by default because it is known to have problems in
2381certain motherboard implementations.
2382In order for this option to take effect, option
2383.Em CYRIX_CACHE_WORKS
2384must also be specified.
2385.It Cd options PCIBIOS
2386Enable support for initializing the
2387.Tn PCI
2388bus using information from the
2389.Tn BIOS .
2390See
2391.Xr pcibios 4
2392for details.
2393.It Cd options KSTACK_CHECK_DR0
2394Detect kernel stack overflow using DR0 register.
2395This option uses DR0 register exclusively so you can't use DR0 register for
2396other purpose (e.g., hardware breakpoint) if you turn this on.
2397.It Cd options MTRR
2398Include support for accessing MTRR registers from user-space.
2399See
2400.Xr i386_get_mtrr 2 .
2401.It Cd options BEEP_ONHALT
2402Make the system speaker emit several beeps when it is completely safe to
2403power down the computer after a
2404.Xr halt 8
2405command.
2406Requires
2407.Xr sysbeep 4
2408support.
2409.It Cd options BEEP_ONHALT_COUNT=times
2410Number of times to beep the speaker when
2411.Cd options BEEP_ONHALT
2412is enabled.
2413Defaults to 3.
2414.It Cd options BEEP_ONHALT_PITCH=hz
2415The tone frequency used when
2416.Cd options BEEP_ONHALT
2417option, in hertz.
2418Defaults to 1500.
2419.It Cd options BEEP_ONHALT_PERIOD=msecs
2420The duration of each beep when
2421.Cd options BEEP_ONHALT
2422is enabled, in milliseconds.
2423Defaults to 250.
2424.It Cd options MULTIBOOT
2425Makes the kernel Multiboot-compliant, allowing it to be booted through
2426a Multiboot-compliant boot manager such as GRUB.
2427See
2428.Xr multiboot 8
2429for more information.
2430.It Cd options MULTIBOOT_SYMTAB_SPACE=nbytes
2431Space (in bytes) reserved in memory to store the kernel's symbol table.
2432If the symbol table is bigger than this space, it will not be loaded;
2433if it is smaller, the unused memory will be claimed by the kernel and
2434used normally.
2435You shouldn't need to touch this variable.
2436Defaults to 1048576 bytes (one megabyte).
2437.El
2438.Ss isa-specific Options
2439Options specific to
2440.Xr isa 4
2441busses.
2442.Bl -ohang
2443.It Cd options PCIC_ISA_ALLOC_IOBASE=address, PCIC_ISA_ALLOC_IOSIZE=size
2444Control the section of IO bus space used for PCMCIA bus space mapping.
2445Ideally the probed defaults are satisfactory, however in practice
2446that is not always the case.
2447See
2448.Xr pcmcia 4
2449for details.
2450.It Cd options PCIC_ISA_INTR_ALLOC_MASK=mask
2451Controls the allowable interrupts that may be used for
2452.Tn PCMCIA
2453devices.
2454This mask is a logical-or of power-of-2s of allowable interrupts:
2455.Bd -literal -offset 04n
2456.Em "IRQ Val      IRQ Val      IRQ Val       IRQ Val"
2457 0  0x0001    4  0x0010    8  0x0100    12  0x1000
2458 1  0x0002    5  0x0020    9  0x0200    13  0x2000
2459 2  0x0004    6  0x0040   10  0x0400    14  0x4000
2460 3  0x0008    7  0x0080   11  0x0800    15  0x8000
2461.Ed
2462.It Cd options PCKBC_CNATTACH_SELFTEST
2463Perform a self test of the keyboard controller before attaching it as a
2464console.
2465This might be necessary on machines where we boot on cold iron, and
2466pckbc refuses to talk until we request a self test.
2467Currently only the netwinder port uses it.
2468.It Cd options PCKBD_CNATTACH_MAY_FAIL
2469If this option is set the PS/2 keyboard will not be used as the console
2470if it cannot be found during boot.
2471This allows other keyboards, like USB, to be the console keyboard.
2472.It Cd options PCKBD_LAYOUT=layout
2473Sets the default keyboard layout, see
2474.Xr pckbd 4 .
2475.El
2476.Ss m68k-specific Options
2477.Bl -ohang
2478.It Cd options FPU_EMULATE
2479Include support for MC68881/MC68882 emulator.
2480.It Cd options FPSP
2481Include support for 68040 floating point.
2482.It Cd options M68020,M68030,M68040,M68060
2483Include support for a specific
2484.Tn CPU ,
2485at least one (the one you are using) should be specified.
2486.It Cd options M060SP
2487Include software support for 68060.
2488This provides emulation of unimplemented
2489integer instructions as well as emulation of unimplemented floating point
2490instructions and data types and software support for floating point traps.
2491.El
2492.Ss powerpc-specific Options
2493.Bl -ohang
2494.It Cd options PMAP_MEMLIMIT=value
2495Limit the amount of memory seen by the kernel to
2496.Ar value
2497bytes (OEA only).
2498.El
2499.Ss sparc-specific Options
2500.Bl -ohang
2501.It Cd options AUDIO_DEBUG
2502Enable simple event debugging of the logging of the
2503.Xr audio 4
2504device.
2505.It Cd options BLINK
2506Enable blinking of LED.
2507Blink rate is full cycle every N seconds for
2508N \*[Lt] then current load average.
2509See
2510.Xr getloadavg 3 .
2511.\" .It Cd options COLORFONT_CACHE
2512.\" What does this do?
2513.It Cd options COUNT_SW_LEFTOVERS
2514Count how many times the sw SCSI device has left 3, 2, 1 and 0 in the
2515sw_3_leftover, sw_2_leftover, sw_1_leftover, and sw_0_leftover
2516variables accessible from
2517.Xr ddb 4 .
2518See
2519.Xr sw 4 .
2520.It Cd options DEBUG_ALIGN
2521Adds debugging messages calls when user-requested alignment fault
2522handling happens.
2523.It Cd options DEBUG_EMUL
2524Adds debugging messages calls for emulated floating point and
2525alignment fixing operations.
2526.It Cd options DEBUG_SVR4
2527Prints registers messages calls for emulated SVR4 getcontext and
2528setcontext operations.
2529See
2530.Em options COMPAT_SVR4 .
2531.It Cd options EXTREME_DEBUG
2532Adds debugging functions callable from
2533.Xr ddb 4 .
2534The debug_pagetables, test_region and print_fe_map
2535functions print information about page tables for the SUN4M
2536platforms only.
2537.It Cd options EXTREME_EXTREME_DEBUG
2538Adds extra info to
2539.Em options EXTREME_DEBUG .
2540.It Cd options FPU_CONTEXT
2541Make
2542.Em options COMPAT_SVR4
2543getcontext and setcontext include floating point registers.
2544.It Cd options MAGMA_DEBUG
2545Adds debugging messages to the
2546.Xr magma 4
2547device.
2548.It Cd options RASTERCONS_FULLSCREEN
2549Use the entire screen for the console.
2550.It Cd options RASTERCONS_SMALLFONT
2551Use the Fixed font on the console, instead of the normal font.
2552.It Cd options SUN4
2553Support sun4 class machines.
2554.It Cd options SUN4C
2555Support sun4c class machines.
2556.It Cd options SUN4M
2557Support sun4m class machines.
2558.It Cd options SUN4_MMU3L
2559.\" XXX ???
2560Enable support for sun4 3-level MMU machines.
2561.It Cd options V9
2562Enable SPARC V9 assembler in
2563.Xr ddb 4 .
2564.El
2565.Ss sparc64-specific Options
2566.Bl -ohang
2567.It Cd options AUDIO_DEBUG
2568Enable simple event debugging of the logging of the
2569.Xr audio 4
2570device.
2571.It Cd options BLINK
2572Enable blinking of LED.
2573Blink rate is full cycle every N seconds for
2574N \*[Lt] then current load average.
2575See
2576.Xr getloadavg 3 .
2577.El
2578.Ss x68k-specific Options
2579.Bl -ohang
2580.It Cd options EXTENDED_MEMORY
2581Include support for extended memory, e.g., TS-6BE16 and 060turbo on-board.
2582.It Cd options JUPITER
2583Include support for Jupiter-X MPU accelerator
2584.It Cd options ZSCONSOLE,ZSCN_SPEED=value
2585Use the built-in serial port as the system-console.
2586Speed is specified in bps, defaults to 9600.
2587.It Cd options ITE_KERNEL_ATTR=value
2588Set the kernel message attribute for ITE.
2589Value, an integer, is a logical or of the following values:
2590.Bl -tag -width 4n -compact -offset indent
2591.It 1
2592color inversed
2593.It 2
2594underlined
2595.It 4
2596bolded
2597.El
2598.El
2599.\" The following requests should be uncommented and used where appropriate.
2600.\" .Sh FILES
2601.\" .Sh EXAMPLES
2602.Sh SEE ALSO
2603.Xr config 1 ,
2604.Xr gdb 1 ,
2605.Xr ktrace 1 ,
2606.Xr pmc 1 ,
2607.Xr quota 1 ,
2608.Xr vndcompress 1 ,
2609.Xr gettimeofday 2 ,
2610.Xr i386_get_mtrr 2 ,
2611.Xr i386_iopl 2 ,
2612.Xr msgctl 2 ,
2613.Xr msgget 2 ,
2614.Xr msgrcv 2 ,
2615.Xr msgsnd 2 ,
2616.Xr ntp_adjtime 2 ,
2617.Xr ntp_gettime 2 ,
2618.Xr semctl 2 ,
2619.Xr semget 2 ,
2620.Xr semop 2 ,
2621.Xr shmat 2 ,
2622.Xr shmctl 2 ,
2623.Xr shmdt 2 ,
2624.Xr shmget 2 ,
2625.Xr sysctl 3 ,
2626.Xr apm 4 ,
2627.Xr ddb 4 ,
2628.Xr inet 4 ,
2629.Xr iso 4 ,
2630.Xr lkm 4 ,
2631.Xr md 4 ,
2632.Xr ns 4 ,
2633.Xr pcibios 4 ,
2634.Xr pcmcia 4 ,
2635.Xr ppp 4 ,
2636.Xr userconf 4 ,
2637.Xr vnd 4 ,
2638.Xr wscons 4 ,
2639.Xr edquota 8 ,
2640.Xr init 8 ,
2641.Xr mdsetimage 8 ,
2642.Xr mount_cd9660 8 ,
2643.Xr mount_fdesc 8 ,
2644.Xr mount_kernfs 8 ,
2645.Xr mount_lfs 8 ,
2646.Xr mount_mfs 8 ,
2647.Xr mount_msdos 8 ,
2648.Xr mount_nfs 8 ,
2649.Xr mount_ntfs 8 ,
2650.Xr mount_null 8 ,
2651.Xr mount_portal 8 ,
2652.Xr mount_procfs 8 ,
2653.Xr mount_udf 8 ,
2654.Xr mount_umap 8 ,
2655.Xr mount_union 8 ,
2656.Xr mrouted 8 ,
2657.Xr newfs_lfs 8 ,
2658.Xr ntpd 8 ,
2659.Xr quotaon 8 ,
2660.Xr rpc.rquotad 8 ,
2661.Xr sysctl 8
2662.Sh HISTORY
2663The
2664.Nm
2665man page first appeared in
2666.Nx 1.3 .
2667.Sh BUGS
2668The
2669.Em EON
2670option should be a pseudo-device, and is also very fragile.
2671