sysctl.9

.\"
.\" Copyright (c) 2006 Robert N. M. Watson
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\"    notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\"    notice, this list of conditions and the following disclaimer in the
.\"    documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.Dd September 1, 2020
.Dt SYSCTL 9
.Os
.Sh NAME
.Nm SYSCTL_DECL ,
.Nm SYSCTL_ADD_BOOL ,
.Nm SYSCTL_ADD_COUNTER_U64 ,
.Nm SYSCTL_ADD_COUNTER_U64_ARRAY ,
.Nm SYSCTL_ADD_INT ,
.Nm SYSCTL_ADD_LONG ,
.Nm SYSCTL_ADD_NODE ,
.Nm SYSCTL_ADD_NODE_WITH_LABEL ,
.Nm SYSCTL_ADD_OPAQUE ,
.Nm SYSCTL_ADD_PROC ,
.Nm SYSCTL_ADD_QUAD ,
.Nm SYSCTL_ADD_ROOT_NODE ,
.Nm SYSCTL_ADD_S8 ,
.Nm SYSCTL_ADD_S16 ,
.Nm SYSCTL_ADD_S32 ,
.Nm SYSCTL_ADD_S64 ,
.Nm SYSCTL_ADD_SBINTIME_MSEC ,
.Nm SYSCTL_ADD_SBINTIME_USEC ,
.Nm SYSCTL_ADD_STRING ,
.Nm SYSCTL_ADD_CONST_STRING ,
.Nm SYSCTL_ADD_STRUCT ,
.Nm SYSCTL_ADD_TIMEVAL_SEC ,
.Nm SYSCTL_ADD_U8 ,
.Nm SYSCTL_ADD_U16 ,
.Nm SYSCTL_ADD_U32 ,
.Nm SYSCTL_ADD_U64 ,
.Nm SYSCTL_ADD_UAUTO ,
.Nm SYSCTL_ADD_UINT ,
.Nm SYSCTL_ADD_ULONG ,
.Nm SYSCTL_ADD_UMA_CUR ,
.Nm SYSCTL_ADD_UMA_MAX ,
.Nm SYSCTL_ADD_UQUAD ,
.Nm SYSCTL_CHILDREN ,
.Nm SYSCTL_STATIC_CHILDREN ,
.Nm SYSCTL_NODE_CHILDREN ,
.Nm SYSCTL_PARENT ,
.Nm SYSCTL_BOOL ,
.Nm SYSCTL_COUNTER_U64 ,
.Nm SYSCTL_COUNTER_U64_ARRAY ,
.Nm SYSCTL_INT ,
.Nm SYSCTL_INT_WITH_LABEL ,
.Nm SYSCTL_LONG ,
.Nm sysctl_msec_to_ticks ,
.Nm SYSCTL_NODE ,
.Nm SYSCTL_NODE_WITH_LABEL ,
.Nm SYSCTL_OPAQUE ,
.Nm SYSCTL_PROC ,
.Nm SYSCTL_QUAD ,
.Nm SYSCTL_ROOT_NODE ,
.Nm SYSCTL_S8 ,
.Nm SYSCTL_S16 ,
.Nm SYSCTL_S32 ,
.Nm SYSCTL_S64 ,
.Nm SYSCTL_SBINTIME_MSEC ,
.Nm SYSCTL_SBINTIME_USEC ,
.Nm SYSCTL_STRING ,
.Nm SYSCTL_CONST_STRING ,
.Nm SYSCTL_STRUCT ,
.Nm SYSCTL_TIMEVAL_SEC ,
.Nm SYSCTL_U8 ,
.Nm SYSCTL_U16 ,
.Nm SYSCTL_U32 ,
.Nm SYSCTL_U64 ,
.Nm SYSCTL_UINT ,
.Nm SYSCTL_ULONG ,
.Nm SYSCTL_UMA_CUR ,
.Nm SYSCTL_UMA_MAX ,
.Nm SYSCTL_UQUAD
.Nd Dynamic and static sysctl MIB creation functions
.Sh SYNOPSIS
.In sys/param.h
.In sys/sysctl.h
.Fn SYSCTL_DECL name
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_BOOL
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "bool *ptr"
.Fa "uint8_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_COUNTER_U64
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "counter_u64_t *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_COUNTER_U64_ARRAY
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "counter_u64_t *ptr"
.Fa "intmax_t len"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_INT
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int *ptr"
.Fa "int val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_LONG
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "long *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_NODE
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int (*handler)(SYSCTL_HANDLER_ARGS)"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_NODE_WITH_LABEL
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int (*handler)(SYSCTL_HANDLER_ARGS)"
.Fa "const char *descr"
.Fa "const char *label"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_OPAQUE
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "void *ptr"
.Fa "intptr_t len"
.Fa "const char *format"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_PROC
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "void *arg1"
.Fa "intptr_t arg2"
.Fa "int (*handler) (SYSCTL_HANDLER_ARGS)"
.Fa "const char *format"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_QUAD
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int64_t *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_ROOT_NODE
.Fa "struct sysctl_ctx_list *ctx"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int (*handler)(SYSCTL_HANDLER_ARGS)"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_S8
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int8_t *ptr"
.Fa "int8_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_S16
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int16_t *ptr"
.Fa "int16_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_S32
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int32_t *ptr"
.Fa "int32_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_S64
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "int64_t *ptr"
.Fa "int64_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_SBINTIME_MSEC
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "sbintime_t *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_SBINTIME_USEC
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "sbintime_t *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_STRING
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "char *ptr"
.Fa "intptr_t len"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_CONST_STRING
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "const char *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_STRUCT
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "void *ptr"
.Fa struct_type
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_TIMEVAL_SEC
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "struct timeval *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_U8
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "uint8_t *ptr"
.Fa "uint8_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_U16
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "uint16_t *ptr"
.Fa "uint16_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_U32
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "uint32_t *ptr"
.Fa "uint32_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_U64
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "uint64_t *ptr"
.Fa "uint64_t val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_UINT
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "unsigned int *ptr"
.Fa "unsigned int val"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_ULONG
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "unsigned long *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_UQUAD
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "uint64_t *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_UMA_CUR
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "uma_zone_t ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_UMA_MAX
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "uma_zone_t ptr"
.Fa "const char *descr"
.Fc
.Fa "const char *descr"
.Ft struct sysctl_oid *
.Fo SYSCTL_ADD_UAUTO
.Fa "struct sysctl_ctx_list *ctx"
.Fa "struct sysctl_oid_list *parent"
.Fa "int number"
.Fa "const char *name"
.Fa "int ctlflags"
.Fa "void *ptr"
.Fa "const char *descr"
.Fc
.Ft struct sysctl_oid_list *
.Fo SYSCTL_CHILDREN
.Fa "struct sysctl_oid *oidp"
.Fc
.Ft struct sysctl_oid_list *
.Fo SYSCTL_STATIC_CHILDREN
.Fa "struct sysctl_oid_list OID_NAME"
.Fc
.Ft struct sysctl_oid_list *
.Fo SYSCTL_NODE_CHILDREN
.Fa "parent"
.Fa "name"
.Fc
.Ft struct sysctl_oid *
.Fo SYSCTL_PARENT
.Fa "struct sysctl_oid *oid"
.Fc
.Fn SYSCTL_BOOL parent number name ctlflags ptr val descr
.Fn SYSCTL_COUNTER_U64 parent number name ctlflags ptr descr
.Fn SYSCTL_COUNTER_U64_ARRAY parent number name ctlflags ptr len descr
.Fn SYSCTL_INT parent number name ctlflags ptr val descr
.Fn SYSCTL_INT_WITH_LABEL parent number name ctlflags ptr val descr label
.Fn SYSCTL_LONG parent number name ctlflags ptr val descr
.Ft int
.Fn sysctl_msec_to_ticks SYSCTL_HANDLER_ARGS
.Fn SYSCTL_NODE parent number name ctlflags handler descr
.Fn SYSCTL_NODE_WITH_LABEL parent number name ctlflags handler descr label
.Fn SYSCTL_OPAQUE parent number name ctlflags ptr len format descr
.Fn SYSCTL_PROC parent number name ctlflags arg1 arg2 handler format descr
.Fn SYSCTL_QUAD parent number name ctlflags ptr val descr
.Fn SYSCTL_ROOT_NODE number name ctlflags handler descr
.Fn SYSCTL_S8 parent number name ctlflags ptr val descr
.Fn SYSCTL_S16 parent number name ctlflags ptr val descr
.Fn SYSCTL_S32 parent number name ctlflags ptr val descr
.Fn SYSCTL_S64 parent number name ctlflags ptr val descr
.Fn SYSCTL_SBINTIME_MSEC parent number name ctlflags ptr descr
.Fn SYSCTL_SBINTIME_USEC parent number name ctlflags ptr descr
.Fn SYSCTL_STRING parent number name ctlflags arg len descr
.Fn SYSCTL_CONST_STRING parent number name ctlflags arg descr
.Fn SYSCTL_STRUCT parent number name ctlflags ptr struct_type descr
.Fn SYSCTL_TIMEVAL_SEC parent number name ctlflags ptr descr
.Fn SYSCTL_U8 parent number name ctlflags ptr val descr
.Fn SYSCTL_U16 parent number name ctlflags ptr val descr
.Fn SYSCTL_U32 parent number name ctlflags ptr val descr
.Fn SYSCTL_U64 parent number name ctlflags ptr val descr
.Fn SYSCTL_UINT parent number name ctlflags ptr val descr
.Fn SYSCTL_ULONG parent number name ctlflags ptr val descr
.Fn SYSCTL_UQUAD parent number name ctlflags ptr val descr
.Fn SYSCTL_UMA_MAX parent number name ctlflags ptr descr
.Fn SYSCTL_UMA_CUR parent number name ctlflags ptr descr
.Sh DESCRIPTION
The
.Nm SYSCTL
kernel interface allows dynamic or static creation of
.Xr sysctl 8
MIB entries.
All static sysctls are automatically destroyed when the module which
they are part of is unloaded.
Most top level categories are created statically and are available to
all kernel code and its modules.
.Sh DESCRIPTION OF ARGUMENTS
.Bl -tag -width ctlflags
.It Fa ctx
Pointer to sysctl context or NULL, if no context.
See
.Xr sysctl_ctx_init 9
for how to create a new sysctl context.
Programmers are strongly advised to use contexts to organize the
dynamic OIDs which they create because when a context is destroyed all
belonging sysctls are destroyed as well.
This makes the sysctl cleanup code much simpler.
Else deletion of all created OIDs is required at module unload.
.It Fa parent
A pointer to a
.Li struct sysctl_oid_list ,
which is the head of the parent's list of children.
This pointer is retrieved using the
.Fn SYSCTL_STATIC_CHILDREN
macro for static sysctls and the
.Fn SYSCTL_CHILDREN
macro for dynamic sysctls.
The
.Fn SYSCTL_PARENT
macro can be used to get the parent of an OID.
The macro returns NULL if there is no parent.
.It Fa number
The OID number that will be assigned to this OID.
In almost all cases this should be set to
.Dv OID_AUTO ,
which will result in the assignment of the next available OID number.
.It Fa name
The name of the OID.
The newly created OID will contain a copy of the name.
.It Fa ctlflags
A bit mask of sysctl control flags.
See the section below describing all the control flags.
.It Fa arg1
First callback argument for procedure sysctls.
.It Fa arg2
Second callback argument for procedure sysctls.
.It Fa len
The length of the data pointed to by the
.Fa ptr
argument.
For string type OIDs a length of zero means that
.Xr strlen 3
will be used to get the length of the string at each access to the OID.
For array type OIDs the length must be greater than zero.
.It Fa ptr
Pointer to sysctl variable or string data.
For sysctl values the pointer can be SYSCTL_NULL_XXX_PTR which means the OID is read-only and the returned value should be taken from the
.Fa val
argument.
.It Fa val
If the
.Fa ptr
argument is SYSCTL_NULL_XXX_PTR, gives the constant value returned by this OID.
Else this argument is not used.
.It Fa struct_type
Name of structure type.
.It Fa handler
A pointer to the function
that is responsible for handling read and write requests
to this OID.
There are several standard handlers
that support operations on nodes,
integers, strings and opaque objects.
It is possible to define custom handlers using the
.Fn SYSCTL_PROC
macro or the
.Fn SYSCTL_ADD_PROC
function.
.It Fa format
A pointer to a string
which specifies the format of the OID in a symbolic way.
This format is used as a hint by
.Xr sysctl 8
to apply proper data formatting for display purposes.
.Pp
Current formats:
.Bl -tag -width "S,TYPE" -compact -offset indent
.It Cm N
node
.It Cm A
.Li "char *"
.It Cm C
.Li "int8_t"
.It Cm CU
.Li "uint8_t"
.It Cm I
.Li "int"
.It Cm IK Ns Op Ar n
temperature in Kelvin, multiplied by an optional single digit
power of ten scaling factor: 1 (default) gives deciKelvin, 0 gives Kelvin, 3
gives milliKelvin
.It Cm IU
.Li "unsigned int"
.It Cm L
.Li "long"
.It Cm LU
.Li "unsigned long"
.It Cm Q
.Li "quad_t"
.It Cm QU
.Li "u_quad_t"
.It Cm S
.Li "int16_t"
.It Cm SU
.Li "uint16_t"
.It Cm "S,TYPE"
.Li "struct TYPE"
structures
.El
.It Fa descr
A pointer to a textual description of the OID.
.It Fa label
A pointer to an aggregation label for this component of the OID.
To make it easier to export sysctl data to monitoring systems that
support aggregations through labels (e.g., Prometheus),
this argument can be used to attach a label name to an OID.
The label acts as a hint that this component's name should not be part
of the metric's name,
but attached to the metric as a label instead.
.Pp
Labels should only be applied to siblings that are structurally similar
and encode the same type of value,
as aggregation is of no use otherwise.
.El
.Sh NODE VALUE TYPES
Most of the macros and functions used to create sysctl nodes export a
read-only constant or in-kernel variable whose type matches the type
of the node's value.
For example,
.Fn SYSCTL_INT
reports the raw value of an associated variable of type
.Vt int .
However, nodes may also export a value that is a translation of an internal
representation.
.Pp
The
.Fn sysctl_msec_to_ticks
handler can be used with
.Fn SYSCTL_PROC
or
.Fn SYSCTL_ADD_PROC
to export a millisecond time interval.
When using this handler,
the
.Fa arg2
parameter points to an in-kernel variable of type
.Vt int
which stores a tick count suitable for use with functions like
.Xr tsleep 9 .
The
.Fn sysctl_msec_to_ticks
function converts this value to milliseconds when reporting the node's value.
Similarly,
.Fn sysctl_msec_to_ticks
accepts new values in milliseconds and stores an equivalent value in ticks to
.Fa *arg2 .
Note that new code should use kernel variables of type
.Vt sbintime_t
instead of tick counts.
.Pp
The
.Fn SYSCTL_ADD_SBINTIME_MSEC
and
.Fn SYSCTL_ADD_SBINTIME_USEC
functions and
.Fn SYSCTL_SBINTIME_MSEC
and
.Fn SYSCTL_SBINTIME_USEC
macros all create nodes which export an in-kernel variable of type
.Vt sbintime_t .
These nodes do not export the raw value of the associated variable.
Instead, they export a 64-bit integer containing a count of either
milliseconds (the MSEC variants) or microseconds (the USEC variants).
.Pp
The
.Fn SYSCTL_ADD_TIMEVAL_SEC
function and
.Fn SYSCTL_TIMEVAL_SEC
macro create nodes which export an in-kernel variable of type
.Vt struct timeval .
These nodes do not export full value of the associated structure.
Instead, they export a count in seconds as a simple integer which is
stored in the
.Fa tv_sec
field of the associated variable.
This function and macro are intended to be used with variables which
store a non-negative interval rather than an absolute time.
As a result, they reject attempts to store negative values.
.Sh CREATING ROOT NODES
Sysctl MIBs or OIDs are created in a hierarchical tree.
The nodes at the bottom of the tree are called root nodes, and have no
parent OID.
To create bottom tree nodes the
.Fn SYSCTL_ROOT_NODE
macro or the
.Fn SYSCTL_ADD_ROOT_NODE
function needs to be used.
By default all static sysctl node OIDs are global and need a
.Fn SYSCTL_DECL
statement prior to their
.Fn SYSCTL_NODE
definition statement, typically in a so-called header file.
.Sh CREATING SYSCTL STRINGS
Zero terminated character strings sysctls are created either using the
.Fn SYSCTL_STRING
macro or the
.Fn SYSCTL_ADD_STRING
function.
If the
.Fa len
argument is zero, the string length is computed at every access to the OID using
.Xr strlen 3 .
Use the
.Fn SYSCTL_CONST_STRING
macro or the
.Fn SYSCTL_ADD_CONST_STRING
function to add a sysctl for a constant string.
.Sh CREATING OPAQUE SYSCTLS
The
.Fn SYSCTL_OPAQUE
or
.Fn SYSCTL_STRUCT
macros or the
.Fn SYSCTL_ADD_OPAQUE
or
.Fn SYSCTL_ADD_STRUCT
functions create an OID that handle any chunk of data
of the size specified by the
.Fa len
argument and data pointed to by the
.Fa ptr
argument.
When using the structure version the type is encoded as part of the
created sysctl.
.Sh CREATING CUSTOM SYSCTLS
The
.Fn SYSCTL_PROC
macro and the
.Fn SYSCTL_ADD_PROC
function
create OIDs with the specified
.Pa handler
function.
The handler is responsible for handling all read and write requests to
the OID.
This OID type is especially useful if the kernel data is not easily
accessible, or needs to be processed before exporting.
.Sh CREATING A STATIC SYSCTL
Static sysctls are declared using one of the
.Fn SYSCTL_BOOL ,
.Fn SYSCTL_COUNTER_U64 ,
.Fn SYSCTL_COUNTER_U64_ARRAY ,
.Fn SYSCTL_INT ,
.Fn SYSCTL_INT_WITH_LABEL ,
.Fn SYSCTL_LONG ,
.Fn SYSCTL_NODE ,
.Fn SYSCTL_NODE_WITH_LABEL ,
.Fn SYSCTL_OPAQUE ,
.Fn SYSCTL_PROC ,
.Fn SYSCTL_QUAD ,
.Fn SYSCTL_ROOT_NODE ,
.Fn SYSCTL_S8 ,
.Fn SYSCTL_S16 ,
.Fn SYSCTL_S32 ,
.Fn SYSCTL_S64 ,
.Fn SYSCTL_SBINTIME_MSEC ,
.Fn SYSCTL_SBINTIME_USEC ,
.Fn SYSCTL_STRING ,
.Fn SYSCTL_CONST_STRING ,
.Fn SYSCTL_STRUCT ,
.Fn SYSCTL_TIMEVAL_SEC ,
.Fn SYSCTL_U8 ,
.Fn SYSCTL_U16 ,
.Fn SYSCTL_U32 ,
.Fn SYSCTL_U64 ,
.Fn SYSCTL_UINT ,
.Fn SYSCTL_ULONG ,
.Fn SYSCTL_UQUAD ,
.Fn SYSCTL_UMA_CUR
or
.Fn SYSCTL_UMA_MAX
macros.
.Sh CREATING A DYNAMIC SYSCTL
Dynamic nodes are created using one of the
.Fn SYSCTL_ADD_BOOL ,
.Fn SYSCTL_ADD_COUNTER_U64 ,
.Fn SYSCTL_ADD_COUNTER_U64_ARRAY ,
.Fn SYSCTL_ADD_INT ,
.Fn SYSCTL_ADD_LONG ,
.Fn SYSCTL_ADD_NODE ,
.Fn SYSCTL_ADD_NODE_WITH_LABEL ,
.Fn SYSCTL_ADD_OPAQUE ,
.Fn SYSCTL_ADD_PROC ,
.Fn SYSCTL_ADD_QUAD ,
.Fn SYSCTL_ADD_ROOT_NODE ,
.Fn SYSCTL_ADD_S8 ,
.Fn SYSCTL_ADD_S16 ,
.Fn SYSCTL_ADD_S32 ,
.Fn SYSCTL_ADD_S64 ,
.Fn SYSCTL_ADD_SBINTIME_MSEC ,
.Fn SYSCTL_ADD_SBINTIME_USEC ,
.Fn SYSCTL_ADD_STRING ,
.Fn SYSCTL_ADD_CONST_STRING ,
.Fn SYSCTL_ADD_STRUCT ,
.Fn SYSCTL_ADD_TIMEVAL_SEC ,
.Fn SYSCTL_ADD_U8 ,
.Fn SYSCTL_ADD_U16 ,
.Fn SYSCTL_ADD_U32 ,
.Fn SYSCTL_ADD_U64 ,
.Fn SYSCTL_ADD_UAUTO ,
.Fn SYSCTL_ADD_UINT ,
.Fn SYSCTL_ADD_ULONG ,
.Fn SYSCTL_ADD_UQUAD ,
.Fn SYSCTL_ADD_UMA_CUR
or
.Fn SYSCTL_ADD_UMA_MAX
functions.
See
.Xr sysctl_remove_oid 9
or
.Xr sysctl_ctx_free 9
for more information on how to destroy a dynamically created OID.
.Sh CONTROL FLAGS
For most of the above functions and macros, declaring a type as part
of the access flags is not necessary \[em] however, when declaring a
sysctl implemented by a function, including a type in the access mask
is required:
.Bl -tag -width ".Dv CTLTYPE_NOFETCH"
.It Dv CTLTYPE_NODE
This is a node intended to be a parent for other nodes.
.It Dv CTLTYPE_INT
This is a signed integer.
.It Dv CTLTYPE_STRING
This is a nul-terminated string stored in a character array.
.It Dv CTLTYPE_S8
This is an 8-bit signed integer.
.It Dv CTLTYPE_S16
This is a 16-bit signed integer.
.It Dv CTLTYPE_S32
This is a 32-bit signed integer.
.It Dv CTLTYPE_S64
This is a 64-bit signed integer.
.It Dv CTLTYPE_OPAQUE
This is an opaque data structure.
.It Dv CTLTYPE_STRUCT
Alias for
.Dv CTLTYPE_OPAQUE .
.It Dv CTLTYPE_U8
This is an 8-bit unsigned integer.
.It Dv CTLTYPE_U16
This is a 16-bit unsigned integer.
.It Dv CTLTYPE_U32
This is a 32-bit unsigned integer.
.It Dv CTLTYPE_U64
This is a 64-bit unsigned integer.
.It Dv CTLTYPE_UINT
This is an unsigned integer.
.It Dv CTLTYPE_LONG
This is a signed long.
.It Dv CTLTYPE_ULONG
This is an unsigned long.
.El
.Pp
All sysctl types except for new node declarations require one of the following
flags to be set indicating the read and write disposition of the sysctl:
.Bl -tag -width ".Dv CTLFLAG_ANYBODY"
.It Dv CTLFLAG_RD
This is a read-only sysctl.
.It Dv CTLFLAG_RDTUN
This is a read-only sysctl and tunable which is tried fetched once
from the system environment early during module load or system boot.
.It Dv CTLFLAG_WR
This is a writable sysctl.
.It Dv CTLFLAG_RW
This sysctl is readable and writable.
.It Dv CTLFLAG_RWTUN
This is a readable and writeable sysctl and tunable which is tried
fetched once from the system environment early during module load or
system boot.
.It Dv CTLFLAG_NOFETCH
In case the node is marked as a tunable using the CTLFLAG_[XX]TUN,
this flag will prevent fetching the initial value from the system
environment.
Typically this flag should only be used for very early
low level system setup code, and not by common drivers and modules.
.It Dv CTLFLAG_MPSAFE
This
.Xr sysctl 9
handler is MP safe.
Do not grab Giant around calls to this handler.
This should only be used for
.Fn SYSCTL_PROC
entries.
.El
.Pp
Additionally, any of the following optional flags may also be specified:
.Bl -tag -width ".Dv CTLFLAG_ANYBODY"
.It Dv CTLFLAG_ANYBODY
Any user or process can write to this sysctl.
.It Dv CTLFLAG_CAPRD
A process in capability mode can read from this sysctl.
.It Dv CTLFLAG_CAPWR
A process in capability mode can write to this sysctl.
.It Dv CTLFLAG_SECURE
This sysctl can be written to only if the effective securelevel of the
process is \[<=] 0.
.It Dv CTLFLAG_PRISON
This sysctl can be written to by processes in
.Xr jail 2 .
.It Dv CTLFLAG_SKIP
When iterating the sysctl name space, do not list this sysctl.
.It Dv CTLFLAG_TUN
Advisory flag that a system tunable also exists for this variable.
The initial sysctl value is tried fetched once from the system
environment early during module load or system boot.
.It Dv CTLFLAG_DYN
Dynamically created OIDs automatically get this flag set.
.It Dv CTLFLAG_VNET
OID references a VIMAGE-enabled variable.
.El
.Sh EXAMPLES
Sample use of
.Fn SYSCTL_DECL
to declare the
.Va security
sysctl tree for use by new nodes:
.Bd -literal -offset indent
SYSCTL_DECL(_security);
.Ed
.Pp
Examples of integer, opaque, string, and procedure sysctls follow:
.Bd -literal -offset indent
/*
 * Example of a constant integer value.  Notice that the control
 * flags are CTLFLAG_RD, the variable pointer is SYSCTL_NULL_INT_PTR,
 * and the value is declared.
 */
SYSCTL_INT(_debug_sizeof, OID_AUTO, bio, CTLFLAG_RD, SYSCTL_NULL_INT_PTR,
    sizeof(struct bio), "sizeof(struct bio)");

/*
 * Example of a variable integer value.  Notice that the control
 * flags are CTLFLAG_RW, the variable pointer is set, and the
 * value is 0.
 */
static int	doingcache = 1;		/* 1 => enable the cache */
SYSCTL_INT(_debug, OID_AUTO, vfscache, CTLFLAG_RW, &doingcache, 0,
    "Enable name cache");

/*
 * Example of a variable string value.  Notice that the control
 * flags are CTLFLAG_RW, that the variable pointer and string
 * size are set.  Unlike newer sysctls, this older sysctl uses a
 * static oid number.
 */
char kernelname[MAXPATHLEN] = "/kernel";	/* XXX bloat */
SYSCTL_STRING(_kern, KERN_BOOTFILE, bootfile, CTLFLAG_RW,
    kernelname, sizeof(kernelname), "Name of kernel file booted");

/*
 * Example of an opaque data type exported by sysctl.  Notice that
 * the variable pointer and size are provided, as well as a format
 * string for sysctl(8).
 */
static l_fp pps_freq;	/* scaled frequency offset (ns/s) */
SYSCTL_OPAQUE(_kern_ntp_pll, OID_AUTO, pps_freq, CTLFLAG_RD,
    &pps_freq, sizeof(pps_freq), "I", "");

/*
 * Example of a procedure based sysctl exporting string
 * information.  Notice that the data type is declared, the NULL
 * variable pointer and 0 size, the function pointer, and the
 * format string for sysctl(8).
 */
SYSCTL_PROC(_kern_timecounter, OID_AUTO, hardware, CTLTYPE_STRING |
    CTLFLAG_RW, NULL, 0, sysctl_kern_timecounter_hardware, "A",
    "");
.Ed
.Pp
The following is an example of
how to create a new top-level category
and how to hook up another subtree to an existing static node.
This example does not use contexts,
which results in tedious management of all intermediate oids,
as they need to be freed later on:
.Bd -literal -offset indent
#include <sys/sysctl.h>
 ...
/*
 * Need to preserve pointers to newly created subtrees,
 * to be able to free them later:
 */
static struct sysctl_oid *root1;
static struct sysctl_oid *root2;
static struct sysctl_oid *oidp;
static int a_int;
static char *string = "dynamic sysctl";
 ...

root1 = SYSCTL_ADD_ROOT_NODE(NULL,
	OID_AUTO, "newtree", CTLFLAG_RW, 0, "new top level tree");
oidp = SYSCTL_ADD_INT(NULL, SYSCTL_CHILDREN(root1),
	OID_AUTO, "newint", CTLFLAG_RW, &a_int, 0, "new int leaf");
 ...
root2 = SYSCTL_ADD_NODE(NULL, SYSCTL_STATIC_CHILDREN(_debug),
	OID_AUTO, "newtree", CTLFLAG_RW, 0, "new tree under debug");
oidp = SYSCTL_ADD_STRING(NULL, SYSCTL_CHILDREN(root2),
	OID_AUTO, "newstring", CTLFLAG_RD, string, 0, "new string leaf");
.Ed
.Pp
This example creates the following subtrees:
.Bd -literal -offset indent
debug.newtree.newstring
newtree.newint
.Ed
.Pp
.Em "Care should be taken to free all OIDs once they are no longer needed!"
.Sh SYSCTL NAMING
When adding, modifying, or removing sysctl names, it is important to be
aware that these interfaces may be used by users, libraries, applications,
or documentation (such as published books), and are implicitly published application interfaces.
As with other application interfaces, caution must be taken not to break
existing applications, and to think about future use of new name spaces so as
to avoid the need to rename or remove interfaces that might be depended on in
the future.
.Pp
The semantics chosen for a new sysctl should be as clear as possible,
and the name of the sysctl must closely reflect its semantics.
Therefore the sysctl name deserves a fair amount of consideration.
It should be short but yet representative of the sysctl meaning.
If the name consists of several words, they should be separated by
underscore characters, as in
.Va compute_summary_at_mount .
Underscore characters may be omitted only if the name consists of not more
than two words, each being not longer than four characters, as in
.Va bootfile .
.Pp
For boolean sysctls, negative logic should be totally avoided.
That is, do not use names like
.Va no_foobar
or
.Va foobar_disable .
They are confusing and lead to configuration errors.
Use positive logic instead:
.Va foobar ,
.Va foobar_enable .
.Pp
A temporary sysctl node OID that should not be relied upon must be designated
as such by a leading underscore character in its name.
For example:
.Va _dirty_hack .
.Sh SEE ALSO
.Xr sysctl 3 ,
.Xr sysctl 8 ,
.Xr device_get_sysctl 9 ,
.Xr sysctl_add_oid 9 ,
.Xr sysctl_ctx_free 9 ,
.Xr sysctl_ctx_init 9 ,
.Xr sysctl_remove_oid 9
.Sh HISTORY
The
.Xr sysctl 8
utility first appeared in
.Bx 4.4 .
.Nm SYSCTL_ADD_CONST_STRING
first appeared in
.Fx 12.1 .
.Sh AUTHORS
.An -nosplit
The
.Nm sysctl
implementation originally found in
.Bx
has been extensively rewritten by
.An Poul-Henning Kamp
in order to add support for name lookups, name space iteration, and dynamic
addition of MIB nodes.
.Pp
This man page was written by
.An Robert N. M. Watson .
.Sh SECURITY CONSIDERATIONS
When creating new sysctls, careful attention should be paid to the security
implications of the monitoring or management interface being created.
Most sysctls present in the kernel are read-only or writable only by the
superuser.
Sysctls exporting extensive information on system data structures and
operation, especially those implemented using procedures, will wish to
implement access control to limit the undesired exposure of information about
other processes, network connections, etc.
.Pp
The following top level sysctl name spaces are commonly used:
.Bl -tag -width ".Va regression"
.It Va compat
Compatibility layer information.
.It Va debug
Debugging information.
Various name spaces exist under
.Va debug .
.It Va hw
Hardware and device driver information.
.It Va kern
Kernel behavior tuning; generally deprecated in favor of more specific
name spaces.
.It Va machdep
Machine-dependent configuration parameters.
.It Va net
Network subsystem.
Various protocols have name spaces under
.Va net .
.It Va regression
Regression test configuration and information.
.It Va security
Security and security-policy configuration and information.
.It Va sysctl
Reserved name space for the implementation of sysctl.
.It Va user
Configuration settings relating to user application behavior.
Generally, configuring applications using kernel sysctls is discouraged.
.It Va vfs
Virtual file system configuration and information.
.It Va vm
Virtual memory subsystem configuration and information.
.El