SYSCTL(9)

📦
version 15.1 · 3 versions
, in freebsd 15.1, nixpkgs unstable

✨ SYSCTL(9) — Kernel Developer's Manual ✨

okay bestie, buckle up 💅 this is the manual page for the SYSCTL(9) family of functions — the whole "make sysctl MIB entries, dynamically or statically" saga. It's giving kernel-developer main character energy, and yes it's long, so we're doing this in parts 🫶

🌸 NAME 🌸

SYSCTL_DECL, SYSCTL_ADD_BOOL, SYSCTL_ADD_COUNTER_U64, SYSCTL_ADD_COUNTER_U64_ARRAY, SYSCTL_ADD_INT, SYSCTL_ADD_LONG, SYSCTL_ADD_NODE, SYSCTL_ADD_NODE_WITH_LABEL, SYSCTL_ADD_OPAQUE, SYSCTL_ADD_PROC, SYSCTL_ADD_QUAD, SYSCTL_ADD_ROOT_NODE, SYSCTL_ADD_S8, SYSCTL_ADD_S16, SYSCTL_ADD_S32, SYSCTL_ADD_S64, SYSCTL_ADD_SBINTIME_MSEC, SYSCTL_ADD_SBINTIME_USEC, SYSCTL_ADD_STRING, SYSCTL_ADD_CONST_STRING, SYSCTL_ADD_STRUCT, SYSCTL_ADD_TIMEVAL_SEC, SYSCTL_ADD_U8, SYSCTL_ADD_U16, SYSCTL_ADD_U32, SYSCTL_ADD_U64, SYSCTL_ADD_UAUTO, SYSCTL_ADD_UINT, SYSCTL_ADD_ULONG, SYSCTL_ADD_UMA_CUR, SYSCTL_ADD_UMA_MAX, SYSCTL_ADD_UQUAD, SYSCTL_CHILDREN, SYSCTL_STATIC_CHILDREN, SYSCTL_NODE_CHILDREN, SYSCTL_PARENT, SYSCTL_BOOL, SYSCTL_COUNTER_U64, SYSCTL_COUNTER_U64_ARRAY, SYSCTL_INT, SYSCTL_INT_WITH_LABEL, SYSCTL_LONG, sysctl_msec_to_ticks, SYSCTL_NODE, SYSCTL_NODE_WITH_LABEL, SYSCTL_OPAQUE, SYSCTL_PROC, SYSCTL_QUAD, SYSCTL_ROOT_NODE, SYSCTL_S8, SYSCTL_S16, SYSCTL_S32, SYSCTL_S64, SYSCTL_SBINTIME_MSEC, SYSCTL_SBINTIME_USEC, SYSCTL_STRING, SYSCTL_CONST_STRING, SYSCTL_STRUCT, SYSCTL_TIMEVAL_SEC, SYSCTL_U8, SYSCTL_U16, SYSCTL_U32, SYSCTL_U64, SYSCTL_UINT, SYSCTL_ULONG, SYSCTL_UMA_CUR, SYSCTL_UMA_MAX, SYSCTL_UQUAD

— and what do they all do? Dynamic and static sysctl MIB creation functions. that's the whole vibe. every single one of these icons is about spinning up sysctl entries, no cap 👑

🦋 SYNOPSIS 🦋

first, the include situation. you literally cannot do any of this without these two, so like, put them at the top:

#include <sys/param.h>
#include <sys/sysctl.h>

then the friend that lets you declare a thing before you define it:

SYSCTL_DECL(name);

okay and now the big block of dynamic creators — every one of these hands you back a struct sysctl_oid *, and they all start with the same ctx, parent, number, name, ctlflags little lineup (icons fr):

struct sysctl_oid *
SYSCTL_ADD_BOOL(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, bool *ptr, uint8_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_COUNTER_U64(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, counter_u64_t *ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_COUNTER_U64_ARRAY(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, counter_u64_t *ptr, intmax_t len, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_INT(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, int *ptr, int val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_LONG(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, long *ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_NODE(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, int (*handler)(SYSCTL_HANDLER_ARGS),
    const char *descr);

struct sysctl_oid *
SYSCTL_ADD_NODE_WITH_LABEL(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, int (*handler)(SYSCTL_HANDLER_ARGS),
    const char *descr, const char *label);

struct sysctl_oid *
SYSCTL_ADD_OPAQUE(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, void *ptr, intptr_t len, const char *format,
    const char *descr);

struct sysctl_oid *
SYSCTL_ADD_PROC(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, void *arg1, intptr_t arg2,
    int (*handler) (SYSCTL_HANDLER_ARGS), const char *format,
    const char *descr);

struct sysctl_oid *
SYSCTL_ADD_QUAD(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, int64_t *ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_ROOT_NODE(struct sysctl_ctx_list *ctx, int number,
    const char *name, int ctlflags, int (*handler)(SYSCTL_HANDLER_ARGS),
    const char *descr);

ok and here come the signed girls — S8 through S64, all cute and fixed-width ✨

struct sysctl_oid *
SYSCTL_ADD_S8(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, int8_t *ptr, int8_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_S16(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, int16_t *ptr, int16_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_S32(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, int32_t *ptr, int32_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_S64(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, int64_t *ptr, int64_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_SBINTIME_MSEC(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, sbintime_t *ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_SBINTIME_USEC(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, sbintime_t *ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_STRING(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, char *ptr, intptr_t len, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_CONST_STRING(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, const char *ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_STRUCT(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, void *ptr, struct_type, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_TIMEVAL_SEC(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, struct timeval *ptr, const char *descr);

and the unsigned besties, because balance 💖

struct sysctl_oid *
SYSCTL_ADD_U8(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, uint8_t *ptr, uint8_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_U16(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, uint16_t *ptr, uint16_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_U32(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, uint32_t *ptr, uint32_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_U64(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, uint64_t *ptr, uint64_t val, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_UINT(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, unsigned int *ptr, unsigned int val,
    const char *descr);

struct sysctl_oid *
SYSCTL_ADD_ULONG(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, unsigned long *ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_UQUAD(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, uint64_t *ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_UMA_CUR(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, uma_zone_t ptr, const char *descr);

struct sysctl_oid *
SYSCTL_ADD_UMA_MAX(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, uma_zone_t ptr, const char *descr);
const char *descr
🫶 real talk: yes, that stray const char *descr line really is chilling there on its own in the original. we don't question it, we just preserve it. it's iconic.

and the auto-magic one that figures out the width for you:

struct sysctl_oid *
SYSCTL_ADD_UAUTO(struct sysctl_ctx_list *ctx,
    struct sysctl_oid_list *parent, int number, const char *name,
    int ctlflags, void *ptr, const char *descr);

now the navigators 🔍 — these are how you get around the tree:

struct sysctl_oid_list *
SYSCTL_CHILDREN(struct sysctl_oid *oidp);

struct sysctl_oid_list *
SYSCTL_STATIC_CHILDREN(struct sysctl_oid_list OID_NAME);

struct sysctl_oid_list *
SYSCTL_NODE_CHILDREN(parent, name);

struct sysctl_oid *
SYSCTL_PARENT(struct sysctl_oid *oid);

then the static macro girlies — same energy, but compile-time and on the stack of your source file:

SYSCTL_BOOL(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_COUNTER_U64(parent, number, name, ctlflags, ptr, descr);

SYSCTL_COUNTER_U64_ARRAY(parent, number, name, ctlflags, ptr, len,
    descr);

SYSCTL_INT(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_INT_WITH_LABEL(parent, number, name, ctlflags, ptr, val, descr,
    label);

SYSCTL_LONG(parent, number, name, ctlflags, ptr, val, descr);

int
sysctl_msec_to_ticks(SYSCTL_HANDLER_ARGS);

SYSCTL_NODE(parent, number, name, ctlflags, handler, descr);

SYSCTL_NODE_WITH_LABEL(parent, number, name, ctlflags, handler, descr,
    label);

SYSCTL_OPAQUE(parent, number, name, ctlflags, ptr, len, format, descr);

SYSCTL_PROC(parent, number, name, ctlflags, arg1, arg2, handler, format,
    descr);

SYSCTL_QUAD(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_ROOT_NODE(number, name, ctlflags, handler, descr);

SYSCTL_S8(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_S16(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_S32(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_S64(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_SBINTIME_MSEC(parent, number, name, ctlflags, ptr, descr);

SYSCTL_SBINTIME_USEC(parent, number, name, ctlflags, ptr, descr);

SYSCTL_STRING(parent, number, name, ctlflags, arg, len, descr);

SYSCTL_CONST_STRING(parent, number, name, ctlflags, arg, descr);

SYSCTL_STRUCT(parent, number, name, ctlflags, ptr, struct_type, descr);

SYSCTL_TIMEVAL_SEC(parent, number, name, ctlflags, ptr, descr);

SYSCTL_U8(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_U16(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_U32(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_U64(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_UINT(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_ULONG(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_UQUAD(parent, number, name, ctlflags, ptr, val, descr);

SYSCTL_UMA_MAX(parent, number, name, ctlflags, ptr, descr);

SYSCTL_UMA_CUR(parent, number, name, ctlflags, ptr, descr);

like — that's the whole list, every single one, no skips 💀 and yes there are a LOT, but that's the assignment.

💅 DESCRIPTION 💅

so here's the tea 🍓: the SYSCTL kernel interface lets you create sysctl(8) MIB entries — either dynamically or statically. choose your fighter.

the genuinely slay part? all static sysctls are automatically destroyed when the module which they are part of is unloaded ✨ like it literally cleans up after itself, we love a self-sufficient icon.

and most top level categories are created statically, which means they're available to all kernel code and its modules — everybody gets access, it's giving sharing is caring 🫶

okay that's the setup bestie, we've got the whole cast of characters and the synopsis locked in — part 2 is where it gets really good 🌈

💬 DESCRIPTION OF ARGUMENTS 💬

Okay bestie, this is the cast of characters — every argument gets its moment, and yes, you need to know all of them. 💅

  • ctx — Pointer to sysctl context or NULL, if no context. See 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.
  • parent — A pointer to a struct sysctl_oid_list, which is the head of the parent's list of children. This pointer is retrieved using the SYSCTL_STATIC_CHILDREN() macro for static sysctls and the SYSCTL_CHILDREN() macro for dynamic sysctls. The SYSCTL_PARENT() macro can be used to get the parent of an OID. The macro returns NULL if there is no parent.
  • number — The OID number that will be assigned to this OID. In almost all cases this should be set to OID_AUTO, which will result in the assignment of the next available OID number.
  • name — The name of the OID. The newly created OID will contain a copy of the name.
  • ctlflags — A bit mask of sysctl control flags. See the section below describing all the control flags.
  • arg1 — First callback argument for procedure sysctls.
  • arg2 — Second callback argument for procedure sysctls.
  • len — The length of the data pointed to by the ptr argument. For string type OIDs a length of zero means that 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.
  • 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 val argument.
  • val — If the ptr argument is SYSCTL_NULL_XXX_PTR, gives the constant value returned by this OID. Else this argument is not used.
  • struct_type — Name of structure type.
  • 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 SYSCTL_PROC() macro or the SYSCTL_ADD_PROC() function.
  • 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 sysctl(8) to apply proper data formatting for display purposes.

Current formats:

  • N — node
  • A — char *
  • C — int8_t
  • CU — uint8_t
  • I — int
  • IK[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
  • IU — unsigned int
  • L — long
  • LU — unsigned long
  • Q — quad_t
  • QU — u_quad_t
  • S — int16_t
  • SU — uint16_t
  • S,TYPE — struct TYPE structures
  • descr — A pointer to a textual description of the OID.
  • 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.

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. Periodt — that's not a vibe suggestion, it's how the label thing actually works.

🔢 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, SYSCTL_INT() reports the raw value of an associated variable of type int. However, nodes may also export a value that is a translation of an internal representation.

The sysctl_msec_to_ticks() handler can be used with SYSCTL_PROC() or SYSCTL_ADD_PROC() to export a millisecond time interval. When using this handler, the arg2 parameter points to an in-kernel variable of type int which stores a tick count suitable for use with functions like tsleep(9). The sysctl_msec_to_ticks() function converts this value to milliseconds when reporting the node's value. Similarly, sysctl_msec_to_ticks() accepts new values in milliseconds and stores an equivalent value in ticks to *arg2. Note that new code should use kernel variables of type sbintime_t instead of tick counts.

The SYSCTL_ADD_SBINTIME_MSEC() and SYSCTL_ADD_SBINTIME_USEC() functions and SYSCTL_SBINTIME_MSEC() and SYSCTL_SBINTIME_USEC() macros all create nodes which export an in-kernel variable of type 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).

The SYSCTL_ADD_TIMEVAL_SEC() function and SYSCTL_TIMEVAL_SEC() macro create nodes which export an in-kernel variable of type 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 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.

🌱 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 SYSCTL_ROOT_NODE() macro or the SYSCTL_ADD_ROOT_NODE() function needs to be used. By default all static sysctl node OIDs are global and need a SYSCTL_DECL() statement prior to their SYSCTL_NODE() definition statement, typically in a so-called header file.

🧵 CREATING SYSCTL STRINGS 🧵

Zero terminated character strings sysctls are created either using the SYSCTL_STRING() macro or the SYSCTL_ADD_STRING() function. If the len argument is zero, the string length is computed at every access to the OID using strlen(3). Use the SYSCTL_CONST_STRING() macro or the SYSCTL_ADD_CONST_STRING() function to add a sysctl for a constant string.

🫥 CREATING OPAQUE SYSCTLS 🫥

The SYSCTL_OPAQUE() or SYSCTL_STRUCT() macros or the SYSCTL_ADD_OPAQUE() or SYSCTL_ADD_STRUCT() functions create an OID that handle any chunk of data of the size specified by the len argument and data pointed to by the ptr argument. When using the structure version the type is encoded as part of the created sysctl.

🛠️ CREATING CUSTOM SYSCTLS 🛠️

The SYSCTL_PROC() macro and the SYSCTL_ADD_PROC() function create OIDs with the specified 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.

🗿 CREATING A STATIC SYSCTL 🗿

Static sysctls are declared using one of these macros, no cap:

  • SYSCTL_BOOL()
  • SYSCTL_COUNTER_U64()
  • SYSCTL_COUNTER_U64_ARRAY()
  • SYSCTL_INT()
  • SYSCTL_INT_WITH_LABEL()
  • SYSCTL_LONG()
  • SYSCTL_NODE()
  • SYSCTL_NODE_WITH_LABEL()
  • SYSCTL_OPAQUE()
  • SYSCTL_PROC()
  • SYSCTL_QUAD()
  • SYSCTL_ROOT_NODE()
  • SYSCTL_S8()
  • SYSCTL_S16()
  • SYSCTL_S32()
  • SYSCTL_S64()
  • SYSCTL_SBINTIME_MSEC()
  • SYSCTL_SBINTIME_USEC()
  • SYSCTL_STRING()
  • SYSCTL_CONST_STRING()
  • SYSCTL_STRUCT()
  • SYSCTL_TIMEVAL_SEC()
  • SYSCTL_U8()
  • SYSCTL_U16()
  • SYSCTL_U32()
  • SYSCTL_U64()
  • SYSCTL_UINT()
  • SYSCTL_ULONG()
  • SYSCTL_UQUAD()
  • SYSCTL_UMA_CUR()
  • SYSCTL_UMA_MAX()

🌀 CREATING A DYNAMIC SYSCTL 🌀

Dynamic nodes are created using one of these functions, and yes, the list is long — it's giving "reference manual main character energy":

  • SYSCTL_ADD_BOOL()
  • SYSCTL_ADD_COUNTER_U64()
  • SYSCTL_ADD_COUNTER_U64_ARRAY()
  • SYSCTL_ADD_INT()
  • SYSCTL_ADD_LONG()
  • SYSCTL_ADD_NODE()
  • SYSCTL_ADD_NODE_WITH_LABEL()
  • SYSCTL_ADD_OPAQUE()
  • SYSCTL_ADD_PROC()
  • SYSCTL_ADD_QUAD()
  • SYSCTL_ADD_ROOT_NODE()
  • SYSCTL_ADD_S8()
  • SYSCTL_ADD_S16()
  • SYSCTL_ADD_S32()
  • SYSCTL_ADD_S64()
  • SYSCTL_ADD_SBINTIME_MSEC()
  • SYSCTL_ADD_SBINTIME_USEC()
  • SYSCTL_ADD_STRING()
  • SYSCTL_ADD_CONST_STRING()
  • SYSCTL_ADD_STRUCT()
  • SYSCTL_ADD_TIMEVAL_SEC()
  • SYSCTL_ADD_U8()
  • SYSCTL_ADD_U16()
  • SYSCTL_ADD_U32()
  • SYSCTL_ADD_U64()
  • SYSCTL_ADD_UAUTO()
  • SYSCTL_ADD_UINT()
  • SYSCTL_ADD_ULONG()
  • SYSCTL_ADD_UQUAD()
  • SYSCTL_ADD_UMA_CUR()
  • SYSCTL_ADD_UMA_MAX()

See sysctl_remove_oid(9) or sysctl_ctx_free(9) for more information on how to destroy a dynamically created OID. Cleanup era, bestie — don't leave orphaned dynamic OIDs just vibing in the kernel.

✨ CONTROL FLAGS ✨

okay so for most of the functions and macros up above, you don't have to declare a type as part of the access flags — but bestie, the second you're declaring a sysctl that's implemented by a function, including a type in the access mask is literally required:

  • CTLTYPE_NODE — this one's a node that's meant to be a parent for other nodes 💅
  • CTLTYPE_INT — a signed integer, classic
  • CTLTYPE_STRING — a nul-terminated string stored in a character array
  • CTLTYPE_S8 — an 8-bit signed integer
  • CTLTYPE_S16 — a 16-bit signed integer
  • CTLTYPE_S32 — a 32-bit signed integer
  • CTLTYPE_S64 — a 64-bit signed integer
  • CTLTYPE_OPAQUE — an opaque data structure 👀
  • CTLTYPE_STRUCT — alias for CTLTYPE_OPAQUE
  • CTLTYPE_U8 — an 8-bit unsigned integer
  • CTLTYPE_U16 — a 16-bit unsigned integer
  • CTLTYPE_U32 — a 32-bit unsigned integer
  • CTLTYPE_U64 — a 64-bit unsigned integer
  • CTLTYPE_UINT — an unsigned integer
  • CTLTYPE_LONG — a signed long
  • CTLTYPE_ULONG — an unsigned long

so like, every sysctl type except new node declarations needs one of these flags set, and it tells you the read/write disposition of the sysctl. it's giving "pick your fighter":

  • CTLFLAG_RD — this is a read-only sysctl 🔒
  • CTLFLAG_RDTUN — this is a read-only sysctl and a tunable, which gets fetched once from the system environment early during module load or system boot
  • CTLFLAG_WR — this is a writable sysctl
  • CTLFLAG_RW — this sysctl is readable and writable, we love range 💖
  • CTLFLAG_RWTUN — this is a readable and writeable sysctl and a tunable, which is fetched once from the system environment early during module load or system boot
  • CTLFLAG_NOFETCH — if the node is marked as a tunable using CTLFLAG_[XX]TUN, this flag prevents fetching the initial value from the system environment. ngl this flag should typically only be used for very early low level system setup code, and not by common drivers and modules 🙈
  • CTLFLAG_MPSAFE — this sysctl(9) handler is MP safe, so do not grab Giant around calls to this handler. this should only be used for SYSCTL_PROC() entries 💅

and on top of that, any of these optional flags can also be specified:

  • CTLFLAG_ANYBODY — any user or process can write to this sysctl
  • CTLFLAG_CAPRD — a process in capability mode can read from this sysctl
  • CTLFLAG_CAPWR — a process in capability mode can write to this sysctl
  • CTLFLAG_SECURE — this sysctl can be written to only if the effective securelevel of the process is ≤ 0
  • CTLFLAG_PRISON — this sysctl can be written to by processes in jail(2)
  • CTLFLAG_SKIP — when iterating the sysctl name space, do not list this sysctl (sneaky, hidden, mysterious ✨)
  • CTLFLAG_TUN — advisory flag that a system tunable also exists for this variable. the initial sysctl value is fetched once from the system environment early during module load or system boot
  • CTLFLAG_DYN — dynamically created OIDs automatically get this flag set
  • CTLFLAG_VNET — OID references a VIMAGE-enabled variable

🌈 EXAMPLES 🌈

here's a lil sample of SYSCTL_DECL() being used to declare the security sysctl tree, so new nodes can use it:

SYSCTL_DECL(_security);

and now the moment you've been waiting for — examples of integer, opaque, string, and procedure sysctls, in all their glory:

/*
 * 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(_kern, OID_AUTO, hz_max, CTLFLAG_RD, SYSCTL_NULL_INT_PTR, HZ_MAXIMUM,
    "Maximum hz value supported");

/*
 * 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",
    "");

and here's how you'd create a whole new top-level category, plus how to hook another subtree up to an existing static node. heads up bestie — this example does not use contexts, which means you get the tedious job of managing all the intermediate oids yourself, because they need to be freed later on 😭

#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");

this example creates these subtrees right here:

debug.newtree.newstring
newtree.newint
⚠️ take care to free all OIDs once they are no longer needed! like, actually. please. don't just leave them lying around 🫶

🎀 SYSCTL NAMING 🎀

when you're adding, modifying, or removing sysctl names, you gotta be aware that these interfaces might be used by users, libraries, applications, or documentation (like, published books, bestie), which means they're implicitly published application interfaces. so just like with any other application interface, you gotta be careful not to break existing applications, and you gotta think about future use of new name spaces so you don't end up needing to rename or remove interfaces that someone might depend on later. no cap 🔥

the semantics you pick for a new sysctl should be as clear as possible, and the name of the sysctl must closely reflect its semantics. so yeah, the name deserves a fair amount of consideration. it should be short but still representative of what the sysctl means. if the name is several words, separate them with underscore characters, like in 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, like in bootfile.

for boolean sysctls, negative logic should be totally avoided. like, do not use names like no_foobar or foobar_disable — they're confusing and they lead to configuration errors 😭 use positive logic instead: foobar, foobar_enable. periodt 💅

and if you've got a temporary sysctl node OID that should not be relied upon, you gotta designate it as such with a leading underscore character in its name. for example: _dirty_hack 💀

🦋 SEE ALSO 🦋

sysctl(3), sysctl(8), device_get_sysctl(9), sysctl_add_oid(9), sysctl_ctx_free(9), sysctl_ctx_init(9), sysctl_remove_oid(9)

⭐ HISTORY ⭐

the sysctl(8) utility first appeared in 4.4BSD. and SYSCTL_ADD_CONST_STRING first appeared in FreeBSD 12.1. iconic behavior honestly 👑

👑 AUTHORS 👑

the sysctl implementation originally found in BSD has been extensively rewritten by Poul-Henning Kamp in order to add support for name lookups, name space iteration, and dynamic addition of MIB nodes. absolute legend 🫶

this man page was written by Robert N. M. Watson.

🔐 SECURITY CONSIDERATIONS 🔐

when you're creating new sysctls, pay careful attention to the security implications of the monitoring or management interface you're creating. most sysctls in the kernel are read-only, or writable only by the superuser. sysctls that export 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 info about other processes, network connections, etc. this is not the vibe to be sloppy about 🙈

these top level sysctl name spaces are the ones commonly used:

  • compat — compatibility layer information
  • debug — debugging information. various name spaces exist under debug
  • hw — hardware and device driver information
  • kern — kernel behavior tuning; generally deprecated in favor of more specific name spaces
  • machdep — machine-dependent configuration parameters
  • net — network subsystem. various protocols have name spaces under net
  • regression — regression test configuration and information
  • security — security and security-policy configuration and information
  • sysctl — reserved name space for the implementation of sysctl
  • user — configuration settings relating to user application behavior. generally, configuring applications using kernel sysctls is discouraged
  • vfs — virtual file system configuration and information
  • vm — virtual memory subsystem configuration and information

and that's the whole page bestie!! 🎉 you made it to the end, so here's your lil sign-off: this one's from the Debian source, dated September 28, 2025 — go forth and sysctl responsibly 💖✨🍓

📦 where this page comes from