sysctl(2)

💖 sysctl(2) — the "read/write system parameters" girlie 💖

okay bestie, pull up a chair, because this is the sysctl system call page from section 2 (the System Calls Manual, yes we're in the syscall girlies' territory ✨). I need you to know upfront: this one is a history lesson, not a "go use this right now" situation. She's retired. She's an icon. She no longer exists on current kernels. Let's get into it 💅

---

🌸 NAME 🌸

sysctl — read/write system parameters

---

✨ SYNOPSIS ✨

#include <unistd.h>
#include <linux/sysctl.h>

[[deprecated]] int _sysctl(struct __sysctl_args *args);

not the [[deprecated]] tag already showing up in the synopsis... she is literally telling on herself 😭

---

🦋 DESCRIPTION 🦋

This system call no longer exists on current kernels! See NOTES. (yes, in the original it's an exclamation point, they were stressed.)

The _sysctl() call reads and/or writes kernel parameters. Like, for example, the hostname, or the maximum number of open files. Classic system-parameter vibes.

The argument has the form

struct __sysctl_args {
    int    *name;    /* integer vector describing variable */
    int     nlen;    /* number of elements of this vector */
    void   *oldval;  /* 0 or address where to store old value */
    size_t *oldlenp; /* available room for old value,
                        overwritten by actual size of old value */
    void   *newval;  /* 0 or address of new value */
    size_t  newlen;  /* size of new value */
};

So the whole thing is: this call does a search in a tree structure, possibly resembling a directory tree under /proc/sys, and if the requested item is found, it calls some appropriate routine to read or modify the value. cute. tree-shaped. very demure.

---

🫶 RETURN VALUE 🫶

Upon successful completion, _sysctl() returns 0. Otherwise, a value of -1 is returned and errno is set to indicate the error. standard behavior, no notes ✋

---

👑 ERRORS 👑

• EACCES
• EPERM — no search permission for one of the encountered "directories", or no read permission where oldval was nonzero, or no write permission where newval was nonzero. (the quotes around "directories" are the original's, and honestly they ate 💀)
• EFAULT — the invocation asked for the previous value by setting oldval non-NULL, but allowed zero room in oldlenp. rude of you, honestly.
• ENOTDIR — name was not found.

---

🌈 STANDARDS 🌈

Linux. that's it. that's the whole section. she's Linux-only and she's not apologizing 🫡

---

💜 HISTORY 💜

• 4.4BSD, Linux 1.3.57. Removed in Linux 5.5, glibc 2.32.
• Only Linux has the /proc/sys mirror, and the object naming schemes differ between Linux and 4.4BSD, but the declaration of the sysctl() function is the same in both. so like... cousins. distant cousins. same signature, different lore.

---

🎀 NOTES 🎀

Use of this system call was long discouraged: since Linux 2.6.24, uses of this system call result in warnings in the kernel log, and in Linux 5.5, the system call was finally removed. Use the /proc/sys interface instead. that's the glow-up, babes — /proc/sys is the moment.

Note that on older kernels where this system call still exists, it is available only if the kernel was configured with the CONFIG_SYSCTL_SYSCALL option. Furthermore, glibc does not provide a wrapper for this system call, necessitating the use of syscall(2). so you had to rawdog it through syscall(2). no wrapper, no comfort, no hand-holding 😭

---

🙈 BUGS 🙈

• The object names vary between kernel versions, making this system call worthless for applications. worthless, the manpage said it, not me 💀
• Not all available objects are properly documented.
• It is not yet possible to change operating system by writing to /proc/sys/kernel/ostype. bestie you cannot just declare a new OS. unfortunately.

---

🍓 EXAMPLES 🍓

here's the whole demo, verbatim, no edits, we respect the code:

#define _GNU_SOURCE
#include <stdcountof.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/syscall.h>
#include <unistd.h>

#include <linux/sysctl.h>

int _sysctl(struct __sysctl_args *args);

#define OSNAMESZ 100

int
main(void)
{
    int                   name[] = { CTL_KERN, KERN_OSTYPE };
    char                  osname[OSNAMESZ];
    size_t                osnamelth;
    struct __sysctl_args  args;

    memset(&args, 0, sizeof(args));
    args.name = name;
    args.nlen = countof(name);
    args.oldval = osname;
    args.oldlenp = &osnamelth;

    osnamelth = sizeof(osname);

    if (syscall(SYS__sysctl, &args) == -1) {
        perror("_sysctl");
        exit(EXIT_FAILURE);
    }
    printf("This machine is running %*s\n", (int) osnamelth, osname);
    exit(EXIT_SUCCESS);
}

she's asking the kernel what OS this is. which, given the BUGS section, is a whole journey. anyway, slay 🧚

---

⭐ SEE ALSO ⭐

• proc(5) — the one that actually matters now, fr fr

---

and that's the page, bestie. this was sysctl(2) from Linux man-pages 6.20, dated 2026-02-10. go forth, use /proc/sys, and let _sysctl() rest in peace 🕊️ it's giving retired legend energy and we love her for it 💖✨