xref: /freebsd/usr.bin/top/machine.c (revision 0de89efe5c443f213c7ea28773ef2dc6cf3af2ed)

/*
 * top - a top users display for Unix
 *
 * SYNOPSIS:  For FreeBSD-2.x system
 *
 * DESCRIPTION:
 * Originally written for BSD4.4 system by Christos Zoulas.
 * Ported to FreeBSD 2.x by Steven Wallace && Wolfram Schneider
 *
 * This is the machine-dependent module for FreeBSD 2.2
 * Works for:
 *	FreeBSD 2.2, and probably FreeBSD 2.1.x
 *
 * LIBS: -lkvm
 *
 * AUTHOR:  Christos Zoulas <christos@ee.cornell.edu>
 *          Steven Wallace  <swallace@freebsd.org>
 *          Wolfram Schneider <wosch@FreeBSD.org>
 *
 * $Id: machine.c,v 1.6 1997/08/27 03:48:25 peter Exp $
 */


#include <sys/types.h>
#include <sys/signal.h>
#include <sys/param.h>

#include "os.h"
#include <stdio.h>
#include <nlist.h>
#include <math.h>
#include <kvm.h>
#include <pwd.h>
#include <sys/errno.h>
#include <sys/sysctl.h>
#include <sys/dkstat.h>
#include <sys/file.h>
#include <sys/time.h>
#include <sys/proc.h>
#include <sys/user.h>
#include <sys/vmmeter.h>
#include <sys/resource.h>
#include <sys/rtprio.h>

/* Swap */
#include <stdlib.h>
#include <sys/rlist.h>
#include <sys/conf.h>

#include <osreldate.h> /* for changes in kernel structures */

#include "top.h"
#include "machine.h"

static int check_nlist __P((struct nlist *));
static int getkval __P((unsigned long, int *, int, char *));
extern char* printable __P((char *));
int swapmode __P((int *retavail, int *retfree));
static int smpmode;
static int namelength;
static int cmdlength;


/* get_process_info passes back a handle.  This is what it looks like: */

struct handle
{
    struct kinfo_proc **next_proc;	/* points to next valid proc pointer */
    int remaining;		/* number of pointers remaining */
};

/* declarations for load_avg */
#include "loadavg.h"

#define PP(pp, field) ((pp)->kp_proc . field)
#define EP(pp, field) ((pp)->kp_eproc . field)
#define VP(pp, field) ((pp)->kp_eproc.e_vm . field)

/* define what weighted cpu is.  */
#define weighted_cpu(pct, pp) (PP((pp), p_swtime) == 0 ? 0.0 : \
			 ((pct) / (1.0 - exp(PP((pp), p_swtime) * logcpu))))

/* what we consider to be process size: */
#define PROCSIZE(pp) (VP((pp), vm_tsize) + VP((pp), vm_dsize) + VP((pp), vm_ssize))

/* definitions for indices in the nlist array */


static struct nlist nlst[] = {
#define X_CCPU		0
    { "_ccpu" },		/* 0 */
#define X_CP_TIME	1
    { "_cp_time" },		/* 1 */
#define X_HZ		2
    { "_hz" },		        /* 2 */
#define X_STATHZ	3
    { "_stathz" },		/* 3 */
#define X_AVENRUN	4
    { "_averunnable" },		/* 4 */

/* Swap */
#define VM_SWAPLIST	5
	{ "_swaplist" },/* list of free swap areas */
#define VM_SWDEVT	6
	{ "_swdevt" },	/* list of swap devices and sizes */
#define VM_NSWAP	7
	{ "_nswap" },	/* size of largest swap device */
#define VM_NSWDEV	8
	{ "_nswdev" },	/* number of swap devices */
#define VM_DMMAX	9
	{ "_dmmax" },	/* maximum size of a swap block */
#define X_BUFSPACE	10
	{ "_bufspace" },	/* K in buffer cache */
#define X_CNT           11
    { "_cnt" },		        /* struct vmmeter cnt */

/* Last pid */
#define X_LASTPID	12
    { "_nextpid" },
    { 0 }
};

/*
 *  These definitions control the format of the per-process area
 */

static char smp_header[] =
  "  PID %-*.*s PRI NICE  SIZE    RES STATE C   TIME   WCPU    CPU COMMAND";

#define smp_Proc_format \
	"%5d %-*.*s %3d %3d%7s %6s %-6.6s%1x%7s %5.2f%% %5.2f%% %.*s"

static char up_header[] =
  "  PID %-*.*s PRI NICE  SIZE    RES STATE    TIME   WCPU    CPU COMMAND";

#define up_Proc_format \
	"%5d %-*.*s %3d %3d%7s %6s %-6.6s%.0d%7s %5.2f%% %5.2f%% %.*s"



/* process state names for the "STATE" column of the display */
/* the extra nulls in the string "run" are for adding a slash and
   the processor number when needed */

char *state_abbrev[] =
{
    "", "START", "RUN\0\0\0", "SLEEP", "STOP", "ZOMB",
};


static kvm_t *kd;

/* values that we stash away in _init and use in later routines */

static double logcpu;

/* these are retrieved from the kernel in _init */

static          long hz;
static load_avg  ccpu;

/* these are offsets obtained via nlist and used in the get_ functions */

static unsigned long cp_time_offset;
static unsigned long avenrun_offset;
static unsigned long lastpid_offset;
static long lastpid;
static unsigned long cnt_offset;
static unsigned long bufspace_offset;
static long cnt;

/* these are for calculating cpu state percentages */

static long cp_time[CPUSTATES];
static long cp_old[CPUSTATES];
static long cp_diff[CPUSTATES];

/* these are for detailing the process states */

int process_states[6];
char *procstatenames[] = {
    "", " starting, ", " running, ", " sleeping, ", " stopped, ",
    " zombie, ",
    NULL
};

/* these are for detailing the cpu states */

int cpu_states[CPUSTATES];
char *cpustatenames[] = {
    "user", "nice", "system", "interrupt", "idle", NULL
};

/* these are for detailing the memory statistics */

int memory_stats[7];
char *memorynames[] = {
    "K Active, ", "K Inact, ", "K Wired, ", "K Cache, ", "K Buf, ", "K Free",
    NULL
};

int swap_stats[7];
char *swapnames[] = {
/*   0           1            2           3            4       5 */
    "K Total, ", "K Used, ", "K Free, ", "% Inuse, ", "K In, ", "K Out",
    NULL
};


/* these are for keeping track of the proc array */

static int nproc;
static int onproc = -1;
static int pref_len;
static struct kinfo_proc *pbase;
static struct kinfo_proc **pref;

/* these are for getting the memory statistics */

static int pageshift;		/* log base 2 of the pagesize */

/* define pagetok in terms of pageshift */

#define pagetok(size) ((size) << pageshift)

/* useful externals */
long percentages();

int
machine_init(statics)

struct statics *statics;

{
    register int i = 0;
    register int pagesize;
    int modelen;
    struct passwd *pw;

    modelen = sizeof(smpmode);
    if ((sysctlbyname("machdep.smp_active", &smpmode, &modelen, NULL, 0) < 0 &&
         sysctlbyname("smp.smp_active", &smpmode, &modelen, NULL, 0) < 0) ||
	modelen != sizeof(smpmode))
	    smpmode = 0;

    while ((pw = getpwent()) != NULL) {
	if (strlen(pw->pw_name) > namelength)
	    namelength = strlen(pw->pw_name);
    }
    if (namelength < 8)
	namelength = 8;
    if (namelength > 16)
	namelength = 16;

    if ((kd = kvm_open(NULL, NULL, NULL, O_RDONLY, "kvm_open")) == NULL)
	return -1;


    /* get the list of symbols we want to access in the kernel */
    (void) kvm_nlist(kd, nlst);
    if (nlst[0].n_type == 0)
    {
	fprintf(stderr, "top: nlist failed\n");
	return(-1);
    }

    /* make sure they were all found */
    if (i > 0 && check_nlist(nlst) > 0)
    {
	return(-1);
    }

    /* get the symbol values out of kmem */
    (void) getkval(nlst[X_STATHZ].n_value, (int *)(&hz), sizeof(hz), "!");
    if (!hz) {
	(void) getkval(nlst[X_HZ].n_value, (int *)(&hz), sizeof(hz),
		       nlst[X_HZ].n_name);
    }

    (void) getkval(nlst[X_CCPU].n_value,   (int *)(&ccpu),	sizeof(ccpu),
	    nlst[X_CCPU].n_name);

    /* stash away certain offsets for later use */
    cp_time_offset = nlst[X_CP_TIME].n_value;
    avenrun_offset = nlst[X_AVENRUN].n_value;
    lastpid_offset =  nlst[X_LASTPID].n_value;
    cnt_offset = nlst[X_CNT].n_value;
    bufspace_offset = nlst[X_BUFSPACE].n_value;

    /* this is used in calculating WCPU -- calculate it ahead of time */
    logcpu = log(loaddouble(ccpu));

    pbase = NULL;
    pref = NULL;
    nproc = 0;
    onproc = -1;
    /* get the page size with "getpagesize" and calculate pageshift from it */
    pagesize = getpagesize();
    pageshift = 0;
    while (pagesize > 1)
    {
	pageshift++;
	pagesize >>= 1;
    }

    /* we only need the amount of log(2)1024 for our conversion */
    pageshift -= LOG1024;

    /* fill in the statics information */
    statics->procstate_names = procstatenames;
    statics->cpustate_names = cpustatenames;
    statics->memory_names = memorynames;
    statics->swap_names = swapnames;

    /* all done! */
    return(0);
}

char *format_header(uname_field)

register char *uname_field;

{
    register char *ptr;
    static char Header[128];

    snprintf(Header, sizeof(Header), smpmode ? smp_header : up_header,
	     namelength, namelength, uname_field);

    cmdlength = 80 - strlen(Header) + 6;

    return Header;
}

static int swappgsin = -1;
static int swappgsout = -1;
extern struct timeval timeout;

void
get_system_info(si)

struct system_info *si;

{
    long total;
    load_avg avenrun[3];

    /* get the cp_time array */
    (void) getkval(cp_time_offset, (int *)cp_time, sizeof(cp_time),
		   nlst[X_CP_TIME].n_name);
    (void) getkval(avenrun_offset, (int *)avenrun, sizeof(avenrun),
		   nlst[X_AVENRUN].n_name);

    (void) getkval(lastpid_offset, (int *)(&lastpid), sizeof(lastpid),
		   "!");

    /* convert load averages to doubles */
    {
	register int i;
	register double *infoloadp;
	load_avg *avenrunp;

#ifdef notyet
	struct loadavg sysload;
	int size;
	getkerninfo(KINFO_LOADAVG, &sysload, &size, 0);
#endif

	infoloadp = si->load_avg;
	avenrunp = avenrun;
	for (i = 0; i < 3; i++)
	{
#ifdef notyet
	    *infoloadp++ = ((double) sysload.ldavg[i]) / sysload.fscale;
#endif
	    *infoloadp++ = loaddouble(*avenrunp++);
	}
    }

    /* convert cp_time counts to percentages */
    total = percentages(CPUSTATES, cpu_states, cp_time, cp_old, cp_diff);

    /* sum memory & swap statistics */
    {
	struct vmmeter sum;
	static unsigned int swap_delay = 0;
	static int swapavail = 0;
	static int swapfree = 0;
	static int bufspace = 0;

        (void) getkval(cnt_offset, (int *)(&sum), sizeof(sum),
		   "_cnt");
        (void) getkval(bufspace_offset, (int *)(&bufspace), sizeof(bufspace),
		   "_bufspace");

	/* convert memory stats to Kbytes */
	memory_stats[0] = pagetok(sum.v_active_count);
	memory_stats[1] = pagetok(sum.v_inactive_count);
	memory_stats[2] = pagetok(sum.v_wire_count);
	memory_stats[3] = pagetok(sum.v_cache_count);
	memory_stats[4] = bufspace / 1024;
	memory_stats[5] = pagetok(sum.v_free_count);
	memory_stats[6] = -1;

	/* first interval */
        if (swappgsin < 0) {
	    swap_stats[4] = 0;
	    swap_stats[5] = 0;
	}

	/* compute differences between old and new swap statistic */
	else {
	    swap_stats[4] = pagetok(((sum.v_swappgsin - swappgsin)));
	    swap_stats[5] = pagetok(((sum.v_swappgsout - swappgsout)));
	}

        swappgsin = sum.v_swappgsin;
	swappgsout = sum.v_swappgsout;

	/* call CPU heavy swapmode() only for changes */
        if (swap_stats[4] > 0 || swap_stats[5] > 0 || swap_delay == 0) {
	    swap_stats[3] = swapmode(&swapavail, &swapfree);
	    swap_stats[0] = swapavail;
	    swap_stats[1] = swapavail - swapfree;
	    swap_stats[2] = swapfree;
	}
        swap_delay = 1;
	swap_stats[6] = -1;
    }

    /* set arrays and strings */
    si->cpustates = cpu_states;
    si->memory = memory_stats;
    si->swap = swap_stats;


    if(lastpid > 0) {
	si->last_pid = lastpid;
    } else {
	si->last_pid = -1;
    }
}

static struct handle handle;

caddr_t get_process_info(si, sel, compare)

struct system_info *si;
struct process_select *sel;
int (*compare)();

{
    register int i;
    register int total_procs;
    register int active_procs;
    register struct kinfo_proc **prefp;
    register struct kinfo_proc *pp;

    /* these are copied out of sel for speed */
    int show_idle;
    int show_system;
    int show_uid;
    int show_command;


    pbase = kvm_getprocs(kd, KERN_PROC_ALL, 0, &nproc);
    if (nproc > onproc)
	pref = (struct kinfo_proc **) realloc(pref, sizeof(struct kinfo_proc *)
		* (onproc = nproc));
    if (pref == NULL || pbase == NULL) {
	(void) fprintf(stderr, "top: Out of memory.\n");
	quit(23);
    }
    /* get a pointer to the states summary array */
    si->procstates = process_states;

    /* set up flags which define what we are going to select */
    show_idle = sel->idle;
    show_system = sel->system;
    show_uid = sel->uid != -1;
    show_command = sel->command != NULL;

    /* count up process states and get pointers to interesting procs */
    total_procs = 0;
    active_procs = 0;
    memset((char *)process_states, 0, sizeof(process_states));
    prefp = pref;
    for (pp = pbase, i = 0; i < nproc; pp++, i++)
    {
	/*
	 *  Place pointers to each valid proc structure in pref[].
	 *  Process slots that are actually in use have a non-zero
	 *  status field.  Processes with P_SYSTEM set are system
	 *  processes---these get ignored unless show_sysprocs is set.
	 */
	if (PP(pp, p_stat) != 0 &&
	    (show_system || ((PP(pp, p_flag) & P_SYSTEM) == 0)))
	{
	    total_procs++;
	    process_states[(unsigned char) PP(pp, p_stat)]++;
	    if ((PP(pp, p_stat) != SZOMB) &&
		(show_idle || (PP(pp, p_pctcpu) != 0) ||
		 (PP(pp, p_stat) == SRUN)) &&
		(!show_uid || EP(pp, e_pcred.p_ruid) == (uid_t)sel->uid))
	    {
		*prefp++ = pp;
		active_procs++;
	    }
	}
    }

    /* if requested, sort the "interesting" processes */
    if (compare != NULL)
    {
	qsort((char *)pref, active_procs, sizeof(struct kinfo_proc *), compare);
    }

    /* remember active and total counts */
    si->p_total = total_procs;
    si->p_active = pref_len = active_procs;

    /* pass back a handle */
    handle.next_proc = pref;
    handle.remaining = active_procs;
    return((caddr_t)&handle);
}

char fmt[128];		/* static area where result is built */

char *format_next_process(handle, get_userid)

caddr_t handle;
char *(*get_userid)();

{
    register struct kinfo_proc *pp;
    register long cputime;
    register double pct;
    struct handle *hp;
    char status[16];

    /* find and remember the next proc structure */
    hp = (struct handle *)handle;
    pp = *(hp->next_proc++);
    hp->remaining--;


    /* get the process's user struct and set cputime */
    if ((PP(pp, p_flag) & P_INMEM) == 0) {
	/*
	 * Print swapped processes as <pname>
	 */
	char *comm = PP(pp, p_comm);
#define COMSIZ sizeof(PP(pp, p_comm))
	char buf[COMSIZ];
	(void) strncpy(buf, comm, COMSIZ);
	comm[0] = '<';
	(void) strncpy(&comm[1], buf, COMSIZ - 2);
	comm[COMSIZ - 2] = '\0';
	(void) strncat(comm, ">", COMSIZ - 1);
	comm[COMSIZ - 1] = '\0';
    }

#if 0
    /* This does not produce the correct results */
    cputime = PP(pp, p_uticks) + PP(pp, p_sticks) + PP(pp, p_iticks);
#endif
    cputime = PP(pp, p_rtime).tv_sec;	/* This does not count interrupts */

    /* calculate the base for cpu percentages */
    pct = pctdouble(PP(pp, p_pctcpu));

    /* generate "STATE" field */
    switch (PP(pp, p_stat)) {
	case SRUN:
	    if (smpmode && PP(pp, p_oncpu) >= 0)
		sprintf(status, "CPU%d", PP(pp, p_oncpu));
	    else
		strcpy(status, "RUN");
	    break;
	case SSLEEP:
	    if (PP(pp, p_wmesg) != NULL) {
		sprintf(status, "%.6s", EP(pp, e_wmesg));
		break;
	    }
	    /* fall through */
	default:
	    sprintf(status, "%.6s", state_abbrev[(unsigned char) PP(pp, p_stat)]);
	    break;
    }

    /* format this entry */
    sprintf(fmt,
	    smpmode ? smp_Proc_format : up_Proc_format,
	    PP(pp, p_pid),
	    namelength, namelength,
	    (*get_userid)(EP(pp, e_pcred.p_ruid)),
	    PP(pp, p_priority) - PZERO,

	    /*
	     * normal time      -> nice value -20 - +20
	     * real time 0 - 31 -> nice value -52 - -21
	     * idle time 0 - 31 -> nice value +21 - +52
	     */
	    (PP(pp, p_rtprio.type) ==  RTP_PRIO_NORMAL ?
	    	PP(pp, p_nice) - NZERO :
	    	(PP(pp, p_rtprio.type) ==  RTP_PRIO_REALTIME ?
		    (PRIO_MIN - 1 - RTP_PRIO_MAX + PP(pp, p_rtprio.prio)) :
		    (PRIO_MAX + 1 + PP(pp, p_rtprio.prio)))),
	    format_k2(pagetok(PROCSIZE(pp))),
	    format_k2(pagetok(VP(pp, vm_rssize))),
	    status,
	    smpmode ? PP(pp, p_lastcpu) : 0,
	    format_time(cputime),
	    10000.0 * weighted_cpu(pct, pp) / hz,
	    10000.0 * pct / hz,
	    cmdlength,
	    printable(PP(pp, p_comm)));

    /* return the result */
    return(fmt);
}


/*
 * check_nlist(nlst) - checks the nlist to see if any symbols were not
 *		found.  For every symbol that was not found, a one-line
 *		message is printed to stderr.  The routine returns the
 *		number of symbols NOT found.
 */

static int check_nlist(nlst)

register struct nlist *nlst;

{
    register int i;

    /* check to see if we got ALL the symbols we requested */
    /* this will write one line to stderr for every symbol not found */

    i = 0;
    while (nlst->n_name != NULL)
    {
	if (nlst->n_type == 0)
	{
	    /* this one wasn't found */
	    (void) fprintf(stderr, "kernel: no symbol named `%s'\n",
			   nlst->n_name);
	    i = 1;
	}
	nlst++;
    }

    return(i);
}


/*
 *  getkval(offset, ptr, size, refstr) - get a value out of the kernel.
 *	"offset" is the byte offset into the kernel for the desired value,
 *  	"ptr" points to a buffer into which the value is retrieved,
 *  	"size" is the size of the buffer (and the object to retrieve),
 *  	"refstr" is a reference string used when printing error meessages,
 *	    if "refstr" starts with a '!', then a failure on read will not
 *  	    be fatal (this may seem like a silly way to do things, but I
 *  	    really didn't want the overhead of another argument).
 *
 */

static int getkval(offset, ptr, size, refstr)

unsigned long offset;
int *ptr;
int size;
char *refstr;

{
    if (kvm_read(kd, offset, (char *) ptr, size) != size)
    {
	if (*refstr == '!')
	{
	    return(0);
	}
	else
	{
	    fprintf(stderr, "top: kvm_read for %s: %s\n",
		refstr, strerror(errno));
	    quit(23);
	}
    }
    return(1);
}

/* comparison routine for qsort */

/*
 *  proc_compare - comparison function for "qsort"
 *	Compares the resource consumption of two processes using five
 *  	distinct keys.  The keys (in descending order of importance) are:
 *  	percent cpu, cpu ticks, state, resident set size, total virtual
 *  	memory usage.  The process states are ordered as follows (from least
 *  	to most important):  WAIT, zombie, sleep, stop, start, run.  The
 *  	array declaration below maps a process state index into a number
 *  	that reflects this ordering.
 */

static unsigned char sorted_state[] =
{
    0,	/* not used		*/
    3,	/* sleep		*/
    1,	/* ABANDONED (WAIT)	*/
    6,	/* run			*/
    5,	/* start		*/
    2,	/* zombie		*/
    4	/* stop			*/
};

int
proc_compare(pp1, pp2)

struct proc **pp1;
struct proc **pp2;

{
    register struct kinfo_proc *p1;
    register struct kinfo_proc *p2;
    register int result;
    register pctcpu lresult;

    /* remove one level of indirection */
    p1 = *(struct kinfo_proc **) pp1;
    p2 = *(struct kinfo_proc **) pp2;

    /* compare percent cpu (pctcpu) */
    if ((lresult = PP(p2, p_pctcpu) - PP(p1, p_pctcpu)) == 0)
    {
	/* use cpticks to break the tie */
	if ((result = PP(p2, p_cpticks) - PP(p1, p_cpticks)) == 0)
	{
	    /* use process state to break the tie */
	    if ((result = sorted_state[(unsigned char) PP(p2, p_stat)] -
			  sorted_state[(unsigned char) PP(p1, p_stat)])  == 0)
	    {
		/* use priority to break the tie */
		if ((result = PP(p2, p_priority) - PP(p1, p_priority)) == 0)
		{
		    /* use resident set size (rssize) to break the tie */
		    if ((result = VP(p2, vm_rssize) - VP(p1, vm_rssize)) == 0)
		    {
			/* use total memory to break the tie */
			result = PROCSIZE(p2) - PROCSIZE(p1);
		    }
		}
	    }
	}
    }
    else
    {
	result = lresult < 0 ? -1 : 1;
    }

    return(result);
}


/*
 * proc_owner(pid) - returns the uid that owns process "pid", or -1 if
 *		the process does not exist.
 *		It is EXTREMLY IMPORTANT that this function work correctly.
 *		If top runs setuid root (as in SVR4), then this function
 *		is the only thing that stands in the way of a serious
 *		security problem.  It validates requests for the "kill"
 *		and "renice" commands.
 */

int proc_owner(pid)

int pid;

{
    register int cnt;
    register struct kinfo_proc **prefp;
    register struct kinfo_proc *pp;

    prefp = pref;
    cnt = pref_len;
    while (--cnt >= 0)
    {
	pp = *prefp++;
	if (PP(pp, p_pid) == (pid_t)pid)
	{
	    return((int)EP(pp, e_pcred.p_ruid));
	}
    }
    return(-1);
}


/*
 * swapmode is based on a program called swapinfo written
 * by Kevin Lahey <kml@rokkaku.atl.ga.us>.
 */

#define	SVAR(var) __STRING(var)	/* to force expansion */
#define	KGET(idx, var)							\
	KGET1(idx, &var, sizeof(var), SVAR(var))
#define	KGET1(idx, p, s, msg)						\
	KGET2(nlst[idx].n_value, p, s, msg)
#define	KGET2(addr, p, s, msg)						\
	if (kvm_read(kd, (u_long)(addr), p, s) != s) {		        \
		warnx("cannot read %s: %s", msg, kvm_geterr(kd));       \
		return (0);                                             \
       }
#define	KGETRET(addr, p, s, msg)					\
	if (kvm_read(kd, (u_long)(addr), p, s) != s) {			\
		warnx("cannot read %s: %s", msg, kvm_geterr(kd));	\
		return (0);						\
	}


int
swapmode(retavail, retfree)
	int *retavail;
	int *retfree;
{
	char *header;
	int hlen, nswap, nswdev, dmmax;
	int i, div, avail, nfree, npfree, used;
	struct swdevt *sw;
	long blocksize, *perdev;
	u_long ptr;
	struct rlist head;
#if __FreeBSD_version >= 220000
	struct rlisthdr swaplist;
#else
	struct rlist *swaplist;
#endif
	struct rlist *swapptr;

	/*
	 * Counter for error messages. If we reach the limit,
	 * stop reading information from swap devices and
	 * return zero. This prevent endless 'bad address'
	 * messages.
	 */
	static warning = 10;

	if (warning <= 0) {
	    /* a single warning */
	    if (!warning) {
		warning--;
		fprintf(stderr,
			"Too much errors, stop reading swap devices ...\n");
		(void)sleep(3);
	    }
	    return(0);
	}
	warning--; /* decrease counter, see end of function */

	KGET(VM_NSWAP, nswap);
	if (!nswap) {
		fprintf(stderr, "No swap space available\n");
		return(0);
	}

	KGET(VM_NSWDEV, nswdev);
	KGET(VM_DMMAX, dmmax);
	KGET1(VM_SWAPLIST, &swaplist, sizeof(swaplist), "swaplist");
	if ((sw = (struct swdevt *)malloc(nswdev * sizeof(*sw))) == NULL ||
	    (perdev = (long *)malloc(nswdev * sizeof(*perdev))) == NULL)
		err(1, "malloc");
	KGET1(VM_SWDEVT, &ptr, sizeof ptr, "swdevt");
	KGET2(ptr, sw, nswdev * sizeof(*sw), "*swdevt");

	/* Count up swap space. */
	nfree = 0;
	memset(perdev, 0, nswdev * sizeof(*perdev));
#if  __FreeBSD_version >= 220000
	swapptr = swaplist.rlh_list;
	while (swapptr) {
#else
	while (swaplist) {
#endif
		int	top, bottom, next_block;
#if  __FreeBSD_version >= 220000
		KGET2(swapptr, &head, sizeof(struct rlist), "swapptr");
#else
		KGET2(swaplist, &head, sizeof(struct rlist), "swaplist");
#endif

		top = head.rl_end;
		bottom = head.rl_start;

		nfree += top - bottom + 1;

		/*
		 * Swap space is split up among the configured disks.
		 *
		 * For interleaved swap devices, the first dmmax blocks
		 * of swap space some from the first disk, the next dmmax
		 * blocks from the next, and so on up to nswap blocks.
		 *
		 * The list of free space joins adjacent free blocks,
		 * ignoring device boundries.  If we want to keep track
		 * of this information per device, we'll just have to
		 * extract it ourselves.
		 */
		while (top / dmmax != bottom / dmmax) {
			next_block = ((bottom + dmmax) / dmmax);
			perdev[(bottom / dmmax) % nswdev] +=
				next_block * dmmax - bottom;
			bottom = next_block * dmmax;
		}
		perdev[(bottom / dmmax) % nswdev] +=
			top - bottom + 1;

#if  __FreeBSD_version >= 220000
		swapptr = head.rl_next;
#else
		swaplist = head.rl_next;
#endif
	}

	header = getbsize(&hlen, &blocksize);
	div = blocksize / 512;
	avail = npfree = 0;
	for (i = 0; i < nswdev; i++) {
		int xsize, xfree;

		/*
		 * Don't report statistics for partitions which have not
		 * yet been activated via swapon(8).
		 */

		xsize = sw[i].sw_nblks;
		xfree = perdev[i];
		used = xsize - xfree;
		npfree++;
		avail += xsize;
	}

	/*
	 * If only one partition has been set up via swapon(8), we don't
	 * need to bother with totals.
	 */
	*retavail = avail / 2;
	*retfree = nfree / 2;
	used = avail - nfree;
	free(sw); free(perdev);

	/* increase counter, no errors occurs */
	warning++;

	return  (int)(((double)used / (double)avail * 100.0) + 0.5);
}