/*      $NetBSD: sched_m2.c,v 1.40 2024/01/24 16:11:48 christos Exp $   */

/*
* Copyright (c) 2007, 2008 Mindaugas Rasiukevicius <rmind at NetBSD org>
* 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.
*/

/*
* TODO:
*  - Implementation of fair share queue;
*  - Support for NUMA;
*/

#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: sched_m2.c,v 1.40 2024/01/24 16:11:48 christos Exp $");

#include <sys/param.h>

#include <sys/cpu.h>
#include <sys/callout.h>
#include <sys/errno.h>
#include <sys/kernel.h>
#include <sys/kmem.h>
#include <sys/lwp.h>
#include <sys/mutex.h>
#include <sys/pool.h>
#include <sys/proc.h>
#include <sys/pset.h>
#include <sys/resource.h>
#include <sys/resourcevar.h>
#include <sys/sched.h>
#include <sys/syscallargs.h>
#include <sys/sysctl.h>
#include <sys/types.h>

/*
* Priority related definitions.
*/
#define PRI_TS_COUNT    (NPRI_USER)
#define PRI_RT_COUNT    (PRI_COUNT - PRI_TS_COUNT)
#define PRI_HTS_RANGE   (PRI_TS_COUNT / 10)

#define PRI_HIGHEST_TS  (MAXPRI_USER)

/*
* Time-slices and priorities.
*/
static u_int    min_ts;                 /* Minimal time-slice */
static u_int    max_ts;                 /* Maximal time-slice */
static u_int    ts_map[PRI_COUNT];      /* Map of time-slices */
static pri_t    high_pri[PRI_COUNT];    /* Map for priority increase */
u_int           sched_rrticks;          /* Real-time time-slice */

static void     sched_precalcts(void);

/*
* Initialization and setup.
*/

void
sched_rqinit(void)
{
       if (hz < 100) {
               panic("sched_rqinit: value of HZ is too low\n");
       }

       /* Default timing ranges */
       min_ts = mstohz(20);                    /*  ~20 ms */
       max_ts = mstohz(150);                   /* ~150 ms */
       sched_rrticks = mstohz(100);                    /* ~100 ms */
       sched_precalcts();

#ifdef notyet
       /* Need to set the name etc. This does not belong here */
       /* Attach the primary CPU here */
       sched_cpuattach(curcpu());
#endif

       sched_lwp_fork(NULL, &lwp0);
#ifdef notyet
       /* without attaching the primary CPU l_mutex does not get initialized */
       lwp_lock(&lwp0);
       sched_newts(&lwp0);
       lwp_unlock(&lwp0);
#else
       /* gross */
       lwp0.l_sched.timeslice = ts_map[lwp0.l_auxprio];
#endif
}

/* Pre-calculate the time-slices for the priorities */
static void
sched_precalcts(void)
{
       pri_t p;

       /* Time-sharing range */
       for (p = 0; p <= PRI_HIGHEST_TS; p++) {
               ts_map[p] = max_ts -
                   (p * 100 / (PRI_TS_COUNT - 1) * (max_ts - min_ts) / 100);
               high_pri[p] = (PRI_HIGHEST_TS - PRI_HTS_RANGE) +
                   ((p * PRI_HTS_RANGE) / (PRI_TS_COUNT - 1));
       }

       /* Real-time range */
       for (p = (PRI_HIGHEST_TS + 1); p < PRI_COUNT; p++) {
               ts_map[p] = sched_rrticks;
               high_pri[p] = p;
       }
}

/*
* Hooks.
*/

void
sched_proc_fork(struct proc *parent, struct proc *child)
{
       struct lwp *l;

       LIST_FOREACH(l, &child->p_lwps, l_sibling) {
               lwp_lock(l);
               sched_newts(l);
               lwp_unlock(l);
       }
}

void
sched_proc_exit(struct proc *child, struct proc *parent)
{

}

void
sched_lwp_fork(struct lwp *l1, struct lwp *l2)
{

}

void
sched_lwp_collect(struct lwp *l)
{

}

void
sched_setrunnable(struct lwp *l)
{

}

void
sched_schedclock(struct lwp *l)
{

}

/*
* Priorities and time-slice.
*/

void
sched_nice(struct proc *p, int prio)
{
       struct lwp *l;
       int n;

       KASSERT(mutex_owned(p->p_lock));

       p->p_nice = prio;
       n = (prio - NZERO) >> 2;
       if (n == 0)
               return;

       LIST_FOREACH(l, &p->p_lwps, l_sibling) {
               lwp_lock(l);
               if (l->l_class == SCHED_OTHER) {
                       pri_t pri = l->l_priority - n;
                       pri = (n < 0) ? uimin(pri, PRI_HIGHEST_TS) : imax(pri, 0);
                       lwp_changepri(l, pri);
               }
               lwp_unlock(l);
       }
}

/* Recalculate the time-slice */
void
sched_newts(struct lwp *l)
{

       l->l_sched.timeslice = ts_map[lwp_eprio(l)];
}

void
sched_slept(struct lwp *l)
{

       /*
        * If thread is in time-sharing queue and batch flag is not marked,
        * increase the priority, and run with the lower time-quantum.
        */
       if (l->l_priority < PRI_HIGHEST_TS && (l->l_flag & LW_BATCH) == 0) {
               struct proc *p = l->l_proc;

               KASSERT(l->l_class == SCHED_OTHER);
               if (__predict_false(p->p_nice < NZERO)) {
                       const int n = uimax((NZERO - p->p_nice) >> 2, 1);
                       l->l_priority = uimin(l->l_priority + n, PRI_HIGHEST_TS);
               } else {
                       l->l_priority++;
               }
       }
}

void
sched_wakeup(struct lwp *l)
{

       /* If thread was sleeping a second or more - set a high priority */
       if (l->l_slptime >= 1)
               l->l_priority = high_pri[l->l_priority];
}

void
sched_pstats_hook(struct lwp *l, int batch)
{
       pri_t prio;

       /*
        * Estimate threads on time-sharing queue only, however,
        * exclude the highest priority for performance purposes.
        */
       KASSERT(lwp_locked(l, NULL));
       if (l->l_priority >= PRI_HIGHEST_TS)
               return;
       KASSERT(l->l_class == SCHED_OTHER);

       /* If it is CPU-bound not a first time - decrease the priority */
       prio = l->l_priority;
       if (batch && prio != 0)
               prio--;

       /* If thread was not ran a second or more - set a high priority */
       if (l->l_stat == LSRUN) {
               if (l->l_rticks && (getticks() - l->l_rticks >= hz))
                       prio = high_pri[prio];
               /* Re-enqueue the thread if priority has changed */
               if (prio != l->l_priority)
                       lwp_changepri(l, prio);
       } else {
               /* In other states, change the priority directly */
               l->l_priority = prio;
       }
}

void
sched_oncpu(lwp_t *l)
{
       struct schedstate_percpu *spc = &l->l_cpu->ci_schedstate;

       /* Update the counters */
       KASSERT(l->l_sched.timeslice >= min_ts);
       KASSERT(l->l_sched.timeslice <= max_ts);
       spc->spc_ticks = l->l_sched.timeslice;
}

/*
* Time-driven events.
*/

/*
* Called once per time-quantum, with the running LWP lock held (spc_lwplock).
*/
void
sched_tick(struct cpu_info *ci)
{
       struct schedstate_percpu *spc = &ci->ci_schedstate;
       struct lwp *l = ci->ci_onproc;
       struct proc *p;

       if (__predict_false(CURCPU_IDLE_P()))
               return;

       lwp_lock(l);
       KASSERT(l->l_mutex != spc->spc_mutex);
       switch (l->l_class) {
       case SCHED_FIFO:
               /*
                * Update the time-quantum, and continue running,
                * if thread runs on FIFO real-time policy.
                */
               KASSERT(l->l_priority > PRI_HIGHEST_TS);
               spc->spc_ticks = l->l_sched.timeslice;
               lwp_unlock(l);
               return;
       case SCHED_OTHER:
               /*
                * If thread is in time-sharing queue, decrease the priority,
                * and run with a higher time-quantum.
                */
               KASSERT(l->l_priority <= PRI_HIGHEST_TS);
               if (l->l_priority == 0)
                       break;

               p = l->l_proc;
               if (__predict_false(p->p_nice > NZERO)) {
                       const int n = uimax((p->p_nice - NZERO) >> 2, 1);
                       l->l_priority = imax(l->l_priority - n, 0);
               } else
                       l->l_priority--;
               break;
       }

       /*
        * If there are higher priority threads or threads in the same queue,
        * mark that thread should yield, otherwise, continue running.
        */
       if (lwp_eprio(l) <= spc->spc_maxpriority || l->l_target_cpu) {
               spc->spc_flags |= SPCF_SHOULDYIELD;
               spc_lock(ci);
               sched_resched_cpu(ci, MAXPRI_KTHREAD, true);
               /* spc now unlocked */
       } else
               spc->spc_ticks = l->l_sched.timeslice;
       lwp_unlock(l);
}

/*
* Sysctl nodes and initialization.
*/

static int
sysctl_sched_rtts(SYSCTLFN_ARGS)
{
       struct sysctlnode node;
       int rttsms = hztoms(sched_rrticks);

       node = *rnode;
       node.sysctl_data = &rttsms;
       return sysctl_lookup(SYSCTLFN_CALL(&node));
}

static int
sysctl_sched_mints(SYSCTLFN_ARGS)
{
       struct sysctlnode node;
       struct cpu_info *ci;
       int error, newsize;
       CPU_INFO_ITERATOR cii;

       node = *rnode;
       node.sysctl_data = &newsize;

       newsize = hztoms(min_ts);
       error = sysctl_lookup(SYSCTLFN_CALL(&node));
       if (error || newp == NULL)
               return error;

       newsize = mstohz(newsize);
       if (newsize < 1 || newsize > hz || newsize >= max_ts)
               return EINVAL;

       /* It is safe to do this in such order */
       for (CPU_INFO_FOREACH(cii, ci))
               spc_lock(ci);

       min_ts = newsize;
       sched_precalcts();

       for (CPU_INFO_FOREACH(cii, ci))
               spc_unlock(ci);

       return 0;
}

static int
sysctl_sched_maxts(SYSCTLFN_ARGS)
{
       struct sysctlnode node;
       struct cpu_info *ci;
       int error, newsize;
       CPU_INFO_ITERATOR cii;

       node = *rnode;
       node.sysctl_data = &newsize;

       newsize = hztoms(max_ts);
       error = sysctl_lookup(SYSCTLFN_CALL(&node));
       if (error || newp == NULL)
               return error;

       newsize = mstohz(newsize);
       if (newsize < 10 || newsize > hz || newsize <= min_ts)
               return EINVAL;

       /* It is safe to do this in such order */
       for (CPU_INFO_FOREACH(cii, ci))
               spc_lock(ci);

       max_ts = newsize;
       sched_precalcts();

       for (CPU_INFO_FOREACH(cii, ci))
               spc_unlock(ci);

       return 0;
}

SYSCTL_SETUP(sysctl_sched_m2_setup, "sysctl sched setup")
{
       const struct sysctlnode *node = NULL;

       sysctl_createv(clog, 0, NULL, &node,
               CTLFLAG_PERMANENT,
               CTLTYPE_NODE, "sched",
               SYSCTL_DESCR("Scheduler options"),
               NULL, 0, NULL, 0,
               CTL_KERN, CTL_CREATE, CTL_EOL);

       if (node == NULL)
               return;

       sysctl_createv(NULL, 0, &node, NULL,
               CTLFLAG_PERMANENT,
               CTLTYPE_STRING, "name", NULL,
               NULL, 0, __UNCONST("M2"), 0,
               CTL_CREATE, CTL_EOL);
       sysctl_createv(NULL, 0, &node, NULL,
               CTLFLAG_PERMANENT,
               CTLTYPE_INT, "rtts",
               SYSCTL_DESCR("Round-robin time quantum (in milliseconds)"),
               sysctl_sched_rtts, 0, NULL, 0,
               CTL_CREATE, CTL_EOL);
       sysctl_createv(NULL, 0, &node, NULL,
               CTLFLAG_PERMANENT | CTLFLAG_READWRITE,
               CTLTYPE_INT, "maxts",
               SYSCTL_DESCR("Maximal time quantum (in milliseconds)"),
               sysctl_sched_maxts, 0, &max_ts, 0,
               CTL_CREATE, CTL_EOL);
       sysctl_createv(NULL, 0, &node, NULL,
               CTLFLAG_PERMANENT | CTLFLAG_READWRITE,
               CTLTYPE_INT, "mints",
               SYSCTL_DESCR("Minimal time quantum (in milliseconds)"),
               sysctl_sched_mints, 0, &min_ts, 0,
               CTL_CREATE, CTL_EOL);
}

You are viewing proxied material from ftp.icm.edu.pl. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.