/*- * Copyright (c) 2005-2006, David Xu * 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 unmodified, 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 ``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 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. */ #include __FBSDID("$FreeBSD$"); #include "opt_hwpmc_hooks.h" #include "opt_sched.h" #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifdef KTRACE #include #include #endif #ifdef HWPMC_HOOKS #include #endif #include #include /* get process's nice value, skip value 20 which is not supported */ #define PROC_NICE(p) MIN((p)->p_nice, 19) /* convert nice to kernel thread priority */ #define NICE_TO_PRI(nice) (PUSER + 20 + (nice)) /* get process's static priority */ #define PROC_PRI(p) NICE_TO_PRI(PROC_NICE(p)) /* convert kernel thread priority to user priority */ #define USER_PRI(pri) MIN((pri) - PUSER, 39) /* convert nice value to user priority */ #define PROC_USER_PRI(p) (PROC_NICE(p) + 20) /* maximum user priority, highest prio + 1 */ #define MAX_USER_PRI 40 /* maximum kernel priority its nice is 19 */ #define PUSER_MAX (PUSER + 39) /* ticks and nanosecond converters */ #define NS_TO_HZ(n) ((n) / (1000000000 / hz)) #define HZ_TO_NS(h) ((h) * (1000000000 / hz)) /* ticks and microsecond converters */ #define MS_TO_HZ(m) ((m) / (1000000 / hz)) #define PRI_SCORE_RATIO 25 #define MAX_SCORE (MAX_USER_PRI * PRI_SCORE_RATIO / 100) #define MAX_SLEEP_TIME (def_timeslice * MAX_SCORE) #define NS_MAX_SLEEP_TIME (HZ_TO_NS(MAX_SLEEP_TIME)) #define STARVATION_TIME (MAX_SLEEP_TIME) #define CURRENT_SCORE(ts) \ (MAX_SCORE * NS_TO_HZ((ts)->ts_slptime) / MAX_SLEEP_TIME) #define SCALE_USER_PRI(x, upri) \ MAX(x * (upri + 1) / (MAX_USER_PRI/2), min_timeslice) /* * For a thread whose nice is zero, the score is used to determine * if it is an interactive thread. */ #define INTERACTIVE_BASE_SCORE (MAX_SCORE * 20)/100 /* * Calculate a score which a thread must have to prove itself is * an interactive thread. */ #define INTERACTIVE_SCORE(ts) \ (PROC_NICE((ts)->ts_proc) * MAX_SCORE / 40 + INTERACTIVE_BASE_SCORE) /* Test if a thread is an interactive thread */ #define THREAD_IS_INTERACTIVE(ts) \ ((ts)->ts_thread->td_user_pri <= \ PROC_PRI((ts)->ts_proc) - INTERACTIVE_SCORE(ts)) /* * Calculate how long a thread must sleep to prove itself is an * interactive sleep. */ #define INTERACTIVE_SLEEP_TIME(ts) \ (HZ_TO_NS(MAX_SLEEP_TIME * \ (MAX_SCORE / 2 + INTERACTIVE_SCORE((ts)) + 1) / MAX_SCORE - 1)) #define CHILD_WEIGHT 90 #define PARENT_WEIGHT 90 #define EXIT_WEIGHT 3 #define SCHED_LOAD_SCALE 128UL #define IDLE 0 #define IDLE_IDLE 1 #define NOT_IDLE 2 #define KQB_LEN (8) /* Number of priority status words. */ #define KQB_L2BPW (5) /* Log2(sizeof(rqb_word_t) * NBBY)). */ #define KQB_BPW (1<> KQB_L2BPW) #define KQB_FFS(word) (ffs(word) - 1) #define KQ_NQS 256 /* * Type of run queue status word. */ typedef u_int32_t kqb_word_t; /* * Head of run queues. */ TAILQ_HEAD(krqhead, td_sched); /* * Bit array which maintains the status of a run queue. When a queue is * non-empty the bit corresponding to the queue number will be set. */ struct krqbits { kqb_word_t rqb_bits[KQB_LEN]; }; /* * Run queue structure. Contains an array of run queues on which processes * are placed, and a structure to maintain the status of each queue. */ struct krunq { struct krqbits rq_status; struct krqhead rq_queues[KQ_NQS]; }; /* * The following datastructures are allocated within their parent structure * but are scheduler specific. */ /* * The schedulable entity that can be given a context to run. A process may * have several of these. */ struct td_sched { struct thread *ts_thread; /* (*) Active associated thread. */ TAILQ_ENTRY(td_sched) ts_procq; /* (j/z) Run queue. */ int ts_flags; /* (j) TSF_* flags. */ fixpt_t ts_pctcpu; /* (j) %cpu during p_swtime. */ u_char ts_rqindex; /* (j) Run queue index. */ enum { TSS_THREAD = 0x0, /* slaved to thread state */ TSS_ONRUNQ } ts_state; /* (j) thread sched specific status. */ int ts_slice; /* Time slice in ticks */ struct kseq *ts_kseq; /* Kseq the thread belongs to */ struct krunq *ts_runq; /* Assiociated runqueue */ #ifdef SMP int ts_cpu; /* CPU that we have affinity for. */ int ts_wakeup_cpu; /* CPU that has activated us. */ #endif int ts_activated; /* How is the thread activated. */ uint64_t ts_timestamp; /* Last timestamp dependent on state.*/ unsigned ts_lastran; /* Last timestamp the thread ran. */ /* The following variables are only used for pctcpu calculation */ int ts_ltick; /* Last tick that we were running on */ int ts_ftick; /* First tick that we were running on */ int ts_ticks; /* Tick count */ u_long ts_slptime; /* (j) Number of ticks we vol. slept */ u_long ts_runtime; /* (j) Temp total run time. */ }; #define td_sched td_sched #define ts_proc ts_thread->td_proc /* flags kept in ts_flags */ #define TSF_BOUND 0x0001 /* Thread can not migrate. */ #define TSF_PREEMPTED 0x0002 /* Thread was preempted. */ #define TSF_MIGRATING 0x0004 /* Thread is migrating. */ #define TSF_SLEEP 0x0008 /* Thread did sleep. */ #define TSF_DIDRUN 0x0010 /* Thread actually ran. */ #define TSF_EXIT 0x0020 /* Thread is being killed. */ #define TSF_NEXTRQ 0x0400 /* Thread should be in next queue. */ #define TSF_FIRST_SLICE 0x0800 /* Thread has first time slice left. */ /* * Cpu percentage computation macros and defines. * * SCHED_CPU_TIME: Number of seconds to average the cpu usage across. * SCHED_CPU_TICKS: Number of hz ticks to average the cpu usage across. */ #define SCHED_CPU_TIME 10 #define SCHED_CPU_TICKS (hz * SCHED_CPU_TIME) /* * kseq - per processor runqs and statistics. */ struct kseq { struct krunq *ksq_curr; /* Current queue. */ struct krunq *ksq_next; /* Next timeshare queue. */ struct krunq ksq_timeshare[2]; /* Run queues for !IDLE. */ struct krunq ksq_idle; /* Queue of IDLE threads. */ int ksq_load; uint64_t ksq_last_timestamp; /* Per-cpu last clock tick */ unsigned ksq_expired_tick; /* First expired tick */ signed char ksq_expired_nice; /* Lowest nice in nextq */ }; static struct td_sched kse0; static int min_timeslice = 5; static int def_timeslice = 100; static int granularity = 10; static int realstathz; static int sched_tdcnt; static struct kseq kseq_global; /* * One td_sched queue per processor. */ #ifdef SMP static struct kseq kseq_cpu[MAXCPU]; #define KSEQ_SELF() (&kseq_cpu[PCPU_GET(cpuid)]) #define KSEQ_CPU(x) (&kseq_cpu[(x)]) #define KSEQ_ID(x) ((x) - kseq_cpu) static cpumask_t cpu_sibling[MAXCPU]; #else /* !SMP */ #define KSEQ_SELF() (&kseq_global) #define KSEQ_CPU(x) (&kseq_global) #endif /* decay 95% of `p_pctcpu' in 60 seconds; see CCPU_SHIFT before changing */ static fixpt_t ccpu = 0.95122942450071400909 * FSCALE; /* exp(-1/20) */ SYSCTL_INT(_kern, OID_AUTO, ccpu, CTLFLAG_RD, &ccpu, 0, ""); static void sched_setup(void *dummy); SYSINIT(sched_setup, SI_SUB_RUN_QUEUE, SI_ORDER_FIRST, sched_setup, NULL); static void sched_initticks(void *dummy); SYSINIT(sched_initticks, SI_SUB_CLOCKS, SI_ORDER_THIRD, sched_initticks, NULL) static SYSCTL_NODE(_kern, OID_AUTO, sched, CTLFLAG_RW, 0, "Scheduler"); SYSCTL_STRING(_kern_sched, OID_AUTO, name, CTLFLAG_RD, "CORE", 0, "Scheduler name"); #ifdef SMP /* Enable forwarding of wakeups to all other cpus */ SYSCTL_NODE(_kern_sched, OID_AUTO, ipiwakeup, CTLFLAG_RD, NULL, "Kernel SMP"); static int runq_fuzz = 0; SYSCTL_INT(_kern_sched, OID_AUTO, runq_fuzz, CTLFLAG_RW, &runq_fuzz, 0, ""); static int forward_wakeup_enabled = 1; SYSCTL_INT(_kern_sched_ipiwakeup, OID_AUTO, enabled, CTLFLAG_RW, &forward_wakeup_enabled, 0, "Forwarding of wakeup to idle CPUs"); static int forward_wakeups_requested = 0; SYSCTL_INT(_kern_sched_ipiwakeup, OID_AUTO, requested, CTLFLAG_RD, &forward_wakeups_requested, 0, "Requests for Forwarding of wakeup to idle CPUs"); static int forward_wakeups_delivered = 0; SYSCTL_INT(_kern_sched_ipiwakeup, OID_AUTO, delivered, CTLFLAG_RD, &forward_wakeups_delivered, 0, "Completed Forwarding of wakeup to idle CPUs"); static int forward_wakeup_use_mask = 1; SYSCTL_INT(_kern_sched_ipiwakeup, OID_AUTO, usemask, CTLFLAG_RW, &forward_wakeup_use_mask, 0, "Use the mask of idle cpus"); static int forward_wakeup_use_loop = 0; SYSCTL_INT(_kern_sched_ipiwakeup, OID_AUTO, useloop, CTLFLAG_RW, &forward_wakeup_use_loop, 0, "Use a loop to find idle cpus"); static int forward_wakeup_use_single = 0; SYSCTL_INT(_kern_sched_ipiwakeup, OID_AUTO, onecpu, CTLFLAG_RW, &forward_wakeup_use_single, 0, "Only signal one idle cpu"); static int forward_wakeup_use_htt = 0; SYSCTL_INT(_kern_sched_ipiwakeup, OID_AUTO, htt2, CTLFLAG_RW, &forward_wakeup_use_htt, 0, "account for htt"); #endif static void krunq_add(struct krunq *, struct td_sched *); static struct td_sched *krunq_choose(struct krunq *); static void krunq_clrbit(struct krunq *rq, int pri); static int krunq_findbit(struct krunq *rq); static void krunq_init(struct krunq *); static void krunq_remove(struct krunq *, struct td_sched *); static struct td_sched * kseq_choose(struct kseq *); static void kseq_load_add(struct kseq *, struct td_sched *); static void kseq_load_rem(struct kseq *, struct td_sched *); static void kseq_runq_add(struct kseq *, struct td_sched *); static void kseq_runq_rem(struct kseq *, struct td_sched *); static void kseq_setup(struct kseq *); static int sched_is_timeshare(struct thread *td); static struct td_sched *sched_choose(void); static int sched_calc_pri(struct td_sched *ts); static int sched_starving(struct kseq *, unsigned, struct td_sched *); static void sched_pctcpu_update(struct td_sched *); static void sched_thread_priority(struct thread *, u_char); static uint64_t sched_timestamp(void); static int sched_recalc_pri(struct td_sched *ts, uint64_t now); static int sched_timeslice(struct td_sched *ts); static void sched_update_runtime(struct td_sched *ts, uint64_t now); static void sched_commit_runtime(struct td_sched *ts); /* * Initialize a run structure. */ static void krunq_init(struct krunq *rq) { int i; bzero(rq, sizeof *rq); for (i = 0; i < KQ_NQS; i++) TAILQ_INIT(&rq->rq_queues[i]); } /* * Clear the status bit of the queue corresponding to priority level pri, * indicating that it is empty. */ static inline void krunq_clrbit(struct krunq *rq, int pri) { struct krqbits *rqb; rqb = &rq->rq_status; rqb->rqb_bits[KQB_WORD(pri)] &= ~KQB_BIT(pri); } /* * Find the index of the first non-empty run queue. This is done by * scanning the status bits, a set bit indicates a non-empty queue. */ static int krunq_findbit(struct krunq *rq) { struct krqbits *rqb; int pri; int i; rqb = &rq->rq_status; for (i = 0; i < KQB_LEN; i++) { if (rqb->rqb_bits[i]) { pri = KQB_FFS(rqb->rqb_bits[i]) + (i << KQB_L2BPW); return (pri); } } return (-1); } static int krunq_check(struct krunq *rq) { struct krqbits *rqb; int i; rqb = &rq->rq_status; for (i = 0; i < KQB_LEN; i++) { if (rqb->rqb_bits[i]) return (1); } return (0); } /* * Set the status bit of the queue corresponding to priority level pri, * indicating that it is non-empty. */ static inline void krunq_setbit(struct krunq *rq, int pri) { struct krqbits *rqb; rqb = &rq->rq_status; rqb->rqb_bits[KQB_WORD(pri)] |= KQB_BIT(pri); } /* * Add the KSE to the queue specified by its priority, and set the * corresponding status bit. */ static void krunq_add(struct krunq *rq, struct td_sched *ts) { struct krqhead *rqh; int pri; pri = ts->ts_thread->td_priority; ts->ts_rqindex = pri; krunq_setbit(rq, pri); rqh = &rq->rq_queues[pri]; if (ts->ts_flags & TSF_PREEMPTED) TAILQ_INSERT_HEAD(rqh, ts, ts_procq); else TAILQ_INSERT_TAIL(rqh, ts, ts_procq); } /* * Find the highest priority process on the run queue. */ static struct td_sched * krunq_choose(struct krunq *rq) { struct krqhead *rqh; struct td_sched *ts; int pri; mtx_assert(&sched_lock, MA_OWNED); if ((pri = krunq_findbit(rq)) != -1) { rqh = &rq->rq_queues[pri]; ts = TAILQ_FIRST(rqh); KASSERT(ts != NULL, ("krunq_choose: no thread on busy queue")); #ifdef SMP if (pri <= PRI_MAX_ITHD || runq_fuzz <= 0) return (ts); /* * In the first couple of entries, check if * there is one for our CPU as a preference. */ struct td_sched *ts2 = ts; const int mycpu = PCPU_GET(cpuid); const int mymask = 1 << mycpu; int count = runq_fuzz; while (count-- && ts2) { const int cpu = ts2->ts_wakeup_cpu; if (cpu_sibling[cpu] & mymask) { ts = ts2; break; } ts2 = TAILQ_NEXT(ts2, ts_procq); } #endif return (ts); } return (NULL); } /* * Remove the KSE from the queue specified by its priority, and clear the * corresponding status bit if the queue becomes empty. * Caller must set ts->ts_state afterwards. */ static void krunq_remove(struct krunq *rq, struct td_sched *ts) { struct krqhead *rqh; int pri; KASSERT(ts->ts_proc->p_sflag & PS_INMEM, ("runq_remove: process swapped out")); pri = ts->ts_rqindex; rqh = &rq->rq_queues[pri]; KASSERT(ts != NULL, ("krunq_remove: no proc on busy queue")); TAILQ_REMOVE(rqh, ts, ts_procq); if (TAILQ_EMPTY(rqh)) krunq_clrbit(rq, pri); } static inline void kseq_runq_add(struct kseq *kseq, struct td_sched *ts) { krunq_add(ts->ts_runq, ts); ts->ts_kseq = kseq; } static inline void kseq_runq_rem(struct kseq *kseq, struct td_sched *ts) { krunq_remove(ts->ts_runq, ts); ts->ts_kseq = NULL; ts->ts_runq = NULL; } static inline void kseq_load_add(struct kseq *kseq, struct td_sched *ts) { kseq->ksq_load++; if ((ts->ts_proc->p_flag & P_NOLOAD) == 0) sched_tdcnt++; } static inline void kseq_load_rem(struct kseq *kseq, struct td_sched *ts) { kseq->ksq_load--; if ((ts->ts_proc->p_flag & P_NOLOAD) == 0) sched_tdcnt--; } /* * Pick the highest priority task we have and return it. */ static struct td_sched * kseq_choose(struct kseq *kseq) { struct krunq *swap; struct td_sched *ts; mtx_assert(&sched_lock, MA_OWNED); ts = krunq_choose(kseq->ksq_curr); if (ts != NULL) return (ts); kseq->ksq_expired_nice = PRIO_MAX + 1; kseq->ksq_expired_tick = 0; swap = kseq->ksq_curr; kseq->ksq_curr = kseq->ksq_next; kseq->ksq_next = swap; ts = krunq_choose(kseq->ksq_curr); if (ts != NULL) return (ts); return krunq_choose(&kseq->ksq_idle); } static inline uint64_t sched_timestamp(void) { uint64_t now = cputick2usec(cpu_ticks()) * 1000; return (now); } static inline int sched_timeslice(struct td_sched *ts) { struct proc *p = ts->ts_proc; if (ts->ts_proc->p_nice < 0) return SCALE_USER_PRI(def_timeslice*4, PROC_USER_PRI(p)); else return SCALE_USER_PRI(def_timeslice, PROC_USER_PRI(p)); } static inline int sched_is_timeshare(struct thread *td) { return (td->td_pri_class == PRI_TIMESHARE); } static int sched_calc_pri(struct td_sched *ts) { int score, pri; if (sched_is_timeshare(ts->ts_thread)) { score = CURRENT_SCORE(ts) - MAX_SCORE / 2; pri = PROC_PRI(ts->ts_proc) - score; if (pri < PUSER) pri = PUSER; else if (pri > PUSER_MAX) pri = PUSER_MAX; return (pri); } return (ts->ts_thread->td_base_user_pri); } static int sched_recalc_pri(struct td_sched *ts, uint64_t now) { uint64_t delta; unsigned int sleep_time; delta = now - ts->ts_timestamp; if (__predict_false(!sched_is_timeshare(ts->ts_thread))) return (ts->ts_thread->td_base_user_pri); if (delta > NS_MAX_SLEEP_TIME) sleep_time = NS_MAX_SLEEP_TIME; else sleep_time = (unsigned int)delta; if (__predict_false(sleep_time == 0)) goto out; if (ts->ts_activated != -1 && sleep_time > INTERACTIVE_SLEEP_TIME(ts)) { ts->ts_slptime = HZ_TO_NS(MAX_SLEEP_TIME - def_timeslice); } else { sleep_time *= (MAX_SCORE - CURRENT_SCORE(ts)) ? : 1; /* * If thread is waking from uninterruptible sleep, it is * unlikely an interactive sleep, limit its sleep time to * prevent it from being an interactive thread. */ if (ts->ts_activated == -1) { if (ts->ts_slptime >= INTERACTIVE_SLEEP_TIME(ts)) sleep_time = 0; else if (ts->ts_slptime + sleep_time >= INTERACTIVE_SLEEP_TIME(ts)) { ts->ts_slptime = INTERACTIVE_SLEEP_TIME(ts); sleep_time = 0; } } /* * Thread gets priority boost here. */ ts->ts_slptime += sleep_time; /* Sleep time should never be larger than maximum */ if (ts->ts_slptime > NS_MAX_SLEEP_TIME) ts->ts_slptime = NS_MAX_SLEEP_TIME; } out: return (sched_calc_pri(ts)); } static void sched_update_runtime(struct td_sched *ts, uint64_t now) { uint64_t runtime; if (sched_is_timeshare(ts->ts_thread)) { if ((int64_t)(now - ts->ts_timestamp) < NS_MAX_SLEEP_TIME) { runtime = now - ts->ts_timestamp; if ((int64_t)(now - ts->ts_timestamp) < 0) runtime = 0; } else { runtime = NS_MAX_SLEEP_TIME; } runtime /= (CURRENT_SCORE(ts) ? : 1); ts->ts_runtime += runtime; ts->ts_timestamp = now; } } static void sched_commit_runtime(struct td_sched *ts) { if (ts->ts_runtime > ts->ts_slptime) ts->ts_slptime = 0; else ts->ts_slptime -= ts->ts_runtime; ts->ts_runtime = 0; } static void kseq_setup(struct kseq *kseq) { krunq_init(&kseq->ksq_timeshare[0]); krunq_init(&kseq->ksq_timeshare[1]); krunq_init(&kseq->ksq_idle); kseq->ksq_curr = &kseq->ksq_timeshare[0]; kseq->ksq_next = &kseq->ksq_timeshare[1]; kseq->ksq_expired_nice = PRIO_MAX + 1; kseq->ksq_expired_tick = 0; } static void sched_setup(void *dummy) { #ifdef SMP int i; #endif /* * To avoid divide-by-zero, we set realstathz a dummy value * in case which sched_clock() called before sched_initticks(). */ realstathz = hz; min_timeslice = MAX(5 * hz / 1000, 1); def_timeslice = MAX(100 * hz / 1000, 1); granularity = MAX(10 * hz / 1000, 1); kseq_setup(&kseq_global); #ifdef SMP runq_fuzz = MIN(mp_ncpus * 2, 8); /* * Initialize the kseqs. */ for (i = 0; i < MAXCPU; i++) { struct kseq *ksq; ksq = &kseq_cpu[i]; kseq_setup(&kseq_cpu[i]); cpu_sibling[i] = 1 << i; } if (smp_topology != NULL) { int i, j; cpumask_t visited; struct cpu_group *cg; visited = 0; for (i = 0; i < smp_topology->ct_count; i++) { cg = &smp_topology->ct_group[i]; if (cg->cg_mask & visited) panic("duplicated cpumask in ct_group."); if (cg->cg_mask == 0) continue; visited |= cg->cg_mask; for (j = 0; j < MAXCPU; j++) { if ((cg->cg_mask & (1 << j)) != 0) cpu_sibling[j] |= cg->cg_mask; } } } #endif mtx_lock_spin(&sched_lock); kseq_load_add(KSEQ_SELF(), &kse0); mtx_unlock_spin(&sched_lock); } /* ARGSUSED */ static void sched_initticks(void *dummy) { mtx_lock_spin(&sched_lock); realstathz = stathz ? stathz : hz; mtx_unlock_spin(&sched_lock); } /* * Very early in the boot some setup of scheduler-specific * parts of proc0 and of soem scheduler resources needs to be done. * Called from: * proc0_init() */ void schedinit(void) { /* * Set up the scheduler specific parts of proc0. */ proc0.p_sched = NULL; /* XXX */ thread0.td_sched = &kse0; kse0.ts_thread = &thread0; kse0.ts_state = TSS_THREAD; kse0.ts_slice = 100; } /* * This is only somewhat accurate since given many processes of the same * priority they will switch when their slices run out, which will be * at most SCHED_SLICE_MAX. */ int sched_rr_interval(void) { return (def_timeslice); } static void sched_pctcpu_update(struct td_sched *ts) { /* * Adjust counters and watermark for pctcpu calc. */ if (ts->ts_ltick > ticks - SCHED_CPU_TICKS) { /* * Shift the tick count out so that the divide doesn't * round away our results. */ ts->ts_ticks <<= 10; ts->ts_ticks = (ts->ts_ticks / (ticks - ts->ts_ftick)) * SCHED_CPU_TICKS; ts->ts_ticks >>= 10; } else ts->ts_ticks = 0; ts->ts_ltick = ticks; ts->ts_ftick = ts->ts_ltick - SCHED_CPU_TICKS; } static void sched_thread_priority(struct thread *td, u_char prio) { struct td_sched *ts; ts = td->td_sched; mtx_assert(&sched_lock, MA_OWNED); if (__predict_false(td->td_priority == prio)) return; if (TD_ON_RUNQ(td)) { /* * If the priority has been elevated due to priority * propagation, we may have to move ourselves to a new * queue. We still call adjustrunqueue below in case td_sched * needs to fix things up. */ if (prio < td->td_priority && ts->ts_runq != NULL && ts->ts_runq != ts->ts_kseq->ksq_curr) { krunq_remove(ts->ts_runq, ts); ts->ts_runq = ts->ts_kseq->ksq_curr; krunq_add(ts->ts_runq, ts); } adjustrunqueue(td, prio); } else td->td_priority = prio; } /* * Update a thread's priority when it is lent another thread's * priority. */ void sched_lend_prio(struct thread *td, u_char prio) { td->td_flags |= TDF_BORROWING; sched_thread_priority(td, prio); } /* * Restore a thread's priority when priority propagation is * over. The prio argument is the minimum priority the thread * needs to have to satisfy other possible priority lending * requests. If the thread's regular priority is less * important than prio, the thread will keep a priority boost * of prio. */ void sched_unlend_prio(struct thread *td, u_char prio) { u_char base_pri; if (td->td_base_pri >= PRI_MIN_TIMESHARE && td->td_base_pri <= PRI_MAX_TIMESHARE) base_pri = td->td_user_pri; else base_pri = td->td_base_pri; if (prio >= base_pri) { td->td_flags &= ~TDF_BORROWING; sched_thread_priority(td, base_pri); } else sched_lend_prio(td, prio); } void sched_prio(struct thread *td, u_char prio) { u_char oldprio; if (td->td_pri_class == PRI_TIMESHARE) prio = MIN(prio, PUSER_MAX); /* First, update the base priority. */ td->td_base_pri = prio; /* * If the thread is borrowing another thread's priority, don't * ever lower the priority. */ if (td->td_flags & TDF_BORROWING && td->td_priority < prio) return; /* Change the real priority. */ oldprio = td->td_priority; sched_thread_priority(td, prio); /* * If the thread is on a turnstile, then let the turnstile update * its state. */ if (TD_ON_LOCK(td) && oldprio != prio) turnstile_adjust(td, oldprio); } void sched_user_prio(struct thread *td, u_char prio) { u_char oldprio; td->td_base_user_pri = prio; if (td->td_flags & TDF_UBORROWING && td->td_user_pri <= prio) return; oldprio = td->td_user_pri; td->td_user_pri = prio; if (TD_ON_UPILOCK(td) && oldprio != prio) umtx_pi_adjust(td, oldprio); } void sched_lend_user_prio(struct thread *td, u_char prio) { u_char oldprio; td->td_flags |= TDF_UBORROWING; oldprio = td->td_user_pri; td->td_user_pri = prio; if (TD_ON_UPILOCK(td) && oldprio != prio) umtx_pi_adjust(td, oldprio); } void sched_unlend_user_prio(struct thread *td, u_char prio) { u_char base_pri; base_pri = td->td_base_user_pri; if (prio >= base_pri) { td->td_flags &= ~TDF_UBORROWING; sched_user_prio(td, base_pri); } else sched_lend_user_prio(td, prio); } void sched_switch(struct thread *td, struct thread *newtd, int flags) { struct kseq *ksq; struct td_sched *ts; uint64_t now; mtx_assert(&sched_lock, MA_OWNED); now = sched_timestamp(); ts = td->td_sched; ksq = KSEQ_SELF(); td->td_lastcpu = td->td_oncpu; td->td_oncpu = NOCPU; td->td_flags &= ~TDF_NEEDRESCHED; td->td_owepreempt = 0; if (td == PCPU_GET(idlethread)) { TD_SET_CAN_RUN(td); } else { sched_update_runtime(ts, now); /* We are ending our run so make our slot available again */ kseq_load_rem(ksq, ts); if (TD_IS_RUNNING(td)) { setrunqueue(td, (flags & SW_PREEMPT) ? SRQ_OURSELF|SRQ_YIELDING|SRQ_PREEMPTED : SRQ_OURSELF|SRQ_YIELDING); } else { ts->ts_flags &= ~TSF_NEXTRQ; } } if (newtd != NULL) { /* * If we bring in a thread account for it as if it had been * added to the run queue and then chosen. */ newtd->td_sched->ts_flags |= TSF_DIDRUN; newtd->td_sched->ts_timestamp = now; TD_SET_RUNNING(newtd); kseq_load_add(ksq, newtd->td_sched); } else { newtd = choosethread(); /* sched_choose sets ts_timestamp, just reuse it */ } if (td != newtd) { ts->ts_lastran = tick; #ifdef HWPMC_HOOKS if (PMC_PROC_IS_USING_PMCS(td->td_proc)) PMC_SWITCH_CONTEXT(td, PMC_FN_CSW_OUT); #endif cpu_switch(td, newtd); #ifdef HWPMC_HOOKS if (PMC_PROC_IS_USING_PMCS(td->td_proc)) PMC_SWITCH_CONTEXT(td, PMC_FN_CSW_IN); #endif } sched_lock.mtx_lock = (uintptr_t)td; td->td_oncpu = PCPU_GET(cpuid); } void sched_nice(struct proc *p, int nice) { struct thread *td; PROC_LOCK_ASSERT(p, MA_OWNED); mtx_assert(&sched_lock, MA_OWNED); p->p_nice = nice; FOREACH_THREAD_IN_PROC(p, td) { if (td->td_pri_class == PRI_TIMESHARE) { sched_user_prio(td, sched_calc_pri(td->td_sched)); td->td_flags |= TDF_NEEDRESCHED; } } } void sched_sleep(struct thread *td) { struct td_sched *ts; mtx_assert(&sched_lock, MA_OWNED); ts = td->td_sched; if (td->td_flags & TDF_SINTR) ts->ts_activated = 0; else ts->ts_activated = -1; ts->ts_flags |= TSF_SLEEP; } void sched_wakeup(struct thread *td) { struct td_sched *ts; struct kseq *kseq, *mykseq; uint64_t now; mtx_assert(&sched_lock, MA_OWNED); ts = td->td_sched; mykseq = KSEQ_SELF(); if (ts->ts_flags & TSF_SLEEP) { ts->ts_flags &= ~TSF_SLEEP; if (sched_is_timeshare(td)) { sched_commit_runtime(ts); now = sched_timestamp(); kseq = KSEQ_CPU(td->td_lastcpu); #ifdef SMP if (kseq != mykseq) now = now - mykseq->ksq_last_timestamp + kseq->ksq_last_timestamp; #endif sched_user_prio(td, sched_recalc_pri(ts, now)); } } setrunqueue(td, SRQ_BORING); } /* * Penalize the parent for creating a new child and initialize the child's * priority. */ void sched_fork(struct thread *td, struct thread *childtd) { mtx_assert(&sched_lock, MA_OWNED); sched_fork_thread(td, childtd); } void sched_fork_thread(struct thread *td, struct thread *child) { struct td_sched *ts; struct td_sched *ts2; sched_newthread(child); ts = td->td_sched; ts2 = child->td_sched; ts2->ts_slptime = ts2->ts_slptime * CHILD_WEIGHT / 100; if (child->td_pri_class == PRI_TIMESHARE) sched_user_prio(child, sched_calc_pri(ts2)); ts->ts_slptime = ts->ts_slptime * PARENT_WEIGHT / 100; ts2->ts_slice = (ts->ts_slice + 1) >> 1; ts2->ts_flags |= TSF_FIRST_SLICE | (ts->ts_flags & TSF_NEXTRQ); ts2->ts_activated = 0; ts->ts_slice >>= 1; if (ts->ts_slice == 0) { ts->ts_slice = 1; sched_tick(); } /* Grab our parents cpu estimation information. */ ts2->ts_ticks = ts->ts_ticks; ts2->ts_ltick = ts->ts_ltick; ts2->ts_ftick = ts->ts_ftick; } void sched_class(struct thread *td, int class) { mtx_assert(&sched_lock, MA_OWNED); td->td_pri_class = class; } /* * Return some of the child's priority and interactivity to the parent. */ void sched_exit(struct proc *p, struct thread *childtd) { mtx_assert(&sched_lock, MA_OWNED); sched_exit_thread(FIRST_THREAD_IN_PROC(p), childtd); } void sched_exit_thread(struct thread *td, struct thread *childtd) { struct td_sched *childke = childtd->td_sched; struct td_sched *parentke = td->td_sched; if (childke->ts_slptime < parentke->ts_slptime) { parentke->ts_slptime = parentke->ts_slptime / (EXIT_WEIGHT) * (EXIT_WEIGHT - 1) + parentke->ts_slptime / EXIT_WEIGHT; } kseq_load_rem(KSEQ_SELF(), childke); sched_update_runtime(childke, sched_timestamp()); sched_commit_runtime(childke); if ((childke->ts_flags & TSF_FIRST_SLICE) && td->td_proc == childtd->td_proc->p_pptr) { parentke->ts_slice += childke->ts_slice; if (parentke->ts_slice > sched_timeslice(parentke)) parentke->ts_slice = sched_timeslice(parentke); } } static int sched_starving(struct kseq *ksq, unsigned now, struct td_sched *ts) { uint64_t delta; if (ts->ts_proc->p_nice > ksq->ksq_expired_nice) return (1); if (ksq->ksq_expired_tick == 0) return (0); delta = HZ_TO_NS((uint64_t)now - ksq->ksq_expired_tick); if (delta > STARVATION_TIME * ksq->ksq_load) return (1); return (0); } /* * An interactive thread has smaller time slice granularity, * a cpu hog can have larger granularity. */ static inline int sched_timeslice_split(struct td_sched *ts) { int score, g; score = (int)(MAX_SCORE - CURRENT_SCORE(ts)); if (score == 0) score = 1; #ifdef SMP g = granularity * ((1 << score) - 1) * smp_cpus; #else g = granularity * ((1 << score) - 1); #endif return (ts->ts_slice >= g && ts->ts_slice % g == 0); } void sched_tick(void) { struct thread *td; struct proc *p; struct td_sched *ts; struct kseq *kseq; uint64_t now; int cpuid; int class; mtx_assert(&sched_lock, MA_OWNED); td = curthread; ts = td->td_sched; p = td->td_proc; class = PRI_BASE(td->td_pri_class); now = sched_timestamp(); cpuid = PCPU_GET(cpuid); kseq = KSEQ_CPU(cpuid); kseq->ksq_last_timestamp = now; if (class == PRI_IDLE) { /* * Processes of equal idle priority are run round-robin. */ if (td != PCPU_GET(idlethread) && --ts->ts_slice <= 0) { ts->ts_slice = def_timeslice; td->td_flags |= TDF_NEEDRESCHED; } return; } if (class == PRI_REALTIME) { /* * Realtime scheduling, do round robin for RR class, FIFO * is not affected. */ if (PRI_NEED_RR(td->td_pri_class) && --ts->ts_slice <= 0) { ts->ts_slice = def_timeslice; td->td_flags |= TDF_NEEDRESCHED; } return; } /* * We skip kernel thread, though it may be classified as TIMESHARE. */ if (class != PRI_TIMESHARE || (p->p_flag & P_KTHREAD) != 0) return; if (--ts->ts_slice <= 0) { td->td_flags |= TDF_NEEDRESCHED; sched_update_runtime(ts, now); sched_commit_runtime(ts); sched_user_prio(td, sched_calc_pri(ts)); ts->ts_slice = sched_timeslice(ts); ts->ts_flags &= ~TSF_FIRST_SLICE; if (ts->ts_flags & TSF_BOUND || td->td_pinned) { if (kseq->ksq_expired_tick == 0) kseq->ksq_expired_tick = tick; } else { if (kseq_global.ksq_expired_tick == 0) kseq_global.ksq_expired_tick = tick; } if (!THREAD_IS_INTERACTIVE(ts) || sched_starving(kseq, tick, ts) || sched_starving(&kseq_global, tick, ts)) { /* The thead becomes cpu hog, schedule it off. */ ts->ts_flags |= TSF_NEXTRQ; if (ts->ts_flags & TSF_BOUND || td->td_pinned) { if (p->p_nice < kseq->ksq_expired_nice) kseq->ksq_expired_nice = p->p_nice; } else { if (p->p_nice < kseq_global.ksq_expired_nice) kseq_global.ksq_expired_nice = p->p_nice; } } } else { /* * Don't allow an interactive thread which has long timeslice * to monopolize CPU, split the long timeslice into small * chunks. This essentially does round-robin between * interactive threads. */ if (THREAD_IS_INTERACTIVE(ts) && sched_timeslice_split(ts)) td->td_flags |= TDF_NEEDRESCHED; } } void sched_clock(struct thread *td) { struct td_sched *ts; mtx_assert(&sched_lock, MA_OWNED); ts = td->td_sched; /* Adjust ticks for pctcpu */ ts->ts_ticks++; ts->ts_ltick = ticks; /* Go up to one second beyond our max and then trim back down */ if (ts->ts_ftick + SCHED_CPU_TICKS + hz < ts->ts_ltick) sched_pctcpu_update(ts); } static int kseq_runnable(struct kseq *kseq) { return (krunq_check(kseq->ksq_curr) || krunq_check(kseq->ksq_next) || krunq_check(&kseq->ksq_idle)); } int sched_runnable(void) { #ifdef SMP return (kseq_runnable(&kseq_global) || kseq_runnable(KSEQ_SELF())); #else return (kseq_runnable(&kseq_global)); #endif } void sched_userret(struct thread *td) { KASSERT((td->td_flags & TDF_BORROWING) == 0, ("thread with borrowed priority returning to userland")); if (td->td_priority != td->td_user_pri) { mtx_lock_spin(&sched_lock); td->td_priority = td->td_user_pri; td->td_base_pri = td->td_user_pri; mtx_unlock_spin(&sched_lock); } } struct td_sched * sched_choose(void) { struct td_sched *ts; struct kseq *kseq; #ifdef SMP struct td_sched *kecpu; mtx_assert(&sched_lock, MA_OWNED); kseq = &kseq_global; ts = kseq_choose(&kseq_global); kecpu = kseq_choose(KSEQ_SELF()); if (ts == NULL || (kecpu != NULL && kecpu->ts_thread->td_priority < ts->ts_thread->td_priority)) { ts = kecpu; kseq = KSEQ_SELF(); } #else kseq = &kseq_global; ts = kseq_choose(kseq); #endif if (ts != NULL) { kseq_runq_rem(kseq, ts); ts->ts_state = TSS_THREAD; ts->ts_flags &= ~TSF_PREEMPTED; ts->ts_timestamp = sched_timestamp(); } return (ts); } #ifdef SMP static int forward_wakeup(int cpunum, cpumask_t me) { cpumask_t map, dontuse; cpumask_t map2; struct pcpu *pc; cpumask_t id, map3; mtx_assert(&sched_lock, MA_OWNED); CTR0(KTR_RUNQ, "forward_wakeup()"); if ((!forward_wakeup_enabled) || (forward_wakeup_use_mask == 0 && forward_wakeup_use_loop == 0)) return (0); if (!smp_started || cold || panicstr) return (0); forward_wakeups_requested++; /* * check the idle mask we received against what we calculated before * in the old version. */ /* * don't bother if we should be doing it ourself.. */ if ((me & idle_cpus_mask) && (cpunum == NOCPU || me == (1 << cpunum))) return (0); dontuse = me | stopped_cpus | hlt_cpus_mask; map3 = 0; if (forward_wakeup_use_loop) { SLIST_FOREACH(pc, &cpuhead, pc_allcpu) { id = pc->pc_cpumask; if ( (id & dontuse) == 0 && pc->pc_curthread == pc->pc_idlethread) { map3 |= id; } } } if (forward_wakeup_use_mask) { map = 0; map = idle_cpus_mask & ~dontuse; /* If they are both on, compare and use loop if different */ if (forward_wakeup_use_loop) { if (map != map3) { printf("map (%02X) != map3 (%02X)\n", map, map3); map = map3; } } } else { map = map3; } /* If we only allow a specific CPU, then mask off all the others */ if (cpunum != NOCPU) { KASSERT((cpunum <= mp_maxcpus),("forward_wakeup: bad cpunum.")); map &= (1 << cpunum); } else { /* Try choose an idle die. */ if (forward_wakeup_use_htt) { map2 = (map & (map >> 1)) & 0x5555; if (map2) { map = map2; } } /* set only one bit */ if (forward_wakeup_use_single) { map = map & ((~map) + 1); } } if (map) { forward_wakeups_delivered++; ipi_selected(map, IPI_AST); return (1); } return (0); } #endif void sched_add(struct thread *td, int flags) { struct kseq *ksq; struct td_sched *ts; struct thread *mytd; int class; int nextrq; int need_resched = 0; #ifdef SMP int cpu; int mycpu; int pinned; struct kseq *myksq; #endif mtx_assert(&sched_lock, MA_OWNED); mytd = curthread; ts = td->td_sched; KASSERT(ts->ts_state != TSS_ONRUNQ, ("sched_add: td_sched %p (%s) already in run queue", ts, ts->ts_proc->p_comm)); KASSERT(ts->ts_proc->p_sflag & PS_INMEM, ("sched_add: process swapped out")); KASSERT(ts->ts_runq == NULL, ("sched_add: KSE %p is still assigned to a run queue", ts)); class = PRI_BASE(td->td_pri_class); #ifdef SMP mycpu = PCPU_GET(cpuid); myksq = KSEQ_CPU(mycpu); ts->ts_wakeup_cpu = mycpu; #endif nextrq = (ts->ts_flags & TSF_NEXTRQ); ts->ts_flags &= ~TSF_NEXTRQ; if (flags & SRQ_PREEMPTED) ts->ts_flags |= TSF_PREEMPTED; ksq = &kseq_global; #ifdef SMP if (td->td_pinned != 0) { cpu = td->td_lastcpu; ksq = KSEQ_CPU(cpu); pinned = 1; } else if ((ts)->ts_flags & TSF_BOUND) { cpu = ts->ts_cpu; ksq = KSEQ_CPU(cpu); pinned = 1; } else { pinned = 0; cpu = NOCPU; } #endif switch (class) { case PRI_ITHD: case PRI_REALTIME: ts->ts_runq = ksq->ksq_curr; break; case PRI_TIMESHARE: if ((td->td_flags & TDF_BORROWING) == 0 && nextrq) ts->ts_runq = ksq->ksq_next; else ts->ts_runq = ksq->ksq_curr; break; case PRI_IDLE: /* * This is for priority prop. */ if (td->td_priority < PRI_MIN_IDLE) ts->ts_runq = ksq->ksq_curr; else ts->ts_runq = &ksq->ksq_idle; break; default: panic("Unknown pri class."); break; } #ifdef SMP if ((ts->ts_runq == kseq_global.ksq_curr || ts->ts_runq == myksq->ksq_curr) && td->td_priority < mytd->td_priority) { #else if (ts->ts_runq == kseq_global.ksq_curr && td->td_priority < mytd->td_priority) { #endif struct krunq *rq; rq = ts->ts_runq; ts->ts_runq = NULL; if ((flags & SRQ_YIELDING) == 0 && maybe_preempt(td)) return; ts->ts_runq = rq; need_resched = TDF_NEEDRESCHED; } ts->ts_state = TSS_ONRUNQ; kseq_runq_add(ksq, ts); kseq_load_add(ksq, ts); #ifdef SMP if (pinned) { if (cpu != mycpu) { struct thread *running = pcpu_find(cpu)->pc_curthread; if (ksq->ksq_curr == ts->ts_runq && running->td_priority < td->td_priority) { if (td->td_priority <= PRI_MAX_ITHD) ipi_selected(1 << cpu, IPI_PREEMPT); else { running->td_flags |= TDF_NEEDRESCHED; ipi_selected(1 << cpu, IPI_AST); } } } else curthread->td_flags |= need_resched; } else { cpumask_t me = 1 << mycpu; cpumask_t idle = idle_cpus_mask & me; int forwarded = 0; if (!idle && ((flags & SRQ_INTR) == 0) && (idle_cpus_mask & ~(hlt_cpus_mask | me))) forwarded = forward_wakeup(cpu, me); if (forwarded == 0) curthread->td_flags |= need_resched; } #else mytd->td_flags |= need_resched; #endif } void sched_rem(struct thread *td) { struct kseq *kseq; struct td_sched *ts; mtx_assert(&sched_lock, MA_OWNED); ts = td->td_sched; KASSERT((ts->ts_state == TSS_ONRUNQ), ("sched_rem: KSE not on run queue")); kseq = ts->ts_kseq; kseq_runq_rem(kseq, ts); kseq_load_rem(kseq, ts); ts->ts_state = TSS_THREAD; } fixpt_t sched_pctcpu(struct thread *td) { fixpt_t pctcpu; struct td_sched *ts; pctcpu = 0; ts = td->td_sched; if (ts == NULL) return (0); mtx_lock_spin(&sched_lock); if (ts->ts_ticks) { int rtick; /* * Don't update more frequently than twice a second. Allowing * this causes the cpu usage to decay away too quickly due to * rounding errors. */ if (ts->ts_ftick + SCHED_CPU_TICKS < ts->ts_ltick || ts->ts_ltick < (ticks - (hz / 2))) sched_pctcpu_update(ts); /* How many rtick per second ? */ rtick = MIN(ts->ts_ticks / SCHED_CPU_TIME, SCHED_CPU_TICKS); pctcpu = (FSCALE * ((FSCALE * rtick)/realstathz)) >> FSHIFT; } ts->ts_proc->p_swtime = ts->ts_ltick - ts->ts_ftick; mtx_unlock_spin(&sched_lock); return (pctcpu); } void sched_bind(struct thread *td, int cpu) { struct td_sched *ts; mtx_assert(&sched_lock, MA_OWNED); ts = td->td_sched; ts->ts_flags |= TSF_BOUND; #ifdef SMP ts->ts_cpu = cpu; if (PCPU_GET(cpuid) == cpu) return; mi_switch(SW_VOL, NULL); #endif } void sched_unbind(struct thread *td) { mtx_assert(&sched_lock, MA_OWNED); td->td_sched->ts_flags &= ~TSF_BOUND; } int sched_is_bound(struct thread *td) { mtx_assert(&sched_lock, MA_OWNED); return (td->td_sched->ts_flags & TSF_BOUND); } int sched_load(void) { return (sched_tdcnt); } void sched_relinquish(struct thread *td) { mtx_lock_spin(&sched_lock); if (sched_is_timeshare(td)) { sched_prio(td, PRI_MAX_TIMESHARE); td->td_sched->ts_flags |= TSF_NEXTRQ; } mi_switch(SW_VOL, NULL); mtx_unlock_spin(&sched_lock); } int sched_sizeof_proc(void) { return (sizeof(struct proc)); } int sched_sizeof_thread(void) { return (sizeof(struct thread) + sizeof(struct td_sched)); } #define KERN_SWITCH_INCLUDE 1 #include "kern/kern_switch.c"