xref: /netbsd-src/sys/kern/kern_synch.c (revision df0caa2637da0538ecdf6b878c4d08e684b43d8f)
1 /*	$NetBSD: kern_synch.c,v 1.149 2005/05/29 22:24:15 christos Exp $	*/
2 
3 /*-
4  * Copyright (c) 1999, 2000, 2004 The NetBSD Foundation, Inc.
5  * All rights reserved.
6  *
7  * This code is derived from software contributed to The NetBSD Foundation
8  * by Jason R. Thorpe of the Numerical Aerospace Simulation Facility,
9  * NASA Ames Research Center.
10  * This code is derived from software contributed to The NetBSD Foundation
11  * by Charles M. Hannum.
12  *
13  * Redistribution and use in source and binary forms, with or without
14  * modification, are permitted provided that the following conditions
15  * are met:
16  * 1. Redistributions of source code must retain the above copyright
17  *    notice, this list of conditions and the following disclaimer.
18  * 2. Redistributions in binary form must reproduce the above copyright
19  *    notice, this list of conditions and the following disclaimer in the
20  *    documentation and/or other materials provided with the distribution.
21  * 3. All advertising materials mentioning features or use of this software
22  *    must display the following acknowledgement:
23  *	This product includes software developed by the NetBSD
24  *	Foundation, Inc. and its contributors.
25  * 4. Neither the name of The NetBSD Foundation nor the names of its
26  *    contributors may be used to endorse or promote products derived
27  *    from this software without specific prior written permission.
28  *
29  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
30  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
31  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
32  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
33  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
34  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
35  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
36  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
37  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
38  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
39  * POSSIBILITY OF SUCH DAMAGE.
40  */
41 
42 /*-
43  * Copyright (c) 1982, 1986, 1990, 1991, 1993
44  *	The Regents of the University of California.  All rights reserved.
45  * (c) UNIX System Laboratories, Inc.
46  * All or some portions of this file are derived from material licensed
47  * to the University of California by American Telephone and Telegraph
48  * Co. or Unix System Laboratories, Inc. and are reproduced herein with
49  * the permission of UNIX System Laboratories, Inc.
50  *
51  * Redistribution and use in source and binary forms, with or without
52  * modification, are permitted provided that the following conditions
53  * are met:
54  * 1. Redistributions of source code must retain the above copyright
55  *    notice, this list of conditions and the following disclaimer.
56  * 2. Redistributions in binary form must reproduce the above copyright
57  *    notice, this list of conditions and the following disclaimer in the
58  *    documentation and/or other materials provided with the distribution.
59  * 3. Neither the name of the University nor the names of its contributors
60  *    may be used to endorse or promote products derived from this software
61  *    without specific prior written permission.
62  *
63  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
64  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
65  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
66  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
67  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
68  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
69  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
70  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
71  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
72  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
73  * SUCH DAMAGE.
74  *
75  *	@(#)kern_synch.c	8.9 (Berkeley) 5/19/95
76  */
77 
78 #include <sys/cdefs.h>
79 __KERNEL_RCSID(0, "$NetBSD: kern_synch.c,v 1.149 2005/05/29 22:24:15 christos Exp $");
80 
81 #include "opt_ddb.h"
82 #include "opt_ktrace.h"
83 #include "opt_kstack.h"
84 #include "opt_lockdebug.h"
85 #include "opt_multiprocessor.h"
86 #include "opt_perfctrs.h"
87 
88 #include <sys/param.h>
89 #include <sys/systm.h>
90 #include <sys/callout.h>
91 #include <sys/proc.h>
92 #include <sys/kernel.h>
93 #include <sys/buf.h>
94 #if defined(PERFCTRS)
95 #include <sys/pmc.h>
96 #endif
97 #include <sys/signalvar.h>
98 #include <sys/resourcevar.h>
99 #include <sys/sched.h>
100 #include <sys/sa.h>
101 #include <sys/savar.h>
102 
103 #include <uvm/uvm_extern.h>
104 
105 #ifdef KTRACE
106 #include <sys/ktrace.h>
107 #endif
108 
109 #include <machine/cpu.h>
110 
111 int	lbolt;			/* once a second sleep address */
112 int	rrticks;		/* number of hardclock ticks per roundrobin() */
113 
114 /*
115  * The global scheduler state.
116  */
117 struct prochd sched_qs[RUNQUE_NQS];	/* run queues */
118 __volatile u_int32_t sched_whichqs;	/* bitmap of non-empty queues */
119 struct slpque sched_slpque[SLPQUE_TABLESIZE]; /* sleep queues */
120 
121 struct simplelock sched_lock = SIMPLELOCK_INITIALIZER;
122 
123 void schedcpu(void *);
124 void updatepri(struct lwp *);
125 void endtsleep(void *);
126 
127 __inline void sa_awaken(struct lwp *);
128 __inline void awaken(struct lwp *);
129 
130 struct callout schedcpu_ch = CALLOUT_INITIALIZER_SETFUNC(schedcpu, NULL);
131 
132 
133 
134 /*
135  * Force switch among equal priority processes every 100ms.
136  * Called from hardclock every hz/10 == rrticks hardclock ticks.
137  */
138 /* ARGSUSED */
139 void
140 roundrobin(struct cpu_info *ci)
141 {
142 	struct schedstate_percpu *spc = &ci->ci_schedstate;
143 
144 	spc->spc_rrticks = rrticks;
145 
146 	if (curlwp != NULL) {
147 		if (spc->spc_flags & SPCF_SEENRR) {
148 			/*
149 			 * The process has already been through a roundrobin
150 			 * without switching and may be hogging the CPU.
151 			 * Indicate that the process should yield.
152 			 */
153 			spc->spc_flags |= SPCF_SHOULDYIELD;
154 		} else
155 			spc->spc_flags |= SPCF_SEENRR;
156 	}
157 	need_resched(curcpu());
158 }
159 
160 /*
161  * Constants for digital decay and forget:
162  *	90% of (p_estcpu) usage in 5 * loadav time
163  *	95% of (p_pctcpu) usage in 60 seconds (load insensitive)
164  *          Note that, as ps(1) mentions, this can let percentages
165  *          total over 100% (I've seen 137.9% for 3 processes).
166  *
167  * Note that hardclock updates p_estcpu and p_cpticks independently.
168  *
169  * We wish to decay away 90% of p_estcpu in (5 * loadavg) seconds.
170  * That is, the system wants to compute a value of decay such
171  * that the following for loop:
172  * 	for (i = 0; i < (5 * loadavg); i++)
173  * 		p_estcpu *= decay;
174  * will compute
175  * 	p_estcpu *= 0.1;
176  * for all values of loadavg:
177  *
178  * Mathematically this loop can be expressed by saying:
179  * 	decay ** (5 * loadavg) ~= .1
180  *
181  * The system computes decay as:
182  * 	decay = (2 * loadavg) / (2 * loadavg + 1)
183  *
184  * We wish to prove that the system's computation of decay
185  * will always fulfill the equation:
186  * 	decay ** (5 * loadavg) ~= .1
187  *
188  * If we compute b as:
189  * 	b = 2 * loadavg
190  * then
191  * 	decay = b / (b + 1)
192  *
193  * We now need to prove two things:
194  *	1) Given factor ** (5 * loadavg) ~= .1, prove factor == b/(b+1)
195  *	2) Given b/(b+1) ** power ~= .1, prove power == (5 * loadavg)
196  *
197  * Facts:
198  *         For x close to zero, exp(x) =~ 1 + x, since
199  *              exp(x) = 0! + x**1/1! + x**2/2! + ... .
200  *              therefore exp(-1/b) =~ 1 - (1/b) = (b-1)/b.
201  *         For x close to zero, ln(1+x) =~ x, since
202  *              ln(1+x) = x - x**2/2 + x**3/3 - ...     -1 < x < 1
203  *              therefore ln(b/(b+1)) = ln(1 - 1/(b+1)) =~ -1/(b+1).
204  *         ln(.1) =~ -2.30
205  *
206  * Proof of (1):
207  *    Solve (factor)**(power) =~ .1 given power (5*loadav):
208  *	solving for factor,
209  *      ln(factor) =~ (-2.30/5*loadav), or
210  *      factor =~ exp(-1/((5/2.30)*loadav)) =~ exp(-1/(2*loadav)) =
211  *          exp(-1/b) =~ (b-1)/b =~ b/(b+1).                    QED
212  *
213  * Proof of (2):
214  *    Solve (factor)**(power) =~ .1 given factor == (b/(b+1)):
215  *	solving for power,
216  *      power*ln(b/(b+1)) =~ -2.30, or
217  *      power =~ 2.3 * (b + 1) = 4.6*loadav + 2.3 =~ 5*loadav.  QED
218  *
219  * Actual power values for the implemented algorithm are as follows:
220  *      loadav: 1       2       3       4
221  *      power:  5.68    10.32   14.94   19.55
222  */
223 
224 /* calculations for digital decay to forget 90% of usage in 5*loadav sec */
225 #define	loadfactor(loadav)	(2 * (loadav))
226 #define	decay_cpu(loadfac, cpu)	(((loadfac) * (cpu)) / ((loadfac) + FSCALE))
227 
228 /* decay 95% of `p_pctcpu' in 60 seconds; see CCPU_SHIFT before changing */
229 fixpt_t	ccpu = 0.95122942450071400909 * FSCALE;		/* exp(-1/20) */
230 
231 /*
232  * If `ccpu' is not equal to `exp(-1/20)' and you still want to use the
233  * faster/more-accurate formula, you'll have to estimate CCPU_SHIFT below
234  * and possibly adjust FSHIFT in "param.h" so that (FSHIFT >= CCPU_SHIFT).
235  *
236  * To estimate CCPU_SHIFT for exp(-1/20), the following formula was used:
237  *	1 - exp(-1/20) ~= 0.0487 ~= 0.0488 == 1 (fixed pt, *11* bits).
238  *
239  * If you dont want to bother with the faster/more-accurate formula, you
240  * can set CCPU_SHIFT to (FSHIFT + 1) which will use a slower/less-accurate
241  * (more general) method of calculating the %age of CPU used by a process.
242  */
243 #define	CCPU_SHIFT	11
244 
245 /*
246  * Recompute process priorities, every hz ticks.
247  */
248 /* ARGSUSED */
249 void
250 schedcpu(void *arg)
251 {
252 	fixpt_t loadfac = loadfactor(averunnable.ldavg[0]);
253 	struct lwp *l;
254 	struct proc *p;
255 	int s, minslp;
256 	unsigned int newcpu;
257 	int clkhz;
258 
259 	proclist_lock_read();
260 	PROCLIST_FOREACH(p, &allproc) {
261 		/*
262 		 * Increment time in/out of memory and sleep time
263 		 * (if sleeping).  We ignore overflow; with 16-bit int's
264 		 * (remember them?) overflow takes 45 days.
265 		 */
266 		minslp = 2;
267 		LIST_FOREACH(l, &p->p_lwps, l_sibling) {
268 			l->l_swtime++;
269 			if (l->l_stat == LSSLEEP || l->l_stat == LSSTOP ||
270 			    l->l_stat == LSSUSPENDED) {
271 				l->l_slptime++;
272 				minslp = min(minslp, l->l_slptime);
273 			} else
274 				minslp = 0;
275 		}
276 		p->p_pctcpu = (p->p_pctcpu * ccpu) >> FSHIFT;
277 		/*
278 		 * If the process has slept the entire second,
279 		 * stop recalculating its priority until it wakes up.
280 		 */
281 		if (minslp > 1)
282 			continue;
283 		s = splstatclock();	/* prevent state changes */
284 		/*
285 		 * p_pctcpu is only for ps.
286 		 */
287 		clkhz = stathz != 0 ? stathz : hz;
288 #if	(FSHIFT >= CCPU_SHIFT)
289 		p->p_pctcpu += (clkhz == 100)?
290 			((fixpt_t) p->p_cpticks) << (FSHIFT - CCPU_SHIFT):
291                 	100 * (((fixpt_t) p->p_cpticks)
292 				<< (FSHIFT - CCPU_SHIFT)) / clkhz;
293 #else
294 		p->p_pctcpu += ((FSCALE - ccpu) *
295 			(p->p_cpticks * FSCALE / clkhz)) >> FSHIFT;
296 #endif
297 		p->p_cpticks = 0;
298 		newcpu = (u_int)decay_cpu(loadfac, p->p_estcpu);
299 		p->p_estcpu = newcpu;
300 		splx(s);	/* Done with the process CPU ticks update */
301 		SCHED_LOCK(s);
302 		LIST_FOREACH(l, &p->p_lwps, l_sibling) {
303 			if (l->l_slptime > 1)
304 				continue;
305 			resetpriority(l);
306 			if (l->l_priority >= PUSER) {
307 				if (l->l_stat == LSRUN &&
308 				    (l->l_flag & L_INMEM) &&
309 				    (l->l_priority / PPQ) != (l->l_usrpri / PPQ)) {
310 					remrunqueue(l);
311 					l->l_priority = l->l_usrpri;
312 					setrunqueue(l);
313 				} else
314 					l->l_priority = l->l_usrpri;
315 			}
316 		}
317 		SCHED_UNLOCK(s);
318 	}
319 	proclist_unlock_read();
320 	uvm_meter();
321 	wakeup((caddr_t)&lbolt);
322 	callout_schedule(&schedcpu_ch, hz);
323 }
324 
325 /*
326  * Recalculate the priority of a process after it has slept for a while.
327  * For all load averages >= 1 and max p_estcpu of 255, sleeping for at
328  * least six times the loadfactor will decay p_estcpu to zero.
329  */
330 void
331 updatepri(struct lwp *l)
332 {
333 	struct proc *p = l->l_proc;
334 	unsigned int newcpu;
335 	fixpt_t loadfac;
336 
337 	SCHED_ASSERT_LOCKED();
338 
339 	newcpu = p->p_estcpu;
340 	loadfac = loadfactor(averunnable.ldavg[0]);
341 
342 	if (l->l_slptime > 5 * loadfac)
343 		p->p_estcpu = 0; /* XXX NJWLWP */
344 	else {
345 		l->l_slptime--;	/* the first time was done in schedcpu */
346 		while (newcpu && --l->l_slptime)
347 			newcpu = (int) decay_cpu(loadfac, newcpu);
348 		p->p_estcpu = newcpu;
349 	}
350 	resetpriority(l);
351 }
352 
353 /*
354  * During autoconfiguration or after a panic, a sleep will simply
355  * lower the priority briefly to allow interrupts, then return.
356  * The priority to be used (safepri) is machine-dependent, thus this
357  * value is initialized and maintained in the machine-dependent layers.
358  * This priority will typically be 0, or the lowest priority
359  * that is safe for use on the interrupt stack; it can be made
360  * higher to block network software interrupts after panics.
361  */
362 int safepri;
363 
364 /*
365  * General sleep call.  Suspends the current process until a wakeup is
366  * performed on the specified identifier.  The process will then be made
367  * runnable with the specified priority.  Sleeps at most timo/hz seconds
368  * (0 means no timeout).  If pri includes PCATCH flag, signals are checked
369  * before and after sleeping, else signals are not checked.  Returns 0 if
370  * awakened, EWOULDBLOCK if the timeout expires.  If PCATCH is set and a
371  * signal needs to be delivered, ERESTART is returned if the current system
372  * call should be restarted if possible, and EINTR is returned if the system
373  * call should be interrupted by the signal (return EINTR).
374  *
375  * The interlock is held until the scheduler_slock is acquired.  The
376  * interlock will be locked before returning back to the caller
377  * unless the PNORELOCK flag is specified, in which case the
378  * interlock will always be unlocked upon return.
379  */
380 int
381 ltsleep(__volatile const void *ident, int priority, const char *wmesg, int timo,
382     __volatile struct simplelock *interlock)
383 {
384 	struct lwp *l = curlwp;
385 	struct proc *p = l ? l->l_proc : NULL;
386 	struct slpque *qp;
387 	int sig, s;
388 	int catch = priority & PCATCH;
389 	int relock = (priority & PNORELOCK) == 0;
390 	int exiterr = (priority & PNOEXITERR) == 0;
391 
392 	/*
393 	 * XXXSMP
394 	 * This is probably bogus.  Figure out what the right
395 	 * thing to do here really is.
396 	 * Note that not sleeping if ltsleep is called with curlwp == NULL
397 	 * in the shutdown case is disgusting but partly necessary given
398 	 * how shutdown (barely) works.
399 	 */
400 	if (cold || (doing_shutdown && (panicstr || (l == NULL)))) {
401 		/*
402 		 * After a panic, or during autoconfiguration,
403 		 * just give interrupts a chance, then just return;
404 		 * don't run any other procs or panic below,
405 		 * in case this is the idle process and already asleep.
406 		 */
407 		s = splhigh();
408 		splx(safepri);
409 		splx(s);
410 		if (interlock != NULL && relock == 0)
411 			simple_unlock(interlock);
412 		return (0);
413 	}
414 
415 	KASSERT(p != NULL);
416 	LOCK_ASSERT(interlock == NULL || simple_lock_held(interlock));
417 
418 #ifdef KTRACE
419 	if (KTRPOINT(p, KTR_CSW))
420 		ktrcsw(p, 1, 0);
421 #endif
422 
423 	SCHED_LOCK(s);
424 
425 #ifdef DIAGNOSTIC
426 	if (ident == NULL)
427 		panic("ltsleep: ident == NULL");
428 	if (l->l_stat != LSONPROC)
429 		panic("ltsleep: l_stat %d != LSONPROC", l->l_stat);
430 	if (l->l_back != NULL)
431 		panic("ltsleep: p_back != NULL");
432 #endif
433 
434 	l->l_wchan = ident;
435 	l->l_wmesg = wmesg;
436 	l->l_slptime = 0;
437 	l->l_priority = priority & PRIMASK;
438 
439 	qp = SLPQUE(ident);
440 	if (qp->sq_head == 0)
441 		qp->sq_head = l;
442 	else {
443 		*qp->sq_tailp = l;
444 	}
445 	*(qp->sq_tailp = &l->l_forw) = 0;
446 
447 	if (timo)
448 		callout_reset(&l->l_tsleep_ch, timo, endtsleep, l);
449 
450 	/*
451 	 * We can now release the interlock; the scheduler_slock
452 	 * is held, so a thread can't get in to do wakeup() before
453 	 * we do the switch.
454 	 *
455 	 * XXX We leave the code block here, after inserting ourselves
456 	 * on the sleep queue, because we might want a more clever
457 	 * data structure for the sleep queues at some point.
458 	 */
459 	if (interlock != NULL)
460 		simple_unlock(interlock);
461 
462 	/*
463 	 * We put ourselves on the sleep queue and start our timeout
464 	 * before calling CURSIG, as we could stop there, and a wakeup
465 	 * or a SIGCONT (or both) could occur while we were stopped.
466 	 * A SIGCONT would cause us to be marked as SSLEEP
467 	 * without resuming us, thus we must be ready for sleep
468 	 * when CURSIG is called.  If the wakeup happens while we're
469 	 * stopped, p->p_wchan will be 0 upon return from CURSIG.
470 	 */
471 	if (catch) {
472 		l->l_flag |= L_SINTR;
473 		if (((sig = CURSIG(l)) != 0) ||
474 		    ((p->p_flag & P_WEXIT) && p->p_nlwps > 1)) {
475 			if (l->l_wchan != NULL)
476 				unsleep(l);
477 			l->l_stat = LSONPROC;
478 			SCHED_UNLOCK(s);
479 			goto resume;
480 		}
481 		if (l->l_wchan == NULL) {
482 			catch = 0;
483 			SCHED_UNLOCK(s);
484 			goto resume;
485 		}
486 	} else
487 		sig = 0;
488 	l->l_stat = LSSLEEP;
489 	p->p_nrlwps--;
490 	p->p_stats->p_ru.ru_nvcsw++;
491 	SCHED_ASSERT_LOCKED();
492 	if (l->l_flag & L_SA)
493 		sa_switch(l, SA_UPCALL_BLOCKED);
494 	else
495 		mi_switch(l, NULL);
496 
497 #if	defined(DDB) && !defined(GPROF)
498 	/* handy breakpoint location after process "wakes" */
499 	__asm(".globl bpendtsleep\nbpendtsleep:");
500 #endif
501 	/*
502 	 * p->p_nrlwps is incremented by whoever made us runnable again,
503 	 * either setrunnable() or awaken().
504 	 */
505 
506 	SCHED_ASSERT_UNLOCKED();
507 	splx(s);
508 
509  resume:
510 	KDASSERT(l->l_cpu != NULL);
511 	KDASSERT(l->l_cpu == curcpu());
512 	l->l_cpu->ci_schedstate.spc_curpriority = l->l_usrpri;
513 
514 	l->l_flag &= ~L_SINTR;
515 	if (l->l_flag & L_TIMEOUT) {
516 		l->l_flag &= ~(L_TIMEOUT|L_CANCELLED);
517 		if (sig == 0) {
518 #ifdef KTRACE
519 			if (KTRPOINT(p, KTR_CSW))
520 				ktrcsw(p, 0, 0);
521 #endif
522 			if (relock && interlock != NULL)
523 				simple_lock(interlock);
524 			return (EWOULDBLOCK);
525 		}
526 	} else if (timo)
527 		callout_stop(&l->l_tsleep_ch);
528 
529 	if (catch) {
530 		const int cancelled = l->l_flag & L_CANCELLED;
531 		l->l_flag &= ~L_CANCELLED;
532 		if (sig != 0 || (sig = CURSIG(l)) != 0 || cancelled) {
533 #ifdef KTRACE
534 			if (KTRPOINT(p, KTR_CSW))
535 				ktrcsw(p, 0, 0);
536 #endif
537 			if (relock && interlock != NULL)
538 				simple_lock(interlock);
539 			/*
540 			 * If this sleep was canceled, don't let the syscall
541 			 * restart.
542 			 */
543 			if (cancelled ||
544 			    (SIGACTION(p, sig).sa_flags & SA_RESTART) == 0)
545 				return (EINTR);
546 			return (ERESTART);
547 		}
548 	}
549 
550 #ifdef KTRACE
551 	if (KTRPOINT(p, KTR_CSW))
552 		ktrcsw(p, 0, 0);
553 #endif
554 	if (relock && interlock != NULL)
555 		simple_lock(interlock);
556 
557 	/* XXXNJW this is very much a kluge.
558 	 * revisit. a better way of preventing looping/hanging syscalls like
559 	 * wait4() and _lwp_wait() from wedging an exiting process
560 	 * would be preferred.
561 	 */
562 	if (catch && ((p->p_flag & P_WEXIT) && p->p_nlwps > 1 && exiterr))
563 		return (EINTR);
564 	return (0);
565 }
566 
567 /*
568  * Implement timeout for tsleep.
569  * If process hasn't been awakened (wchan non-zero),
570  * set timeout flag and undo the sleep.  If proc
571  * is stopped, just unsleep so it will remain stopped.
572  */
573 void
574 endtsleep(void *arg)
575 {
576 	struct lwp *l;
577 	int s;
578 
579 	l = (struct lwp *)arg;
580 	SCHED_LOCK(s);
581 	if (l->l_wchan) {
582 		if (l->l_stat == LSSLEEP)
583 			setrunnable(l);
584 		else
585 			unsleep(l);
586 		l->l_flag |= L_TIMEOUT;
587 	}
588 	SCHED_UNLOCK(s);
589 }
590 
591 /*
592  * Remove a process from its wait queue
593  */
594 void
595 unsleep(struct lwp *l)
596 {
597 	struct slpque *qp;
598 	struct lwp **hp;
599 
600 	SCHED_ASSERT_LOCKED();
601 
602 	if (l->l_wchan) {
603 		hp = &(qp = SLPQUE(l->l_wchan))->sq_head;
604 		while (*hp != l)
605 			hp = &(*hp)->l_forw;
606 		*hp = l->l_forw;
607 		if (qp->sq_tailp == &l->l_forw)
608 			qp->sq_tailp = hp;
609 		l->l_wchan = 0;
610 	}
611 }
612 
613 __inline void
614 sa_awaken(struct lwp *l)
615 {
616 
617 	SCHED_ASSERT_LOCKED();
618 
619 	if (l == l->l_savp->savp_lwp && l->l_flag & L_SA_YIELD)
620 		l->l_flag &= ~L_SA_IDLE;
621 }
622 
623 /*
624  * Optimized-for-wakeup() version of setrunnable().
625  */
626 __inline void
627 awaken(struct lwp *l)
628 {
629 
630 	SCHED_ASSERT_LOCKED();
631 
632 	if (l->l_proc->p_sa)
633 		sa_awaken(l);
634 
635 	if (l->l_slptime > 1)
636 		updatepri(l);
637 	l->l_slptime = 0;
638 	l->l_stat = LSRUN;
639 	l->l_proc->p_nrlwps++;
640 	/*
641 	 * Since curpriority is a user priority, p->p_priority
642 	 * is always better than curpriority on the last CPU on
643 	 * which it ran.
644 	 *
645 	 * XXXSMP See affinity comment in resched_proc().
646 	 */
647 	if (l->l_flag & L_INMEM) {
648 		setrunqueue(l);
649 		KASSERT(l->l_cpu != NULL);
650 		need_resched(l->l_cpu);
651 	} else
652 		sched_wakeup(&proc0);
653 }
654 
655 #if defined(MULTIPROCESSOR) || defined(LOCKDEBUG)
656 void
657 sched_unlock_idle(void)
658 {
659 
660 	simple_unlock(&sched_lock);
661 }
662 
663 void
664 sched_lock_idle(void)
665 {
666 
667 	simple_lock(&sched_lock);
668 }
669 #endif /* MULTIPROCESSOR || LOCKDEBUG */
670 
671 /*
672  * Make all processes sleeping on the specified identifier runnable.
673  */
674 
675 void
676 wakeup(__volatile const void *ident)
677 {
678 	int s;
679 
680 	SCHED_ASSERT_UNLOCKED();
681 
682 	SCHED_LOCK(s);
683 	sched_wakeup(ident);
684 	SCHED_UNLOCK(s);
685 }
686 
687 void
688 sched_wakeup(__volatile const void *ident)
689 {
690 	struct slpque *qp;
691 	struct lwp *l, **q;
692 
693 	SCHED_ASSERT_LOCKED();
694 
695 	qp = SLPQUE(ident);
696  restart:
697 	for (q = &qp->sq_head; (l = *q) != NULL; ) {
698 #ifdef DIAGNOSTIC
699 		if (l->l_back || (l->l_stat != LSSLEEP &&
700 		    l->l_stat != LSSTOP && l->l_stat != LSSUSPENDED))
701 			panic("wakeup");
702 #endif
703 		if (l->l_wchan == ident) {
704 			l->l_wchan = 0;
705 			*q = l->l_forw;
706 			if (qp->sq_tailp == &l->l_forw)
707 				qp->sq_tailp = q;
708 			if (l->l_stat == LSSLEEP) {
709 				awaken(l);
710 				goto restart;
711 			}
712 		} else
713 			q = &l->l_forw;
714 	}
715 }
716 
717 /*
718  * Make the highest priority process first in line on the specified
719  * identifier runnable.
720  */
721 void
722 wakeup_one(__volatile const void *ident)
723 {
724 	struct slpque *qp;
725 	struct lwp *l, **q;
726 	struct lwp *best_sleepp, **best_sleepq;
727 	struct lwp *best_stopp, **best_stopq;
728 	int s;
729 
730 	best_sleepp = best_stopp = NULL;
731 	best_sleepq = best_stopq = NULL;
732 
733 	SCHED_LOCK(s);
734 
735 	qp = SLPQUE(ident);
736 
737 	for (q = &qp->sq_head; (l = *q) != NULL; q = &l->l_forw) {
738 #ifdef DIAGNOSTIC
739 		if (l->l_back || (l->l_stat != LSSLEEP &&
740 		    l->l_stat != LSSTOP && l->l_stat != LSSUSPENDED))
741 			panic("wakeup_one");
742 #endif
743 		if (l->l_wchan == ident) {
744 			if (l->l_stat == LSSLEEP) {
745 				if (best_sleepp == NULL ||
746 				    l->l_priority < best_sleepp->l_priority) {
747 					best_sleepp = l;
748 					best_sleepq = q;
749 				}
750 			} else {
751 				if (best_stopp == NULL ||
752 				    l->l_priority < best_stopp->l_priority) {
753 				    	best_stopp = l;
754 					best_stopq = q;
755 				}
756 			}
757 		}
758 	}
759 
760 	/*
761 	 * Consider any SSLEEP process higher than the highest priority SSTOP
762 	 * process.
763 	 */
764 	if (best_sleepp != NULL) {
765 		l = best_sleepp;
766 		q = best_sleepq;
767 	} else {
768 		l = best_stopp;
769 		q = best_stopq;
770 	}
771 
772 	if (l != NULL) {
773 		l->l_wchan = NULL;
774 		*q = l->l_forw;
775 		if (qp->sq_tailp == &l->l_forw)
776 			qp->sq_tailp = q;
777 		if (l->l_stat == LSSLEEP)
778 			awaken(l);
779 	}
780 	SCHED_UNLOCK(s);
781 }
782 
783 /*
784  * General yield call.  Puts the current process back on its run queue and
785  * performs a voluntary context switch.  Should only be called when the
786  * current process explicitly requests it (eg sched_yield(2) in compat code).
787  */
788 void
789 yield(void)
790 {
791 	struct lwp *l = curlwp;
792 	int s;
793 
794 	SCHED_LOCK(s);
795 	l->l_priority = l->l_usrpri;
796 	l->l_stat = LSRUN;
797 	setrunqueue(l);
798 	l->l_proc->p_stats->p_ru.ru_nvcsw++;
799 	mi_switch(l, NULL);
800 	SCHED_ASSERT_UNLOCKED();
801 	splx(s);
802 }
803 
804 /*
805  * General preemption call.  Puts the current process back on its run queue
806  * and performs an involuntary context switch.  If a process is supplied,
807  * we switch to that process.  Otherwise, we use the normal process selection
808  * criteria.
809  */
810 
811 void
812 preempt(int more)
813 {
814 	struct lwp *l = curlwp;
815 	int r, s;
816 
817 	SCHED_LOCK(s);
818 	l->l_priority = l->l_usrpri;
819 	l->l_stat = LSRUN;
820 	setrunqueue(l);
821 	l->l_proc->p_stats->p_ru.ru_nivcsw++;
822 	r = mi_switch(l, NULL);
823 	SCHED_ASSERT_UNLOCKED();
824 	splx(s);
825 	if ((l->l_flag & L_SA) != 0 && r != 0 && more == 0)
826 		sa_preempt(l);
827 }
828 
829 /*
830  * The machine independent parts of context switch.
831  * Must be called at splsched() (no higher!) and with
832  * the sched_lock held.
833  * Switch to "new" if non-NULL, otherwise let cpu_switch choose
834  * the next lwp.
835  *
836  * Returns 1 if another process was actually run.
837  */
838 int
839 mi_switch(struct lwp *l, struct lwp *newl)
840 {
841 	struct schedstate_percpu *spc;
842 	struct rlimit *rlim;
843 	long s, u;
844 	struct timeval tv;
845 	int hold_count;
846 	struct proc *p = l->l_proc;
847 	int retval;
848 
849 	SCHED_ASSERT_LOCKED();
850 
851 	/*
852 	 * Release the kernel_lock, as we are about to yield the CPU.
853 	 * The scheduler lock is still held until cpu_switch()
854 	 * selects a new process and removes it from the run queue.
855 	 */
856 	hold_count = KERNEL_LOCK_RELEASE_ALL();
857 
858 	KDASSERT(l->l_cpu != NULL);
859 	KDASSERT(l->l_cpu == curcpu());
860 
861 	spc = &l->l_cpu->ci_schedstate;
862 
863 #if defined(LOCKDEBUG) || defined(DIAGNOSTIC)
864 	spinlock_switchcheck();
865 #endif
866 #ifdef LOCKDEBUG
867 	simple_lock_switchcheck();
868 #endif
869 
870 	/*
871 	 * Compute the amount of time during which the current
872 	 * process was running.
873 	 */
874 	microtime(&tv);
875 	u = p->p_rtime.tv_usec +
876 	    (tv.tv_usec - spc->spc_runtime.tv_usec);
877 	s = p->p_rtime.tv_sec + (tv.tv_sec - spc->spc_runtime.tv_sec);
878 	if (u < 0) {
879 		u += 1000000;
880 		s--;
881 	} else if (u >= 1000000) {
882 		u -= 1000000;
883 		s++;
884 	}
885 	p->p_rtime.tv_usec = u;
886 	p->p_rtime.tv_sec = s;
887 
888 	/*
889 	 * Check if the process exceeds its CPU resource allocation.
890 	 * If over max, kill it.  In any case, if it has run for more
891 	 * than 10 minutes, reduce priority to give others a chance.
892 	 */
893 	rlim = &p->p_rlimit[RLIMIT_CPU];
894 	if (s >= rlim->rlim_cur) {
895 		/*
896 		 * XXXSMP: we're inside the scheduler lock perimeter;
897 		 * use sched_psignal.
898 		 */
899 		if (s >= rlim->rlim_max)
900 			sched_psignal(p, SIGKILL);
901 		else {
902 			sched_psignal(p, SIGXCPU);
903 			if (rlim->rlim_cur < rlim->rlim_max)
904 				rlim->rlim_cur += 5;
905 		}
906 	}
907 	if (autonicetime && s > autonicetime && p->p_ucred->cr_uid &&
908 	    p->p_nice == NZERO) {
909 		p->p_nice = autoniceval + NZERO;
910 		resetpriority(l);
911 	}
912 
913 	/*
914 	 * Process is about to yield the CPU; clear the appropriate
915 	 * scheduling flags.
916 	 */
917 	spc->spc_flags &= ~SPCF_SWITCHCLEAR;
918 
919 #ifdef KSTACK_CHECK_MAGIC
920 	kstack_check_magic(l);
921 #endif
922 
923 	/*
924 	 * If we are using h/w performance counters, save context.
925 	 */
926 #if PERFCTRS
927 	if (PMC_ENABLED(p))
928 		pmc_save_context(p);
929 #endif
930 
931 	/*
932 	 * Switch to the new current process.  When we
933 	 * run again, we'll return back here.
934 	 */
935 	uvmexp.swtch++;
936 	if (newl == NULL) {
937 		retval = cpu_switch(l, NULL);
938 	} else {
939 		remrunqueue(newl);
940 		cpu_switchto(l, newl);
941 		retval = 0;
942 	}
943 
944 	/*
945 	 * If we are using h/w performance counters, restore context.
946 	 */
947 #if PERFCTRS
948 	if (PMC_ENABLED(p))
949 		pmc_restore_context(p);
950 #endif
951 
952 	/*
953 	 * Make sure that MD code released the scheduler lock before
954 	 * resuming us.
955 	 */
956 	SCHED_ASSERT_UNLOCKED();
957 
958 	/*
959 	 * We're running again; record our new start time.  We might
960 	 * be running on a new CPU now, so don't use the cache'd
961 	 * schedstate_percpu pointer.
962 	 */
963 	KDASSERT(l->l_cpu != NULL);
964 	KDASSERT(l->l_cpu == curcpu());
965 	microtime(&l->l_cpu->ci_schedstate.spc_runtime);
966 
967 	/*
968 	 * Reacquire the kernel_lock now.  We do this after we've
969 	 * released the scheduler lock to avoid deadlock, and before
970 	 * we reacquire the interlock.
971 	 */
972 	KERNEL_LOCK_ACQUIRE_COUNT(hold_count);
973 
974 	return retval;
975 }
976 
977 /*
978  * Initialize the (doubly-linked) run queues
979  * to be empty.
980  */
981 void
982 rqinit()
983 {
984 	int i;
985 
986 	for (i = 0; i < RUNQUE_NQS; i++)
987 		sched_qs[i].ph_link = sched_qs[i].ph_rlink =
988 		    (struct lwp *)&sched_qs[i];
989 }
990 
991 static __inline void
992 resched_proc(struct lwp *l, u_char pri)
993 {
994 	struct cpu_info *ci;
995 
996 	/*
997 	 * XXXSMP
998 	 * Since l->l_cpu persists across a context switch,
999 	 * this gives us *very weak* processor affinity, in
1000 	 * that we notify the CPU on which the process last
1001 	 * ran that it should try to switch.
1002 	 *
1003 	 * This does not guarantee that the process will run on
1004 	 * that processor next, because another processor might
1005 	 * grab it the next time it performs a context switch.
1006 	 *
1007 	 * This also does not handle the case where its last
1008 	 * CPU is running a higher-priority process, but every
1009 	 * other CPU is running a lower-priority process.  There
1010 	 * are ways to handle this situation, but they're not
1011 	 * currently very pretty, and we also need to weigh the
1012 	 * cost of moving a process from one CPU to another.
1013 	 *
1014 	 * XXXSMP
1015 	 * There is also the issue of locking the other CPU's
1016 	 * sched state, which we currently do not do.
1017 	 */
1018 	ci = (l->l_cpu != NULL) ? l->l_cpu : curcpu();
1019 	if (pri < ci->ci_schedstate.spc_curpriority)
1020 		need_resched(ci);
1021 }
1022 
1023 /*
1024  * Change process state to be runnable,
1025  * placing it on the run queue if it is in memory,
1026  * and awakening the swapper if it isn't in memory.
1027  */
1028 void
1029 setrunnable(struct lwp *l)
1030 {
1031 	struct proc *p = l->l_proc;
1032 
1033 	SCHED_ASSERT_LOCKED();
1034 
1035 	switch (l->l_stat) {
1036 	case 0:
1037 	case LSRUN:
1038 	case LSONPROC:
1039 	case LSZOMB:
1040 	case LSDEAD:
1041 	default:
1042 		panic("setrunnable: lwp %p state was %d", l, l->l_stat);
1043 	case LSSTOP:
1044 		/*
1045 		 * If we're being traced (possibly because someone attached us
1046 		 * while we were stopped), check for a signal from the debugger.
1047 		 */
1048 		if ((p->p_flag & P_TRACED) != 0 && p->p_xstat != 0) {
1049 			sigaddset(&p->p_sigctx.ps_siglist, p->p_xstat);
1050 			CHECKSIGS(p);
1051 		}
1052 	case LSSLEEP:
1053 		unsleep(l);		/* e.g. when sending signals */
1054 		break;
1055 
1056 	case LSIDL:
1057 		break;
1058 	case LSSUSPENDED:
1059 		break;
1060 	}
1061 
1062 	if (l->l_proc->p_sa)
1063 		sa_awaken(l);
1064 
1065 	l->l_stat = LSRUN;
1066 	p->p_nrlwps++;
1067 
1068 	if (l->l_flag & L_INMEM)
1069 		setrunqueue(l);
1070 
1071 	if (l->l_slptime > 1)
1072 		updatepri(l);
1073 	l->l_slptime = 0;
1074 	if ((l->l_flag & L_INMEM) == 0)
1075 		sched_wakeup((caddr_t)&proc0);
1076 	else
1077 		resched_proc(l, l->l_priority);
1078 }
1079 
1080 /*
1081  * Compute the priority of a process when running in user mode.
1082  * Arrange to reschedule if the resulting priority is better
1083  * than that of the current process.
1084  */
1085 void
1086 resetpriority(struct lwp *l)
1087 {
1088 	unsigned int newpriority;
1089 	struct proc *p = l->l_proc;
1090 
1091 	SCHED_ASSERT_LOCKED();
1092 
1093 	newpriority = PUSER + p->p_estcpu +
1094 			NICE_WEIGHT * (p->p_nice - NZERO);
1095 	newpriority = min(newpriority, MAXPRI);
1096 	l->l_usrpri = newpriority;
1097 	resched_proc(l, l->l_usrpri);
1098 }
1099 
1100 /*
1101  * Recompute priority for all LWPs in a process.
1102  */
1103 void
1104 resetprocpriority(struct proc *p)
1105 {
1106 	struct lwp *l;
1107 
1108 	LIST_FOREACH(l, &p->p_lwps, l_sibling)
1109 	    resetpriority(l);
1110 }
1111 
1112 /*
1113  * We adjust the priority of the current process.  The priority of a process
1114  * gets worse as it accumulates CPU time.  The CPU usage estimator (p_estcpu)
1115  * is increased here.  The formula for computing priorities (in kern_synch.c)
1116  * will compute a different value each time p_estcpu increases. This can
1117  * cause a switch, but unless the priority crosses a PPQ boundary the actual
1118  * queue will not change.  The CPU usage estimator ramps up quite quickly
1119  * when the process is running (linearly), and decays away exponentially, at
1120  * a rate which is proportionally slower when the system is busy.  The basic
1121  * principle is that the system will 90% forget that the process used a lot
1122  * of CPU time in 5 * loadav seconds.  This causes the system to favor
1123  * processes which haven't run much recently, and to round-robin among other
1124  * processes.
1125  */
1126 
1127 void
1128 schedclock(struct lwp *l)
1129 {
1130 	struct proc *p = l->l_proc;
1131 	int s;
1132 
1133 	p->p_estcpu = ESTCPULIM(p->p_estcpu + 1);
1134 	SCHED_LOCK(s);
1135 	resetpriority(l);
1136 	SCHED_UNLOCK(s);
1137 
1138 	if (l->l_priority >= PUSER)
1139 		l->l_priority = l->l_usrpri;
1140 }
1141 
1142 void
1143 suspendsched()
1144 {
1145 	struct lwp *l;
1146 	int s;
1147 
1148 	/*
1149 	 * Convert all non-P_SYSTEM LSSLEEP or LSRUN processes to
1150 	 * LSSUSPENDED.
1151 	 */
1152 	proclist_lock_read();
1153 	SCHED_LOCK(s);
1154 	LIST_FOREACH(l, &alllwp, l_list) {
1155 		if ((l->l_proc->p_flag & P_SYSTEM) != 0)
1156 			continue;
1157 
1158 		switch (l->l_stat) {
1159 		case LSRUN:
1160 			l->l_proc->p_nrlwps--;
1161 			if ((l->l_flag & L_INMEM) != 0)
1162 				remrunqueue(l);
1163 			/* FALLTHROUGH */
1164 		case LSSLEEP:
1165 			l->l_stat = LSSUSPENDED;
1166 			break;
1167 		case LSONPROC:
1168 			/*
1169 			 * XXX SMP: we need to deal with processes on
1170 			 * others CPU !
1171 			 */
1172 			break;
1173 		default:
1174 			break;
1175 		}
1176 	}
1177 	SCHED_UNLOCK(s);
1178 	proclist_unlock_read();
1179 }
1180 
1181 /*
1182  * Low-level routines to access the run queue.  Optimised assembler
1183  * routines can override these.
1184  */
1185 
1186 #ifndef __HAVE_MD_RUNQUEUE
1187 
1188 /*
1189  * On some architectures, it's faster to use a MSB ordering for the priorites
1190  * than the traditional LSB ordering.
1191  */
1192 #ifdef __HAVE_BIGENDIAN_BITOPS
1193 #define	RQMASK(n) (0x80000000 >> (n))
1194 #else
1195 #define	RQMASK(n) (0x00000001 << (n))
1196 #endif
1197 
1198 /*
1199  * The primitives that manipulate the run queues.  whichqs tells which
1200  * of the 32 queues qs have processes in them.  Setrunqueue puts processes
1201  * into queues, remrunqueue removes them from queues.  The running process is
1202  * on no queue, other processes are on a queue related to p->p_priority,
1203  * divided by 4 actually to shrink the 0-127 range of priorities into the 32
1204  * available queues.
1205  */
1206 
1207 #ifdef RQDEBUG
1208 static void
1209 checkrunqueue(int whichq, struct lwp *l)
1210 {
1211 	const struct prochd * const rq = &sched_qs[whichq];
1212 	struct lwp *l2;
1213 	int found = 0;
1214 	int die = 0;
1215 	int empty = 1;
1216 	for (l2 = rq->ph_link; l2 != (void*) rq; l2 = l2->l_forw) {
1217 		if (l2->l_stat != LSRUN) {
1218 			printf("checkrunqueue[%d]: lwp %p state (%d) "
1219 			    " != LSRUN\n", whichq, l2, l2->l_stat);
1220 		}
1221 		if (l2->l_back->l_forw != l2) {
1222 			printf("checkrunqueue[%d]: lwp %p back-qptr (%p) "
1223 			    "corrupt %p\n", whichq, l2, l2->l_back,
1224 			    l2->l_back->l_forw);
1225 			die = 1;
1226 		}
1227 		if (l2->l_forw->l_back != l2) {
1228 			printf("checkrunqueue[%d]: lwp %p forw-qptr (%p) "
1229 			    "corrupt %p\n", whichq, l2, l2->l_forw,
1230 			    l2->l_forw->l_back);
1231 			die = 1;
1232 		}
1233 		if (l2 == l)
1234 			found = 1;
1235 		empty = 0;
1236 	}
1237 	if (empty && (sched_whichqs & RQMASK(whichq)) != 0) {
1238 		printf("checkrunqueue[%d]: bit set for empty run-queue %p\n",
1239 		    whichq, rq);
1240 		die = 1;
1241 	} else if (!empty && (sched_whichqs & RQMASK(whichq)) == 0) {
1242 		printf("checkrunqueue[%d]: bit clear for non-empty "
1243 		    "run-queue %p\n", whichq, rq);
1244 		die = 1;
1245 	}
1246 	if (l != NULL && (sched_whichqs & RQMASK(whichq)) == 0) {
1247 		printf("checkrunqueue[%d]: bit clear for active lwp %p\n",
1248 		    whichq, l);
1249 		die = 1;
1250 	}
1251 	if (l != NULL && empty) {
1252 		printf("checkrunqueue[%d]: empty run-queue %p with "
1253 		    "active lwp %p\n", whichq, rq, l);
1254 		die = 1;
1255 	}
1256 	if (l != NULL && !found) {
1257 		printf("checkrunqueue[%d]: lwp %p not in runqueue %p!",
1258 		    whichq, l, rq);
1259 		die = 1;
1260 	}
1261 	if (die)
1262 		panic("checkrunqueue: inconsistency found");
1263 }
1264 #endif /* RQDEBUG */
1265 
1266 void
1267 setrunqueue(struct lwp *l)
1268 {
1269 	struct prochd *rq;
1270 	struct lwp *prev;
1271 	const int whichq = l->l_priority / 4;
1272 
1273 #ifdef RQDEBUG
1274 	checkrunqueue(whichq, NULL);
1275 #endif
1276 #ifdef DIAGNOSTIC
1277 	if (l->l_back != NULL || l->l_wchan != NULL || l->l_stat != LSRUN)
1278 		panic("setrunqueue");
1279 #endif
1280 	sched_whichqs |= RQMASK(whichq);
1281 	rq = &sched_qs[whichq];
1282 	prev = rq->ph_rlink;
1283 	l->l_forw = (struct lwp *)rq;
1284 	rq->ph_rlink = l;
1285 	prev->l_forw = l;
1286 	l->l_back = prev;
1287 #ifdef RQDEBUG
1288 	checkrunqueue(whichq, l);
1289 #endif
1290 }
1291 
1292 void
1293 remrunqueue(struct lwp *l)
1294 {
1295 	struct lwp *prev, *next;
1296 	const int whichq = l->l_priority / 4;
1297 #ifdef RQDEBUG
1298 	checkrunqueue(whichq, l);
1299 #endif
1300 #ifdef DIAGNOSTIC
1301 	if (((sched_whichqs & RQMASK(whichq)) == 0))
1302 		panic("remrunqueue: bit %d not set", whichq);
1303 #endif
1304 	prev = l->l_back;
1305 	l->l_back = NULL;
1306 	next = l->l_forw;
1307 	prev->l_forw = next;
1308 	next->l_back = prev;
1309 	if (prev == next)
1310 		sched_whichqs &= ~RQMASK(whichq);
1311 #ifdef RQDEBUG
1312 	checkrunqueue(whichq, NULL);
1313 #endif
1314 }
1315 
1316 #undef RQMASK
1317 #endif /* !defined(__HAVE_MD_RUNQUEUE) */
1318