Home | History | Annotate | Line # | Download | only in kern
kern_synch.c revision 1.177.2.19
      1  1.177.2.19     rmind /*	$NetBSD: kern_synch.c,v 1.177.2.19 2007/03/24 16:50:26 rmind Exp $	*/
      2        1.63   thorpej 
      3        1.63   thorpej /*-
      4       1.174        ad  * Copyright (c) 1999, 2000, 2004, 2006, 2007 The NetBSD Foundation, Inc.
      5        1.63   thorpej  * All rights reserved.
      6        1.63   thorpej  *
      7        1.63   thorpej  * This code is derived from software contributed to The NetBSD Foundation
      8        1.63   thorpej  * by Jason R. Thorpe of the Numerical Aerospace Simulation Facility,
      9   1.177.2.5     rmind  * NASA Ames Research Center, by Charles M. Hannum, Andrew Doran and
     10   1.177.2.5     rmind  * Daniel Sieger.
     11        1.63   thorpej  *
     12        1.63   thorpej  * Redistribution and use in source and binary forms, with or without
     13        1.63   thorpej  * modification, are permitted provided that the following conditions
     14        1.63   thorpej  * are met:
     15        1.63   thorpej  * 1. Redistributions of source code must retain the above copyright
     16        1.63   thorpej  *    notice, this list of conditions and the following disclaimer.
     17        1.63   thorpej  * 2. Redistributions in binary form must reproduce the above copyright
     18        1.63   thorpej  *    notice, this list of conditions and the following disclaimer in the
     19        1.63   thorpej  *    documentation and/or other materials provided with the distribution.
     20        1.63   thorpej  * 3. All advertising materials mentioning features or use of this software
     21        1.63   thorpej  *    must display the following acknowledgement:
     22        1.63   thorpej  *	This product includes software developed by the NetBSD
     23        1.63   thorpej  *	Foundation, Inc. and its contributors.
     24        1.63   thorpej  * 4. Neither the name of The NetBSD Foundation nor the names of its
     25        1.63   thorpej  *    contributors may be used to endorse or promote products derived
     26        1.63   thorpej  *    from this software without specific prior written permission.
     27        1.63   thorpej  *
     28        1.63   thorpej  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
     29        1.63   thorpej  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
     30        1.63   thorpej  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
     31        1.63   thorpej  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
     32        1.63   thorpej  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
     33        1.63   thorpej  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
     34        1.63   thorpej  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
     35        1.63   thorpej  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
     36        1.63   thorpej  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
     37        1.63   thorpej  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
     38        1.63   thorpej  * POSSIBILITY OF SUCH DAMAGE.
     39        1.63   thorpej  */
     40        1.26       cgd 
     41        1.26       cgd /*-
     42        1.26       cgd  * Copyright (c) 1982, 1986, 1990, 1991, 1993
     43        1.26       cgd  *	The Regents of the University of California.  All rights reserved.
     44        1.26       cgd  * (c) UNIX System Laboratories, Inc.
     45        1.26       cgd  * All or some portions of this file are derived from material licensed
     46        1.26       cgd  * to the University of California by American Telephone and Telegraph
     47        1.26       cgd  * Co. or Unix System Laboratories, Inc. and are reproduced herein with
     48        1.26       cgd  * the permission of UNIX System Laboratories, Inc.
     49        1.26       cgd  *
     50        1.26       cgd  * Redistribution and use in source and binary forms, with or without
     51        1.26       cgd  * modification, are permitted provided that the following conditions
     52        1.26       cgd  * are met:
     53        1.26       cgd  * 1. Redistributions of source code must retain the above copyright
     54        1.26       cgd  *    notice, this list of conditions and the following disclaimer.
     55        1.26       cgd  * 2. Redistributions in binary form must reproduce the above copyright
     56        1.26       cgd  *    notice, this list of conditions and the following disclaimer in the
     57        1.26       cgd  *    documentation and/or other materials provided with the distribution.
     58       1.136       agc  * 3. Neither the name of the University nor the names of its contributors
     59        1.26       cgd  *    may be used to endorse or promote products derived from this software
     60        1.26       cgd  *    without specific prior written permission.
     61        1.26       cgd  *
     62        1.26       cgd  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     63        1.26       cgd  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     64        1.26       cgd  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     65        1.26       cgd  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     66        1.26       cgd  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     67        1.26       cgd  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     68        1.26       cgd  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     69        1.26       cgd  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     70        1.26       cgd  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     71        1.26       cgd  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     72        1.26       cgd  * SUCH DAMAGE.
     73        1.26       cgd  *
     74        1.50      fvdl  *	@(#)kern_synch.c	8.9 (Berkeley) 5/19/95
     75        1.26       cgd  */
     76       1.106     lukem 
     77       1.106     lukem #include <sys/cdefs.h>
     78  1.177.2.19     rmind __KERNEL_RCSID(0, "$NetBSD: kern_synch.c,v 1.177.2.19 2007/03/24 16:50:26 rmind Exp $");
     79        1.48       mrg 
     80       1.109      yamt #include "opt_kstack.h"
     81        1.82   thorpej #include "opt_lockdebug.h"
     82        1.83   thorpej #include "opt_multiprocessor.h"
     83       1.110    briggs #include "opt_perfctrs.h"
     84        1.26       cgd 
     85       1.174        ad #define	__MUTEX_PRIVATE
     86       1.174        ad 
     87        1.26       cgd #include <sys/param.h>
     88        1.26       cgd #include <sys/systm.h>
     89        1.26       cgd #include <sys/proc.h>
     90        1.26       cgd #include <sys/kernel.h>
     91       1.111    briggs #if defined(PERFCTRS)
     92       1.110    briggs #include <sys/pmc.h>
     93       1.111    briggs #endif
     94   1.177.2.8      yamt #include <sys/cpu.h>
     95        1.26       cgd #include <sys/resourcevar.h>
     96        1.55      ross #include <sys/sched.h>
     97  1.177.2.10      yamt #include <sys/syscall_stats.h>
     98       1.174        ad #include <sys/sleepq.h>
     99       1.174        ad #include <sys/lockdebug.h>
    100        1.47       mrg 
    101        1.47       mrg #include <uvm/uvm_extern.h>
    102        1.47       mrg 
    103        1.26       cgd int	lbolt;			/* once a second sleep address */
    104        1.26       cgd 
    105  1.177.2.10      yamt static void	sched_unsleep(struct lwp *);
    106  1.177.2.10      yamt static void	sched_changepri(struct lwp *, pri_t);
    107  1.177.2.10      yamt static void	sched_lendpri(struct lwp *, pri_t);
    108       1.122   thorpej 
    109       1.174        ad syncobj_t sleep_syncobj = {
    110       1.174        ad 	SOBJ_SLEEPQ_SORTED,
    111       1.174        ad 	sleepq_unsleep,
    112  1.177.2.10      yamt 	sleepq_changepri,
    113  1.177.2.10      yamt 	sleepq_lendpri,
    114  1.177.2.10      yamt 	syncobj_noowner,
    115       1.174        ad };
    116       1.174        ad 
    117       1.174        ad syncobj_t sched_syncobj = {
    118       1.174        ad 	SOBJ_SLEEPQ_SORTED,
    119       1.174        ad 	sched_unsleep,
    120  1.177.2.10      yamt 	sched_changepri,
    121  1.177.2.10      yamt 	sched_lendpri,
    122  1.177.2.10      yamt 	syncobj_noowner,
    123       1.174        ad };
    124       1.122   thorpej 
    125        1.26       cgd /*
    126       1.174        ad  * During autoconfiguration or after a panic, a sleep will simply lower the
    127       1.174        ad  * priority briefly to allow interrupts, then return.  The priority to be
    128       1.174        ad  * used (safepri) is machine-dependent, thus this value is initialized and
    129       1.174        ad  * maintained in the machine-dependent layers.  This priority will typically
    130       1.174        ad  * be 0, or the lowest priority that is safe for use on the interrupt stack;
    131       1.174        ad  * it can be made higher to block network software interrupts after panics.
    132        1.26       cgd  */
    133       1.174        ad int	safepri;
    134        1.26       cgd 
    135        1.26       cgd /*
    136       1.174        ad  * OBSOLETE INTERFACE
    137       1.174        ad  *
    138        1.26       cgd  * General sleep call.  Suspends the current process until a wakeup is
    139        1.26       cgd  * performed on the specified identifier.  The process will then be made
    140       1.174        ad  * runnable with the specified priority.  Sleeps at most timo/hz seconds (0
    141       1.174        ad  * means no timeout).  If pri includes PCATCH flag, signals are checked
    142        1.26       cgd  * before and after sleeping, else signals are not checked.  Returns 0 if
    143        1.26       cgd  * awakened, EWOULDBLOCK if the timeout expires.  If PCATCH is set and a
    144        1.26       cgd  * signal needs to be delivered, ERESTART is returned if the current system
    145        1.26       cgd  * call should be restarted if possible, and EINTR is returned if the system
    146        1.26       cgd  * call should be interrupted by the signal (return EINTR).
    147        1.77   thorpej  *
    148       1.174        ad  * The interlock is held until we are on a sleep queue. The interlock will
    149       1.174        ad  * be locked before returning back to the caller unless the PNORELOCK flag
    150       1.174        ad  * is specified, in which case the interlock will always be unlocked upon
    151       1.174        ad  * return.
    152        1.26       cgd  */
    153        1.26       cgd int
    154  1.177.2.10      yamt ltsleep(wchan_t ident, pri_t priority, const char *wmesg, int timo,
    155       1.174        ad 	volatile struct simplelock *interlock)
    156        1.26       cgd {
    157       1.122   thorpej 	struct lwp *l = curlwp;
    158       1.174        ad 	sleepq_t *sq;
    159       1.174        ad 	int error, catch;
    160        1.26       cgd 
    161       1.174        ad 	if (sleepq_dontsleep(l)) {
    162       1.174        ad 		(void)sleepq_abort(NULL, 0);
    163       1.174        ad 		if ((priority & PNORELOCK) != 0)
    164        1.77   thorpej 			simple_unlock(interlock);
    165       1.174        ad 		return 0;
    166        1.26       cgd 	}
    167        1.78  sommerfe 
    168       1.174        ad 	sq = sleeptab_lookup(&sleeptab, ident);
    169       1.174        ad 	sleepq_enter(sq, l);
    170        1.42       cgd 
    171       1.174        ad 	if (interlock != NULL) {
    172       1.174        ad 		LOCK_ASSERT(simple_lock_held(interlock));
    173       1.174        ad 		simple_unlock(interlock);
    174       1.150       chs 	}
    175       1.150       chs 
    176       1.174        ad 	catch = priority & PCATCH;
    177       1.174        ad 	sleepq_block(sq, priority & PRIMASK, ident, wmesg, timo, catch,
    178       1.174        ad 	    &sleep_syncobj);
    179       1.174        ad 	error = sleepq_unblock(timo, catch);
    180       1.126        pk 
    181       1.174        ad 	if (interlock != NULL && (priority & PNORELOCK) == 0)
    182       1.126        pk 		simple_lock(interlock);
    183       1.174        ad 
    184       1.174        ad 	return error;
    185        1.26       cgd }
    186        1.26       cgd 
    187  1.177.2.15     rmind int
    188  1.177.2.15     rmind mtsleep(wchan_t ident, pri_t priority, const char *wmesg, int timo,
    189  1.177.2.15     rmind 	kmutex_t *mtx)
    190  1.177.2.15     rmind {
    191  1.177.2.15     rmind 	struct lwp *l = curlwp;
    192  1.177.2.15     rmind 	sleepq_t *sq;
    193  1.177.2.15     rmind 	int error, catch;
    194  1.177.2.15     rmind 
    195  1.177.2.15     rmind 	if (sleepq_dontsleep(l)) {
    196  1.177.2.15     rmind 		(void)sleepq_abort(mtx, (priority & PNORELOCK) != 0);
    197  1.177.2.15     rmind 		return 0;
    198  1.177.2.15     rmind 	}
    199  1.177.2.15     rmind 
    200  1.177.2.15     rmind 	sq = sleeptab_lookup(&sleeptab, ident);
    201  1.177.2.15     rmind 	sleepq_enter(sq, l);
    202  1.177.2.15     rmind 	mutex_exit(mtx);
    203  1.177.2.15     rmind 
    204  1.177.2.15     rmind 	catch = priority & PCATCH;
    205  1.177.2.15     rmind 	sleepq_block(sq, priority & PRIMASK, ident, wmesg, timo, catch,
    206  1.177.2.15     rmind 	    &sleep_syncobj);
    207  1.177.2.15     rmind 	error = sleepq_unblock(timo, catch);
    208  1.177.2.15     rmind 
    209  1.177.2.15     rmind 	if ((priority & PNORELOCK) == 0)
    210  1.177.2.15     rmind 		mutex_enter(mtx);
    211  1.177.2.15     rmind 
    212  1.177.2.15     rmind 	return error;
    213  1.177.2.15     rmind }
    214  1.177.2.15     rmind 
    215        1.26       cgd /*
    216       1.174        ad  * General sleep call for situations where a wake-up is not expected.
    217        1.26       cgd  */
    218       1.174        ad int
    219  1.177.2.10      yamt kpause(const char *wmesg, bool intr, int timo, kmutex_t *mtx)
    220        1.26       cgd {
    221       1.174        ad 	struct lwp *l = curlwp;
    222       1.174        ad 	sleepq_t *sq;
    223       1.174        ad 	int error;
    224        1.26       cgd 
    225       1.174        ad 	if (sleepq_dontsleep(l))
    226       1.174        ad 		return sleepq_abort(NULL, 0);
    227        1.26       cgd 
    228       1.174        ad 	if (mtx != NULL)
    229       1.174        ad 		mutex_exit(mtx);
    230       1.174        ad 	sq = sleeptab_lookup(&sleeptab, l);
    231       1.174        ad 	sleepq_enter(sq, l);
    232       1.174        ad 	sleepq_block(sq, sched_kpri(l), l, wmesg, timo, intr, &sleep_syncobj);
    233       1.174        ad 	error = sleepq_unblock(timo, intr);
    234       1.174        ad 	if (mtx != NULL)
    235       1.174        ad 		mutex_enter(mtx);
    236        1.83   thorpej 
    237       1.174        ad 	return error;
    238       1.139        cl }
    239       1.139        cl 
    240        1.26       cgd /*
    241       1.174        ad  * OBSOLETE INTERFACE
    242       1.174        ad  *
    243        1.26       cgd  * Make all processes sleeping on the specified identifier runnable.
    244        1.26       cgd  */
    245        1.26       cgd void
    246       1.174        ad wakeup(wchan_t ident)
    247        1.26       cgd {
    248       1.174        ad 	sleepq_t *sq;
    249        1.83   thorpej 
    250       1.174        ad 	if (cold)
    251       1.174        ad 		return;
    252        1.83   thorpej 
    253       1.174        ad 	sq = sleeptab_lookup(&sleeptab, ident);
    254       1.174        ad 	sleepq_wake(sq, ident, (u_int)-1);
    255        1.63   thorpej }
    256        1.63   thorpej 
    257        1.63   thorpej /*
    258       1.174        ad  * OBSOLETE INTERFACE
    259       1.174        ad  *
    260        1.63   thorpej  * Make the highest priority process first in line on the specified
    261        1.63   thorpej  * identifier runnable.
    262        1.63   thorpej  */
    263       1.174        ad void
    264       1.174        ad wakeup_one(wchan_t ident)
    265        1.63   thorpej {
    266       1.174        ad 	sleepq_t *sq;
    267        1.63   thorpej 
    268       1.174        ad 	if (cold)
    269       1.174        ad 		return;
    270       1.174        ad 
    271       1.174        ad 	sq = sleeptab_lookup(&sleeptab, ident);
    272       1.174        ad 	sleepq_wake(sq, ident, 1);
    273       1.174        ad }
    274        1.63   thorpej 
    275       1.117  gmcgarry 
    276       1.117  gmcgarry /*
    277       1.117  gmcgarry  * General yield call.  Puts the current process back on its run queue and
    278       1.117  gmcgarry  * performs a voluntary context switch.  Should only be called when the
    279       1.117  gmcgarry  * current process explicitly requests it (eg sched_yield(2) in compat code).
    280       1.117  gmcgarry  */
    281       1.117  gmcgarry void
    282       1.117  gmcgarry yield(void)
    283       1.117  gmcgarry {
    284       1.122   thorpej 	struct lwp *l = curlwp;
    285       1.117  gmcgarry 
    286       1.174        ad 	KERNEL_UNLOCK_ALL(l, &l->l_biglocks);
    287       1.174        ad 	lwp_lock(l);
    288       1.174        ad 	if (l->l_stat == LSONPROC) {
    289  1.177.2.18     rmind 		KASSERT(lwp_locked(l, l->l_cpu->ci_schedstate.spc_mutex));
    290       1.174        ad 		l->l_priority = l->l_usrpri;
    291       1.174        ad 	}
    292       1.174        ad 	l->l_nvcsw++;
    293  1.177.2.13     rmind 	mi_switch(l);
    294       1.174        ad 	KERNEL_LOCK(l->l_biglocks, l);
    295        1.69   thorpej }
    296        1.69   thorpej 
    297        1.69   thorpej /*
    298        1.69   thorpej  * General preemption call.  Puts the current process back on its run queue
    299       1.156    rpaulo  * and performs an involuntary context switch.
    300        1.69   thorpej  */
    301        1.69   thorpej void
    302       1.174        ad preempt(void)
    303        1.69   thorpej {
    304       1.122   thorpej 	struct lwp *l = curlwp;
    305        1.69   thorpej 
    306       1.174        ad 	KERNEL_UNLOCK_ALL(l, &l->l_biglocks);
    307       1.174        ad 	lwp_lock(l);
    308       1.174        ad 	if (l->l_stat == LSONPROC) {
    309  1.177.2.18     rmind 		KASSERT(lwp_locked(l, l->l_cpu->ci_schedstate.spc_mutex));
    310       1.174        ad 		l->l_priority = l->l_usrpri;
    311       1.174        ad 	}
    312       1.174        ad 	l->l_nivcsw++;
    313  1.177.2.13     rmind 	(void)mi_switch(l);
    314       1.174        ad 	KERNEL_LOCK(l->l_biglocks, l);
    315        1.69   thorpej }
    316        1.69   thorpej 
    317        1.69   thorpej /*
    318   1.177.2.1      yamt  * sched_switch_unlock: update 'curlwp' and release old lwp.
    319   1.177.2.1      yamt  */
    320   1.177.2.1      yamt 
    321   1.177.2.1      yamt void
    322   1.177.2.1      yamt sched_switch_unlock(struct lwp *old, struct lwp *new)
    323   1.177.2.1      yamt {
    324   1.177.2.1      yamt 
    325   1.177.2.1      yamt 	KASSERT(old == NULL || old == curlwp);
    326  1.177.2.12      yamt 	KASSERT(new != NULL);
    327   1.177.2.1      yamt 
    328   1.177.2.1      yamt 	if (old != NULL) {
    329   1.177.2.4      yamt 		LOCKDEBUG_BARRIER(old->l_mutex, 1);
    330  1.177.2.17        ad 		lwp_unlock(old);
    331   1.177.2.1      yamt 	} else {
    332   1.177.2.1      yamt 		LOCKDEBUG_BARRIER(NULL, 1);
    333   1.177.2.1      yamt 	}
    334   1.177.2.1      yamt 	curlwp = new;
    335   1.177.2.1      yamt 	spl0();
    336   1.177.2.1      yamt }
    337   1.177.2.1      yamt 
    338   1.177.2.1      yamt /*
    339   1.177.2.3      yamt  * Compute the amount of time during which the current lwp was running.
    340   1.177.2.3      yamt  *
    341   1.177.2.3      yamt  * - update l_rtime unless it's an idle lwp.
    342   1.177.2.3      yamt  * - update spc_runtime for the next lwp.
    343   1.177.2.3      yamt  */
    344   1.177.2.3      yamt 
    345   1.177.2.3      yamt static inline void
    346   1.177.2.3      yamt updatertime(struct lwp *l, struct schedstate_percpu *spc)
    347   1.177.2.3      yamt {
    348   1.177.2.3      yamt 	struct timeval tv;
    349   1.177.2.3      yamt 	long s, u;
    350   1.177.2.3      yamt 
    351  1.177.2.10      yamt 	if ((l->l_flag & LW_IDLE) != 0) {
    352   1.177.2.3      yamt 		microtime(&spc->spc_runtime);
    353   1.177.2.3      yamt 		return;
    354   1.177.2.3      yamt 	}
    355   1.177.2.3      yamt 
    356   1.177.2.3      yamt 	microtime(&tv);
    357   1.177.2.3      yamt 	u = l->l_rtime.tv_usec + (tv.tv_usec - spc->spc_runtime.tv_usec);
    358   1.177.2.3      yamt 	s = l->l_rtime.tv_sec + (tv.tv_sec - spc->spc_runtime.tv_sec);
    359   1.177.2.3      yamt 	if (u < 0) {
    360   1.177.2.3      yamt 		u += 1000000;
    361   1.177.2.3      yamt 		s--;
    362   1.177.2.3      yamt 	} else if (u >= 1000000) {
    363   1.177.2.3      yamt 		u -= 1000000;
    364   1.177.2.3      yamt 		s++;
    365   1.177.2.3      yamt 	}
    366   1.177.2.3      yamt 	l->l_rtime.tv_usec = u;
    367   1.177.2.3      yamt 	l->l_rtime.tv_sec = s;
    368   1.177.2.3      yamt 
    369   1.177.2.3      yamt 	spc->spc_runtime = tv;
    370   1.177.2.3      yamt }
    371   1.177.2.3      yamt 
    372   1.177.2.3      yamt /*
    373  1.177.2.14     rmind  * The machine independent parts of context switch.
    374       1.130   nathanw  *
    375       1.122   thorpej  * Returns 1 if another process was actually run.
    376        1.26       cgd  */
    377       1.122   thorpej int
    378  1.177.2.13     rmind mi_switch(struct lwp *l)
    379        1.26       cgd {
    380        1.76   thorpej 	struct schedstate_percpu *spc;
    381  1.177.2.13     rmind 	struct lwp *newl;
    382       1.174        ad 	int retval, oldspl;
    383        1.26       cgd 
    384       1.174        ad 	LOCK_ASSERT(lwp_locked(l, NULL));
    385       1.174        ad 
    386       1.174        ad #ifdef LOCKDEBUG
    387       1.174        ad 	spinlock_switchcheck();
    388       1.174        ad 	simple_lock_switchcheck();
    389       1.174        ad #endif
    390       1.174        ad #ifdef KSTACK_CHECK_MAGIC
    391       1.174        ad 	kstack_check_magic(l);
    392       1.174        ad #endif
    393        1.83   thorpej 
    394        1.90  sommerfe 	/*
    395       1.174        ad 	 * It's safe to read the per CPU schedstate unlocked here, as all we
    396       1.174        ad 	 * are after is the run time and that's guarenteed to have been last
    397       1.174        ad 	 * updated by this CPU.
    398        1.90  sommerfe 	 */
    399       1.122   thorpej 	KDASSERT(l->l_cpu == curcpu());
    400       1.122   thorpej 	spc = &l->l_cpu->ci_schedstate;
    401        1.76   thorpej 
    402  1.177.2.10      yamt 	/* Count time spent in current system call */
    403  1.177.2.10      yamt 	SYSCALL_TIME_SLEEP(l);
    404  1.177.2.10      yamt 
    405        1.26       cgd 	/*
    406       1.174        ad 	 * XXXSMP If we are using h/w performance counters, save context.
    407        1.69   thorpej 	 */
    408       1.174        ad #if PERFCTRS
    409       1.175  christos 	if (PMC_ENABLED(l->l_proc)) {
    410       1.175  christos 		pmc_save_context(l->l_proc);
    411       1.174        ad 	}
    412       1.109      yamt #endif
    413        1.26       cgd 
    414       1.113  gmcgarry 	/*
    415  1.177.2.16     rmind 	 * If on the CPU and we have gotten this far, then we must yield.
    416  1.177.2.16     rmind 	 */
    417  1.177.2.16     rmind 	KASSERT(l->l_stat != LSRUN);
    418  1.177.2.16     rmind 	if (l->l_stat == LSONPROC) {
    419  1.177.2.18     rmind 		KASSERT(lwp_locked(l, l->l_cpu->ci_schedstate.spc_mutex));
    420  1.177.2.16     rmind 		l->l_stat = LSRUN;
    421  1.177.2.16     rmind 		if ((l->l_flag & LW_IDLE) == 0) {
    422  1.177.2.16     rmind 			sched_enqueue(l, true);
    423  1.177.2.16     rmind 		}
    424  1.177.2.16     rmind 	}
    425  1.177.2.16     rmind 
    426  1.177.2.16     rmind 	/*
    427       1.174        ad 	 * Process is about to yield the CPU; clear the appropriate
    428       1.174        ad 	 * scheduling flags.
    429       1.174        ad 	 */
    430       1.174        ad 	spc->spc_flags &= ~SPCF_SWITCHCLEAR;
    431       1.174        ad 
    432   1.177.2.2      yamt 	LOCKDEBUG_BARRIER(l->l_mutex, 1);
    433       1.174        ad 
    434       1.174        ad 	/*
    435   1.177.2.1      yamt 	 * Switch to the new LWP if necessary.
    436   1.177.2.1      yamt 	 * When we run again, we'll return back here.
    437       1.174        ad 	 */
    438       1.174        ad 	oldspl = MUTEX_SPIN_OLDSPL(l->l_cpu);
    439       1.174        ad 
    440   1.177.2.2      yamt 	/*
    441  1.177.2.18     rmind 	 * Acquire the spc_mutex if necessary.
    442   1.177.2.2      yamt 	 */
    443   1.177.2.2      yamt #if defined(MULTIPROCESSOR) || defined(LOCKDEBUG)
    444  1.177.2.18     rmind 	if (l->l_mutex != spc->spc_mutex) {
    445  1.177.2.18     rmind 		mutex_enter(spc->spc_mutex);
    446   1.177.2.2      yamt 	}
    447   1.177.2.2      yamt #endif
    448  1.177.2.14     rmind 	/*
    449  1.177.2.16     rmind 	 * Let sched_nextlwp() select the LWP to run the CPU next.
    450  1.177.2.14     rmind 	 * If no LWP is runnable, switch to the idle LWP.
    451  1.177.2.14     rmind 	 */
    452  1.177.2.19     rmind 	newl = sched_nextlwp();
    453  1.177.2.16     rmind 	if (newl) {
    454  1.177.2.16     rmind 		sched_dequeue(newl);
    455  1.177.2.16     rmind 	} else {
    456   1.177.2.1      yamt 		newl = l->l_cpu->ci_data.cpu_idlelwp;
    457   1.177.2.1      yamt 		KASSERT(newl != NULL);
    458   1.177.2.1      yamt 	}
    459  1.177.2.18     rmind 	KASSERT(lwp_locked(newl, spc->spc_mutex));
    460   1.177.2.7      yamt 	newl->l_stat = LSONPROC;
    461   1.177.2.7      yamt 	newl->l_cpu = l->l_cpu;
    462  1.177.2.17        ad 	newl->l_flag |= LW_RUNNING;
    463   1.177.2.2      yamt 
    464   1.177.2.2      yamt #if defined(MULTIPROCESSOR) || defined(LOCKDEBUG)
    465  1.177.2.18     rmind 	if (l->l_mutex != spc->spc_mutex) {
    466  1.177.2.18     rmind 		mutex_exit(spc->spc_mutex);
    467   1.177.2.2      yamt 	}
    468   1.177.2.2      yamt #endif
    469   1.177.2.2      yamt 
    470   1.177.2.3      yamt 	updatertime(l, spc);
    471   1.177.2.1      yamt 	if (l != newl) {
    472   1.177.2.1      yamt 		struct lwp *prevlwp;
    473   1.177.2.1      yamt 
    474   1.177.2.1      yamt 		uvmexp.swtch++;
    475   1.177.2.1      yamt 		pmap_deactivate(l);
    476  1.177.2.17        ad 		l->l_flag &= ~LW_RUNNING;
    477   1.177.2.1      yamt 		prevlwp = cpu_switchto(l, newl);
    478   1.177.2.1      yamt 		sched_switch_unlock(prevlwp, l);
    479   1.177.2.1      yamt 		pmap_activate(l);
    480   1.177.2.1      yamt 		retval = 1;
    481   1.177.2.1      yamt 	} else {
    482   1.177.2.1      yamt 		sched_switch_unlock(l, l);
    483       1.122   thorpej 		retval = 0;
    484       1.122   thorpej 	}
    485       1.110    briggs 
    486   1.177.2.1      yamt 	KASSERT(l == curlwp);
    487   1.177.2.1      yamt 	KASSERT(l->l_stat == LSONPROC);
    488   1.177.2.1      yamt 
    489       1.110    briggs 	/*
    490       1.174        ad 	 * XXXSMP If we are using h/w performance counters, restore context.
    491        1.26       cgd 	 */
    492       1.114  gmcgarry #if PERFCTRS
    493       1.175  christos 	if (PMC_ENABLED(l->l_proc)) {
    494       1.175  christos 		pmc_restore_context(l->l_proc);
    495       1.166  christos 	}
    496       1.114  gmcgarry #endif
    497       1.110    briggs 
    498       1.110    briggs 	/*
    499        1.76   thorpej 	 * We're running again; record our new start time.  We might
    500       1.174        ad 	 * be running on a new CPU now, so don't use the cached
    501        1.76   thorpej 	 * schedstate_percpu pointer.
    502        1.76   thorpej 	 */
    503  1.177.2.10      yamt 	SYSCALL_TIME_WAKEUP(l);
    504       1.122   thorpej 	KDASSERT(l->l_cpu == curcpu());
    505       1.169      yamt 
    506   1.177.2.1      yamt 	(void)splsched();
    507   1.177.2.1      yamt 	splx(oldspl);
    508       1.122   thorpej 	return retval;
    509        1.26       cgd }
    510        1.26       cgd 
    511        1.26       cgd /*
    512       1.174        ad  * Change process state to be runnable, placing it on the run queue if it is
    513       1.174        ad  * in memory, and awakening the swapper if it isn't in memory.
    514       1.174        ad  *
    515       1.174        ad  * Call with the process and LWP locked.  Will return with the LWP unlocked.
    516        1.26       cgd  */
    517        1.26       cgd void
    518       1.122   thorpej setrunnable(struct lwp *l)
    519        1.26       cgd {
    520       1.122   thorpej 	struct proc *p = l->l_proc;
    521       1.174        ad 	sigset_t *ss;
    522        1.26       cgd 
    523  1.177.2.10      yamt 	KASSERT((l->l_flag & LW_IDLE) == 0);
    524  1.177.2.10      yamt 	KASSERT(mutex_owned(&p->p_smutex));
    525  1.177.2.10      yamt 	KASSERT(lwp_locked(l, NULL));
    526        1.83   thorpej 
    527       1.122   thorpej 	switch (l->l_stat) {
    528       1.122   thorpej 	case LSSTOP:
    529        1.33   mycroft 		/*
    530        1.33   mycroft 		 * If we're being traced (possibly because someone attached us
    531        1.33   mycroft 		 * while we were stopped), check for a signal from the debugger.
    532        1.33   mycroft 		 */
    533       1.174        ad 		if ((p->p_slflag & PSL_TRACED) != 0 && p->p_xstat != 0) {
    534       1.174        ad 			if ((sigprop[p->p_xstat] & SA_TOLWP) != 0)
    535       1.174        ad 				ss = &l->l_sigpend.sp_set;
    536       1.174        ad 			else
    537       1.174        ad 				ss = &p->p_sigpend.sp_set;
    538       1.174        ad 			sigaddset(ss, p->p_xstat);
    539       1.174        ad 			signotify(l);
    540        1.53   mycroft 		}
    541       1.174        ad 		p->p_nrlwps++;
    542        1.26       cgd 		break;
    543       1.174        ad 	case LSSUSPENDED:
    544  1.177.2.10      yamt 		l->l_flag &= ~LW_WSUSPEND;
    545       1.174        ad 		p->p_nrlwps++;
    546       1.122   thorpej 		break;
    547       1.174        ad 	case LSSLEEP:
    548       1.174        ad 		KASSERT(l->l_wchan != NULL);
    549        1.26       cgd 		break;
    550       1.174        ad 	default:
    551       1.174        ad 		panic("setrunnable: lwp %p state was %d", l, l->l_stat);
    552        1.26       cgd 	}
    553       1.139        cl 
    554       1.174        ad 	/*
    555       1.174        ad 	 * If the LWP was sleeping interruptably, then it's OK to start it
    556       1.174        ad 	 * again.  If not, mark it as still sleeping.
    557       1.174        ad 	 */
    558       1.174        ad 	if (l->l_wchan != NULL) {
    559       1.174        ad 		l->l_stat = LSSLEEP;
    560  1.177.2.10      yamt 		/* lwp_unsleep() will release the lock. */
    561  1.177.2.10      yamt 		lwp_unsleep(l);
    562       1.174        ad 		return;
    563       1.174        ad 	}
    564       1.139        cl 
    565  1.177.2.18     rmind 	LOCK_ASSERT(lwp_locked(l, l->l_cpu->ci_schedstate.spc_mutex));
    566       1.122   thorpej 
    567       1.174        ad 	/*
    568       1.174        ad 	 * If the LWP is still on the CPU, mark it as LSONPROC.  It may be
    569       1.174        ad 	 * about to call mi_switch(), in which case it will yield.
    570       1.174        ad 	 */
    571  1.177.2.17        ad 	if ((l->l_flag & LW_RUNNING) != 0) {
    572       1.174        ad 		l->l_stat = LSONPROC;
    573       1.174        ad 		l->l_slptime = 0;
    574       1.174        ad 		lwp_unlock(l);
    575       1.174        ad 		return;
    576       1.174        ad 	}
    577       1.122   thorpej 
    578       1.174        ad 	/*
    579       1.174        ad 	 * Set the LWP runnable.  If it's swapped out, we need to wake the swapper
    580       1.174        ad 	 * to bring it back in.  Otherwise, enter it into a run queue.
    581       1.174        ad 	 */
    582   1.177.2.5     rmind 	sched_setrunnable(l);
    583       1.174        ad 	l->l_stat = LSRUN;
    584       1.122   thorpej 	l->l_slptime = 0;
    585       1.174        ad 
    586  1.177.2.10      yamt 	if (l->l_flag & LW_INMEM) {
    587  1.177.2.16     rmind 		sched_enqueue(l, false);
    588  1.177.2.10      yamt 		resched_cpu(l);
    589       1.174        ad 		lwp_unlock(l);
    590       1.174        ad 	} else {
    591       1.174        ad 		lwp_unlock(l);
    592       1.177        ad 		uvm_kick_scheduler();
    593       1.174        ad 	}
    594        1.26       cgd }
    595        1.26       cgd 
    596       1.174        ad /*
    597       1.174        ad  * suspendsched:
    598       1.174        ad  *
    599       1.174        ad  *	Convert all non-L_SYSTEM LSSLEEP or LSRUN LWPs to LSSUSPENDED.
    600       1.174        ad  */
    601        1.94    bouyer void
    602       1.174        ad suspendsched(void)
    603        1.94    bouyer {
    604       1.174        ad #ifdef MULTIPROCESSOR
    605       1.174        ad 	CPU_INFO_ITERATOR cii;
    606       1.174        ad 	struct cpu_info *ci;
    607       1.174        ad #endif
    608       1.122   thorpej 	struct lwp *l;
    609       1.174        ad 	struct proc *p;
    610        1.94    bouyer 
    611        1.94    bouyer 	/*
    612       1.174        ad 	 * We do this by process in order not to violate the locking rules.
    613        1.94    bouyer 	 */
    614       1.174        ad 	mutex_enter(&proclist_mutex);
    615       1.174        ad 	PROCLIST_FOREACH(p, &allproc) {
    616       1.174        ad 		mutex_enter(&p->p_smutex);
    617       1.174        ad 
    618  1.177.2.10      yamt 		if ((p->p_flag & PK_SYSTEM) != 0) {
    619       1.174        ad 			mutex_exit(&p->p_smutex);
    620        1.94    bouyer 			continue;
    621       1.174        ad 		}
    622       1.174        ad 
    623       1.174        ad 		p->p_stat = SSTOP;
    624       1.174        ad 
    625       1.174        ad 		LIST_FOREACH(l, &p->p_lwps, l_sibling) {
    626       1.174        ad 			if (l == curlwp)
    627       1.174        ad 				continue;
    628       1.174        ad 
    629       1.174        ad 			lwp_lock(l);
    630       1.122   thorpej 
    631        1.97     enami 			/*
    632       1.174        ad 			 * Set L_WREBOOT so that the LWP will suspend itself
    633       1.174        ad 			 * when it tries to return to user mode.  We want to
    634       1.174        ad 			 * try and get to get as many LWPs as possible to
    635       1.174        ad 			 * the user / kernel boundary, so that they will
    636       1.174        ad 			 * release any locks that they hold.
    637        1.97     enami 			 */
    638  1.177.2.10      yamt 			l->l_flag |= (LW_WREBOOT | LW_WSUSPEND);
    639       1.174        ad 
    640       1.174        ad 			if (l->l_stat == LSSLEEP &&
    641  1.177.2.10      yamt 			    (l->l_flag & LW_SINTR) != 0) {
    642       1.174        ad 				/* setrunnable() will release the lock. */
    643       1.174        ad 				setrunnable(l);
    644       1.174        ad 				continue;
    645       1.174        ad 			}
    646       1.174        ad 
    647       1.174        ad 			lwp_unlock(l);
    648        1.94    bouyer 		}
    649       1.174        ad 
    650       1.174        ad 		mutex_exit(&p->p_smutex);
    651        1.94    bouyer 	}
    652       1.174        ad 	mutex_exit(&proclist_mutex);
    653       1.174        ad 
    654       1.174        ad 	/*
    655       1.174        ad 	 * Kick all CPUs to make them preempt any LWPs running in user mode.
    656       1.174        ad 	 * They'll trap into the kernel and suspend themselves in userret().
    657       1.174        ad 	 */
    658       1.174        ad #ifdef MULTIPROCESSOR
    659       1.174        ad 	for (CPU_INFO_FOREACH(cii, ci))
    660   1.177.2.9      yamt 		cpu_need_resched(ci, 0);
    661       1.174        ad #else
    662   1.177.2.9      yamt 	cpu_need_resched(curcpu(), 0);
    663       1.174        ad #endif
    664        1.94    bouyer }
    665       1.113  gmcgarry 
    666       1.113  gmcgarry /*
    667       1.174        ad  * sched_kpri:
    668       1.174        ad  *
    669       1.174        ad  *	Scale a priority level to a kernel priority level, usually
    670       1.174        ad  *	for an LWP that is about to sleep.
    671       1.174        ad  */
    672  1.177.2.10      yamt pri_t
    673       1.174        ad sched_kpri(struct lwp *l)
    674       1.174        ad {
    675       1.174        ad 	/*
    676       1.174        ad 	 * Scale user priorities (127 -> 50) up to kernel priorities
    677       1.174        ad 	 * in the range (49 -> 8).  Reserve the top 8 kernel priorities
    678       1.174        ad 	 * for high priority kthreads.  Kernel priorities passed in
    679       1.174        ad 	 * are left "as is".  XXX This is somewhat arbitrary.
    680       1.174        ad 	 */
    681       1.174        ad 	static const uint8_t kpri_tab[] = {
    682       1.174        ad 		 0,   1,   2,   3,   4,   5,   6,   7,
    683       1.174        ad 		 8,   9,  10,  11,  12,  13,  14,  15,
    684       1.174        ad 		16,  17,  18,  19,  20,  21,  22,  23,
    685       1.174        ad 		24,  25,  26,  27,  28,  29,  30,  31,
    686       1.174        ad 		32,  33,  34,  35,  36,  37,  38,  39,
    687       1.174        ad 		40,  41,  42,  43,  44,  45,  46,  47,
    688       1.174        ad 		48,  49,   8,   8,   9,   9,  10,  10,
    689       1.174        ad 		11,  11,  12,  12,  13,  14,  14,  15,
    690       1.174        ad 		15,  16,  16,  17,  17,  18,  18,  19,
    691       1.174        ad 		20,  20,  21,  21,  22,  22,  23,  23,
    692       1.174        ad 		24,  24,  25,  26,  26,  27,  27,  28,
    693       1.174        ad 		28,  29,  29,  30,  30,  31,  32,  32,
    694       1.174        ad 		33,  33,  34,  34,  35,  35,  36,  36,
    695       1.174        ad 		37,  38,  38,  39,  39,  40,  40,  41,
    696       1.174        ad 		41,  42,  42,  43,  44,  44,  45,  45,
    697       1.174        ad 		46,  46,  47,  47,  48,  48,  49,  49,
    698       1.174        ad 	};
    699       1.174        ad 
    700  1.177.2.10      yamt 	return (pri_t)kpri_tab[l->l_usrpri];
    701       1.174        ad }
    702       1.174        ad 
    703       1.174        ad /*
    704       1.174        ad  * sched_unsleep:
    705       1.174        ad  *
    706       1.174        ad  *	The is called when the LWP has not been awoken normally but instead
    707       1.174        ad  *	interrupted: for example, if the sleep timed out.  Because of this,
    708       1.174        ad  *	it's not a valid action for running or idle LWPs.
    709       1.174        ad  */
    710  1.177.2.10      yamt static void
    711       1.174        ad sched_unsleep(struct lwp *l)
    712       1.174        ad {
    713       1.174        ad 
    714       1.174        ad 	lwp_unlock(l);
    715       1.174        ad 	panic("sched_unsleep");
    716       1.174        ad }
    717       1.174        ad 
    718   1.177.2.5     rmind inline void
    719  1.177.2.10      yamt resched_cpu(struct lwp *l)
    720       1.174        ad {
    721   1.177.2.5     rmind 	struct cpu_info *ci;
    722  1.177.2.10      yamt 	const pri_t pri = lwp_eprio(l);
    723       1.174        ad 
    724   1.177.2.5     rmind 	/*
    725   1.177.2.5     rmind 	 * XXXSMP
    726   1.177.2.5     rmind 	 * Since l->l_cpu persists across a context switch,
    727   1.177.2.5     rmind 	 * this gives us *very weak* processor affinity, in
    728   1.177.2.5     rmind 	 * that we notify the CPU on which the process last
    729   1.177.2.5     rmind 	 * ran that it should try to switch.
    730   1.177.2.5     rmind 	 *
    731   1.177.2.5     rmind 	 * This does not guarantee that the process will run on
    732   1.177.2.5     rmind 	 * that processor next, because another processor might
    733   1.177.2.5     rmind 	 * grab it the next time it performs a context switch.
    734   1.177.2.5     rmind 	 *
    735   1.177.2.5     rmind 	 * This also does not handle the case where its last
    736   1.177.2.5     rmind 	 * CPU is running a higher-priority process, but every
    737   1.177.2.5     rmind 	 * other CPU is running a lower-priority process.  There
    738   1.177.2.5     rmind 	 * are ways to handle this situation, but they're not
    739   1.177.2.5     rmind 	 * currently very pretty, and we also need to weigh the
    740   1.177.2.5     rmind 	 * cost of moving a process from one CPU to another.
    741   1.177.2.5     rmind 	 */
    742   1.177.2.5     rmind 	ci = (l->l_cpu != NULL) ? l->l_cpu : curcpu();
    743   1.177.2.5     rmind 	if (pri < ci->ci_schedstate.spc_curpriority)
    744   1.177.2.9      yamt 		cpu_need_resched(ci, 0);
    745   1.177.2.1      yamt }
    746  1.177.2.10      yamt 
    747  1.177.2.10      yamt static void
    748  1.177.2.10      yamt sched_changepri(struct lwp *l, pri_t pri)
    749  1.177.2.10      yamt {
    750  1.177.2.10      yamt 
    751  1.177.2.18     rmind 	LOCK_ASSERT(lwp_locked(l, l->l_cpu->ci_schedstate.spc_mutex));
    752  1.177.2.10      yamt 
    753  1.177.2.10      yamt 	l->l_usrpri = pri;
    754  1.177.2.10      yamt 	if (l->l_priority < PUSER)
    755  1.177.2.10      yamt 		return;
    756  1.177.2.10      yamt 
    757  1.177.2.10      yamt 	if (l->l_stat != LSRUN || (l->l_flag & LW_INMEM) == 0) {
    758  1.177.2.10      yamt 		l->l_priority = pri;
    759  1.177.2.10      yamt 		return;
    760  1.177.2.10      yamt 	}
    761  1.177.2.10      yamt 
    762  1.177.2.11      yamt 	sched_dequeue(l);
    763  1.177.2.10      yamt 	l->l_priority = pri;
    764  1.177.2.16     rmind 	sched_enqueue(l, false);
    765  1.177.2.10      yamt 	resched_cpu(l);
    766  1.177.2.10      yamt }
    767  1.177.2.10      yamt 
    768  1.177.2.10      yamt static void
    769  1.177.2.11      yamt sched_lendpri(struct lwp *l, pri_t pri)
    770  1.177.2.10      yamt {
    771  1.177.2.10      yamt 
    772  1.177.2.18     rmind 	LOCK_ASSERT(lwp_locked(l, l->l_cpu->ci_schedstate.spc_mutex));
    773  1.177.2.10      yamt 
    774  1.177.2.10      yamt 	if (l->l_stat != LSRUN || (l->l_flag & LW_INMEM) == 0) {
    775  1.177.2.10      yamt 		l->l_inheritedprio = pri;
    776  1.177.2.10      yamt 		return;
    777  1.177.2.10      yamt 	}
    778  1.177.2.10      yamt 
    779  1.177.2.11      yamt 	sched_dequeue(l);
    780  1.177.2.10      yamt 	l->l_inheritedprio = pri;
    781  1.177.2.16     rmind 	sched_enqueue(l, false);
    782  1.177.2.10      yamt 	resched_cpu(l);
    783  1.177.2.10      yamt }
    784  1.177.2.10      yamt 
    785  1.177.2.10      yamt struct lwp *
    786  1.177.2.10      yamt syncobj_noowner(wchan_t wchan)
    787  1.177.2.10      yamt {
    788  1.177.2.10      yamt 
    789  1.177.2.10      yamt 	return NULL;
    790  1.177.2.10      yamt }
    791