Home | History | Annotate | Line # | Download | only in kern
kern_timeout.c revision 1.11
      1 /*	$NetBSD: kern_timeout.c,v 1.11 2003/09/25 10:44:11 scw Exp $	*/
      2 
      3 /*-
      4  * Copyright (c) 2003 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.
      9  *
     10  * Redistribution and use in source and binary forms, with or without
     11  * modification, are permitted provided that the following conditions
     12  * are met:
     13  * 1. Redistributions of source code must retain the above copyright
     14  *    notice, this list of conditions and the following disclaimer.
     15  * 2. Redistributions in binary form must reproduce the above copyright
     16  *    notice, this list of conditions and the following disclaimer in the
     17  *    documentation and/or other materials provided with the distribution.
     18  * 3. All advertising materials mentioning features or use of this software
     19  *    must display the following acknowledgement:
     20  *	This product includes software developed by the NetBSD
     21  *	Foundation, Inc. and its contributors.
     22  * 4. Neither the name of The NetBSD Foundation nor the names of its
     23  *    contributors may be used to endorse or promote products derived
     24  *    from this software without specific prior written permission.
     25  *
     26  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
     27  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
     28  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
     29  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
     30  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
     31  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
     32  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
     33  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
     34  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
     35  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
     36  * POSSIBILITY OF SUCH DAMAGE.
     37  */
     38 
     39 /*
     40  * Copyright (c) 2001 Thomas Nordin <nordin (at) openbsd.org>
     41  * Copyright (c) 2000-2001 Artur Grabowski <art (at) openbsd.org>
     42  * All rights reserved.
     43  *
     44  * Redistribution and use in source and binary forms, with or without
     45  * modification, are permitted provided that the following conditions
     46  * are met:
     47  *
     48  * 1. Redistributions of source code must retain the above copyright
     49  *    notice, this list of conditions and the following disclaimer.
     50  * 2. Redistributions in binary form must reproduce the above copyright
     51  *    notice, this list of conditions and the following disclaimer in the
     52  *    documentation and/or other materials provided with the distribution.
     53  * 3. The name of the author may not be used to endorse or promote products
     54  *    derived from this software without specific prior written permission.
     55  *
     56  * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES,
     57  * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
     58  * AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
     59  * THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
     60  * EXEMPLARY, OR CONSEQUENTIAL  DAMAGES (INCLUDING, BUT NOT LIMITED TO,
     61  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
     62  * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
     63  * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
     64  * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
     65  * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
     66  */
     67 
     68 #include <sys/cdefs.h>
     69 __KERNEL_RCSID(0, "$NetBSD: kern_timeout.c,v 1.11 2003/09/25 10:44:11 scw Exp $");
     70 
     71 /*
     72  * Adapted from OpenBSD: kern_timeout.c,v 1.15 2002/12/08 04:21:07 art Exp,
     73  * modified to match NetBSD's pre-existing callout API.
     74  */
     75 
     76 #include <sys/param.h>
     77 #include <sys/systm.h>
     78 #include <sys/kernel.h>
     79 #include <sys/lock.h>
     80 #include <sys/callout.h>
     81 
     82 #ifdef DDB
     83 #include <machine/db_machdep.h>
     84 #include <ddb/db_interface.h>
     85 #include <ddb/db_access.h>
     86 #include <ddb/db_sym.h>
     87 #include <ddb/db_output.h>
     88 #endif
     89 
     90 /*
     91  * Timeouts are kept in a hierarchical timing wheel. The c_time is the value
     92  * of the global variable "hardclock_ticks" when the timeout should be called.
     93  * There are four levels with 256 buckets each. See 'Scheme 7' in
     94  * "Hashed and Hierarchical Timing Wheels: Efficient Data Structures for
     95  * Implementing a Timer Facility" by George Varghese and Tony Lauck.
     96  */
     97 #define BUCKETS 1024
     98 #define WHEELSIZE 256
     99 #define WHEELMASK 255
    100 #define WHEELBITS 8
    101 
    102 static struct callout_circq timeout_wheel[BUCKETS];	/* Queues of timeouts */
    103 static struct callout_circq timeout_todo;		/* Worklist */
    104 
    105 #define MASKWHEEL(wheel, time) (((time) >> ((wheel)*WHEELBITS)) & WHEELMASK)
    106 
    107 #define BUCKET(rel, abs)						\
    108     (((rel) <= (1 << (2*WHEELBITS)))					\
    109     	? ((rel) <= (1 << WHEELBITS))					\
    110             ? &timeout_wheel[MASKWHEEL(0, (abs))]			\
    111             : &timeout_wheel[MASKWHEEL(1, (abs)) + WHEELSIZE]		\
    112         : ((rel) <= (1 << (3*WHEELBITS)))				\
    113             ? &timeout_wheel[MASKWHEEL(2, (abs)) + 2*WHEELSIZE]		\
    114             : &timeout_wheel[MASKWHEEL(3, (abs)) + 3*WHEELSIZE])
    115 
    116 #define MOVEBUCKET(wheel, time)						\
    117     CIRCQ_APPEND(&timeout_todo,						\
    118         &timeout_wheel[MASKWHEEL((wheel), (time)) + (wheel)*WHEELSIZE])
    119 
    120 /*
    121  * All wheels are locked with the same lock (which must also block out all
    122  * interrupts).
    123  */
    124 static struct simplelock callout_slock;
    125 
    126 #define	CALLOUT_LOCK(s)							\
    127 do {									\
    128 	s = splsched();							\
    129 	simple_lock(&callout_slock);					\
    130 } while (/*CONSTCOND*/0)
    131 
    132 #define	CALLOUT_UNLOCK(s)						\
    133 do {									\
    134 	simple_unlock(&callout_slock);					\
    135 	splx((s));							\
    136 } while (/*CONSTCOND*/0)
    137 
    138 /*
    139  * Circular queue definitions.
    140  */
    141 
    142 #define CIRCQ_INIT(list)						\
    143 do {									\
    144         (list)->cq_next_l = (list);					\
    145         (list)->cq_prev_l = (list);					\
    146 } while (/*CONSTCOND*/0)
    147 
    148 #define CIRCQ_INSERT(elem, list)					\
    149 do {									\
    150         (elem)->cq_prev_e = (list)->cq_prev_e;				\
    151         (elem)->cq_next_l = (list);					\
    152         (list)->cq_prev_l->cq_next_l = (elem);				\
    153         (list)->cq_prev_l = (elem);					\
    154 } while (/*CONSTCOND*/0)
    155 
    156 #define CIRCQ_APPEND(fst, snd)						\
    157 do {									\
    158         if (!CIRCQ_EMPTY(snd)) {					\
    159                 (fst)->cq_prev_l->cq_next_l = (snd)->cq_next_l;		\
    160                 (snd)->cq_next_l->cq_prev_l = (fst)->cq_prev_l;		\
    161                 (snd)->cq_prev_l->cq_next_l = (fst);			\
    162                 (fst)->cq_prev_l = (snd)->cq_prev_l;			\
    163                 CIRCQ_INIT(snd);					\
    164         }								\
    165 } while (/*CONSTCOND*/0)
    166 
    167 #define CIRCQ_REMOVE(elem)						\
    168 do {									\
    169         (elem)->cq_next_l->cq_prev_e = (elem)->cq_prev_e;		\
    170         (elem)->cq_prev_l->cq_next_e = (elem)->cq_next_e;		\
    171 } while (/*CONSTCOND*/0)
    172 
    173 #define CIRCQ_FIRST(list)	((list)->cq_next_e)
    174 #define CIRCQ_NEXT(elem)	((elem)->cq_next_e)
    175 #define CIRCQ_LAST(elem,list)	((elem)->cq_next_l == (list))
    176 #define CIRCQ_EMPTY(list)	((list)->cq_next_l == (list))
    177 
    178 /*
    179  * Some of the "math" in here is a bit tricky.
    180  *
    181  * We have to beware of wrapping ints.
    182  * We use the fact that any element added to the queue must be added with a
    183  * positive time. That means that any element `to' on the queue cannot be
    184  * scheduled to timeout further in time than INT_MAX, but c->c_time can
    185  * be positive or negative so comparing it with anything is dangerous.
    186  * The only way we can use the c->c_time value in any predictable way
    187  * is when we calculate how far in the future `to' will timeout -
    188  * "c->c_time - hardclock_ticks". The result will always be positive for
    189  * future timeouts and 0 or negative for due timeouts.
    190  */
    191 
    192 #ifdef CALLOUT_EVENT_COUNTERS
    193 static struct evcnt callout_ev_late;
    194 #endif
    195 
    196 /*
    197  * callout_startup:
    198  *
    199  *	Initialize the callout facility, called at system startup time.
    200  */
    201 void
    202 callout_startup(void)
    203 {
    204 	int b;
    205 
    206 	CIRCQ_INIT(&timeout_todo);
    207 	for (b = 0; b < BUCKETS; b++)
    208 		CIRCQ_INIT(&timeout_wheel[b]);
    209 	simple_lock_init(&callout_slock);
    210 
    211 #ifdef CALLOUT_EVENT_COUNTERS
    212 	evcnt_attach_dynamic(&callout_ev_late, EVCNT_TYPE_MISC,
    213 	    NULL, "callout", "late");
    214 #endif
    215 }
    216 
    217 /*
    218  * callout_init:
    219  *
    220  *	Initialize a callout structure.
    221  */
    222 void
    223 callout_init(struct callout *c)
    224 {
    225 
    226 	memset(c, 0, sizeof(*c));
    227 }
    228 
    229 /*
    230  * callout_setfunc:
    231  *
    232  *	Initialize a callout structure and set the function and
    233  *	argument.
    234  */
    235 void
    236 callout_setfunc(struct callout *c, void (*func)(void *), void *arg)
    237 {
    238 
    239 	memset(c, 0, sizeof(*c));
    240 	c->c_func = func;
    241 	c->c_arg = arg;
    242 }
    243 
    244 /*
    245  * callout_reset:
    246  *
    247  *	Reset a callout structure with a new function and argument, and
    248  *	schedule it to run.
    249  */
    250 void
    251 callout_reset(struct callout *c, int to_ticks, void (*func)(void *), void *arg)
    252 {
    253 	int s, old_time;
    254 
    255 	KASSERT(to_ticks >= 0);
    256 
    257 	CALLOUT_LOCK(s);
    258 
    259 	/* Initialize the time here, it won't change. */
    260 	old_time = c->c_time;
    261 	c->c_time = to_ticks + hardclock_ticks;
    262 	c->c_flags &= ~(CALLOUT_FIRED|CALLOUT_INVOKING);
    263 
    264 	c->c_func = func;
    265 	c->c_arg = arg;
    266 
    267 	/*
    268 	 * If this timeout is already scheduled and now is moved
    269 	 * earlier, reschedule it now. Otherwise leave it in place
    270 	 * and let it be rescheduled later.
    271 	 */
    272 	if (callout_pending(c)) {
    273 		if (c->c_time - old_time < 0) {
    274 			CIRCQ_REMOVE(&c->c_list);
    275 			CIRCQ_INSERT(&c->c_list, &timeout_todo);
    276 		}
    277 	} else {
    278 		c->c_flags |= CALLOUT_PENDING;
    279 		CIRCQ_INSERT(&c->c_list, &timeout_todo);
    280 	}
    281 
    282 	CALLOUT_UNLOCK(s);
    283 }
    284 
    285 /*
    286  * callout_schedule:
    287  *
    288  *	Schedule a callout to run.  The function and argument must
    289  *	already be set in the callout structure.
    290  */
    291 void
    292 callout_schedule(struct callout *c, int to_ticks)
    293 {
    294 	int s, old_time;
    295 
    296 	KASSERT(to_ticks >= 0);
    297 
    298 	CALLOUT_LOCK(s);
    299 
    300 	/* Initialize the time here, it won't change. */
    301 	old_time = c->c_time;
    302 	c->c_time = to_ticks + hardclock_ticks;
    303 	c->c_flags &= ~(CALLOUT_FIRED|CALLOUT_INVOKING);
    304 
    305 	/*
    306 	 * If this timeout is already scheduled and now is moved
    307 	 * earlier, reschedule it now. Otherwise leave it in place
    308 	 * and let it be rescheduled later.
    309 	 */
    310 	if (callout_pending(c)) {
    311 		if (c->c_time - old_time < 0) {
    312 			CIRCQ_REMOVE(&c->c_list);
    313 			CIRCQ_INSERT(&c->c_list, &timeout_todo);
    314 		}
    315 	} else {
    316 		c->c_flags |= CALLOUT_PENDING;
    317 		CIRCQ_INSERT(&c->c_list, &timeout_todo);
    318 	}
    319 
    320 	CALLOUT_UNLOCK(s);
    321 }
    322 
    323 /*
    324  * callout_stop:
    325  *
    326  *	Cancel a pending callout.
    327  */
    328 void
    329 callout_stop(struct callout *c)
    330 {
    331 	int s;
    332 
    333 	CALLOUT_LOCK(s);
    334 
    335 	if (callout_pending(c))
    336 		CIRCQ_REMOVE(&c->c_list);
    337 
    338 	c->c_flags &= ~(CALLOUT_PENDING|CALLOUT_FIRED);
    339 
    340 	CALLOUT_UNLOCK(s);
    341 }
    342 
    343 /*
    344  * This is called from hardclock() once every tick.
    345  * We return !0 if we need to schedule a softclock.
    346  */
    347 int
    348 callout_hardclock(void)
    349 {
    350 	int s;
    351 	int needsoftclock;
    352 
    353 	CALLOUT_LOCK(s);
    354 
    355 	MOVEBUCKET(0, hardclock_ticks);
    356 	if (MASKWHEEL(0, hardclock_ticks) == 0) {
    357 		MOVEBUCKET(1, hardclock_ticks);
    358 		if (MASKWHEEL(1, hardclock_ticks) == 0) {
    359 			MOVEBUCKET(2, hardclock_ticks);
    360 			if (MASKWHEEL(2, hardclock_ticks) == 0)
    361 				MOVEBUCKET(3, hardclock_ticks);
    362 		}
    363 	}
    364 
    365 	needsoftclock = !CIRCQ_EMPTY(&timeout_todo);
    366 	CALLOUT_UNLOCK(s);
    367 
    368 	return needsoftclock;
    369 }
    370 
    371 /* ARGSUSED */
    372 void
    373 softclock(void *v)
    374 {
    375 	struct callout *c;
    376 	void (*func)(void *);
    377 	void *arg;
    378 	int s;
    379 
    380 	CALLOUT_LOCK(s);
    381 
    382 	while (!CIRCQ_EMPTY(&timeout_todo)) {
    383 		c = CIRCQ_FIRST(&timeout_todo);
    384 		CIRCQ_REMOVE(&c->c_list);
    385 
    386 		/* If due run it, otherwise insert it into the right bucket. */
    387 		if (c->c_time - hardclock_ticks > 0) {
    388 			CIRCQ_INSERT(&c->c_list,
    389 			    BUCKET((c->c_time - hardclock_ticks), c->c_time));
    390 		} else {
    391 #ifdef CALLOUT_EVENT_COUNTERS
    392 			if (c->c_time - hardclock_ticks < 0)
    393 				callout_ev_late.ev_count++;
    394 #endif
    395 			c->c_flags = (c->c_flags  & ~CALLOUT_PENDING) |
    396 			    (CALLOUT_FIRED|CALLOUT_INVOKING);
    397 
    398 			func = c->c_func;
    399 			arg = c->c_arg;
    400 
    401 			CALLOUT_UNLOCK(s);
    402 			(*func)(arg);
    403 			CALLOUT_LOCK(s);
    404 		}
    405 	}
    406 
    407 	CALLOUT_UNLOCK(s);
    408 }
    409 
    410 #ifdef DDB
    411 static void
    412 db_show_callout_bucket(struct callout_circq *bucket)
    413 {
    414 	struct callout *c;
    415 	db_expr_t offset;
    416 	char *name;
    417 
    418 	if (CIRCQ_EMPTY(bucket))
    419 		return;
    420 
    421 	for (c = CIRCQ_FIRST(bucket); /*nothing*/; c = CIRCQ_NEXT(&c->c_list)) {
    422 		db_find_sym_and_offset((db_addr_t)(intptr_t)c->c_func, &name,
    423 		    &offset);
    424 		name = name ? name : "?";
    425 #ifdef _LP64
    426 #define	POINTER_WIDTH	"%16lx"
    427 #else
    428 #define	POINTER_WIDTH	"%8lx"
    429 #endif
    430 		db_printf("%9d %2d/%-4d " POINTER_WIDTH "  %s\n",
    431 		    c->c_time - hardclock_ticks,
    432 		    (int)((bucket - timeout_wheel) / WHEELSIZE),
    433 		    (int)(bucket - timeout_wheel), (u_long) c->c_arg, name);
    434 
    435 		if (CIRCQ_LAST(&c->c_list, bucket))
    436 			break;
    437 	}
    438 }
    439 
    440 void
    441 db_show_callout(db_expr_t addr, int haddr, db_expr_t count, char *modif)
    442 {
    443 	int b;
    444 
    445 	db_printf("hardclock_ticks now: %d\n", hardclock_ticks);
    446 #ifdef _LP64
    447 	db_printf("    ticks  wheel               arg  func\n");
    448 #else
    449 	db_printf("    ticks  wheel       arg  func\n");
    450 #endif
    451 
    452 	/*
    453 	 * Don't lock the callwheel; all the other CPUs are paused
    454 	 * anyhow, and we might be called in a circumstance where
    455 	 * some other CPU was paused while holding the lock.
    456 	 */
    457 
    458 	db_show_callout_bucket(&timeout_todo);
    459 	for (b = 0; b < BUCKETS; b++)
    460 		db_show_callout_bucket(&timeout_wheel[b]);
    461 }
    462 #endif /* DDB */
    463