Subversion Repositories HelenOS

Rev

Rev 3424 | Go to most recent revision | Only display areas with differences | Regard whitespace | Details | Blame | Last modification | View Log | RSS feed

Rev 3424 Rev 3425
1
/*
1
/*
2
 * Copyright (c) 2001-2007 Jakub Jermar
2
 * Copyright (c) 2001-2007 Jakub Jermar
3
 * All rights reserved.
3
 * All rights reserved.
4
 *
4
 *
5
 * Redistribution and use in source and binary forms, with or without
5
 * Redistribution and use in source and binary forms, with or without
6
 * modification, are permitted provided that the following conditions
6
 * modification, are permitted provided that the following conditions
7
 * are met:
7
 * are met:
8
 *
8
 *
9
 * - Redistributions of source code must retain the above copyright
9
 * - Redistributions of source code must retain the above copyright
10
 *   notice, this list of conditions and the following disclaimer.
10
 *   notice, this list of conditions and the following disclaimer.
11
 * - Redistributions in binary form must reproduce the above copyright
11
 * - Redistributions in binary form must reproduce the above copyright
12
 *   notice, this list of conditions and the following disclaimer in the
12
 *   notice, this list of conditions and the following disclaimer in the
13
 *   documentation and/or other materials provided with the distribution.
13
 *   documentation and/or other materials provided with the distribution.
14
 * - The name of the author may not be used to endorse or promote products
14
 * - The name of the author may not be used to endorse or promote products
15
 *   derived from this software without specific prior written permission.
15
 *   derived from this software without specific prior written permission.
16
 *
16
 *
17
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
17
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
18
 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
18
 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
19
 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
19
 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
20
 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
20
 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
21
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
21
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
22
 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
22
 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23
 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
23
 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24
 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
24
 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
25
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
26
 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26
 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
27
 */
27
 */
28
 
28
 
29
/** @addtogroup genericproc
29
/** @addtogroup genericproc
30
 * @{
30
 * @{
31
 */
31
 */
32
 
32
 
33
/**
33
/**
34
 * @file
34
 * @file
35
 * @brief   Scheduler and load balancing.
35
 * @brief   Scheduler and load balancing.
36
 *
36
 *
37
 * This file contains the scheduler and kcpulb kernel thread which
37
 * This file contains the scheduler and kcpulb kernel thread which
38
 * performs load-balancing of per-CPU run queues.
38
 * performs load-balancing of per-CPU run queues.
39
 */
39
 */
40
 
40
 
41
#include <proc/scheduler.h>
41
#include <proc/scheduler.h>
42
#include <proc/thread.h>
42
#include <proc/thread.h>
43
#include <proc/task.h>
43
#include <proc/task.h>
44
#include <mm/frame.h>
44
#include <mm/frame.h>
45
#include <mm/page.h>
45
#include <mm/page.h>
46
#include <mm/as.h>
46
#include <mm/as.h>
47
#include <time/timeout.h>
47
#include <time/timeout.h>
48
#include <time/delay.h>
48
#include <time/delay.h>
49
#include <arch/asm.h>
49
#include <arch/asm.h>
50
#include <arch/faddr.h>
50
#include <arch/faddr.h>
51
#include <arch/cycle.h>
51
#include <arch/cycle.h>
52
#include <atomic.h>
52
#include <atomic.h>
53
#include <synch/spinlock.h>
53
#include <synch/spinlock.h>
54
#include <config.h>
54
#include <config.h>
55
#include <context.h>
55
#include <context.h>
56
#include <fpu_context.h>
56
#include <fpu_context.h>
57
#include <func.h>
57
#include <func.h>
58
#include <arch.h>
58
#include <arch.h>
59
#include <adt/list.h>
59
#include <adt/list.h>
60
#include <panic.h>
60
#include <panic.h>
61
#include <cpu.h>
61
#include <cpu.h>
62
#include <print.h>
62
#include <print.h>
63
#include <debug.h>
63
#include <debug.h>
64
 
64
 
65
static void before_task_runs(void);
65
static void before_task_runs(void);
66
static void before_thread_runs(void);
66
static void before_thread_runs(void);
67
static void after_thread_ran(void);
67
static void after_thread_ran(void);
68
static void scheduler_separated_stack(void);
68
static void scheduler_separated_stack(void);
69
 
69
 
70
atomic_t nrdy;  /**< Number of ready threads in the system. */
70
atomic_t nrdy;  /**< Number of ready threads in the system. */
71
 
71
 
72
/** Carry out actions before new task runs. */
72
/** Carry out actions before new task runs. */
73
void before_task_runs(void)
73
void before_task_runs(void)
74
{
74
{
75
    before_task_runs_arch();
75
    before_task_runs_arch();
76
}
76
}
77
 
77
 
78
/** Take actions before new thread runs.
78
/** Take actions before new thread runs.
79
 *
79
 *
80
 * Perform actions that need to be
80
 * Perform actions that need to be
81
 * taken before the newly selected
81
 * taken before the newly selected
82
 * tread is passed control.
82
 * tread is passed control.
83
 *
83
 *
84
 * THREAD->lock is locked on entry
84
 * THREAD->lock is locked on entry
85
 *
85
 *
86
 */
86
 */
87
void before_thread_runs(void)
87
void before_thread_runs(void)
88
{
88
{
89
    before_thread_runs_arch();
89
    before_thread_runs_arch();
90
#ifdef CONFIG_FPU_LAZY
90
#ifdef CONFIG_FPU_LAZY
91
    if(THREAD == CPU->fpu_owner)
91
    if(THREAD == CPU->fpu_owner)
92
        fpu_enable();
92
        fpu_enable();
93
    else
93
    else
94
        fpu_disable();
94
        fpu_disable();
95
#else
95
#else
96
    fpu_enable();
96
    fpu_enable();
97
    if (THREAD->fpu_context_exists)
97
    if (THREAD->fpu_context_exists)
98
        fpu_context_restore(THREAD->saved_fpu_context);
98
        fpu_context_restore(THREAD->saved_fpu_context);
99
    else {
99
    else {
100
        fpu_init();
100
        fpu_init();
101
        THREAD->fpu_context_exists = 1;
101
        THREAD->fpu_context_exists = 1;
102
    }
102
    }
103
#endif
103
#endif
104
}
104
}
105
 
105
 
106
/** Take actions after THREAD had run.
106
/** Take actions after THREAD had run.
107
 *
107
 *
108
 * Perform actions that need to be
108
 * Perform actions that need to be
109
 * taken after the running thread
109
 * taken after the running thread
110
 * had been preempted by the scheduler.
110
 * had been preempted by the scheduler.
111
 *
111
 *
112
 * THREAD->lock is locked on entry
112
 * THREAD->lock is locked on entry
113
 *
113
 *
114
 */
114
 */
115
void after_thread_ran(void)
115
void after_thread_ran(void)
116
{
116
{
117
    after_thread_ran_arch();
117
    after_thread_ran_arch();
118
}
118
}
119
 
119
 
120
#ifdef CONFIG_FPU_LAZY
120
#ifdef CONFIG_FPU_LAZY
121
void scheduler_fpu_lazy_request(void)
121
void scheduler_fpu_lazy_request(void)
122
{
122
{
123
restart:
123
restart:
124
    fpu_enable();
124
    fpu_enable();
125
    spinlock_lock(&CPU->lock);
125
    spinlock_lock(&CPU->lock);
126
 
126
 
127
    /* Save old context */
127
    /* Save old context */
128
    if (CPU->fpu_owner != NULL) {  
128
    if (CPU->fpu_owner != NULL) {  
129
        spinlock_lock(&CPU->fpu_owner->lock);
129
        spinlock_lock(&CPU->fpu_owner->lock);
130
        fpu_context_save(CPU->fpu_owner->saved_fpu_context);
130
        fpu_context_save(CPU->fpu_owner->saved_fpu_context);
131
        /* don't prevent migration */
131
        /* don't prevent migration */
132
        CPU->fpu_owner->fpu_context_engaged = 0;
132
        CPU->fpu_owner->fpu_context_engaged = 0;
133
        spinlock_unlock(&CPU->fpu_owner->lock);
133
        spinlock_unlock(&CPU->fpu_owner->lock);
134
        CPU->fpu_owner = NULL;
134
        CPU->fpu_owner = NULL;
135
    }
135
    }
136
 
136
 
137
    spinlock_lock(&THREAD->lock);
137
    spinlock_lock(&THREAD->lock);
138
    if (THREAD->fpu_context_exists) {
138
    if (THREAD->fpu_context_exists) {
139
        fpu_context_restore(THREAD->saved_fpu_context);
139
        fpu_context_restore(THREAD->saved_fpu_context);
140
    } else {
140
    } else {
141
        /* Allocate FPU context */
141
        /* Allocate FPU context */
142
        if (!THREAD->saved_fpu_context) {
142
        if (!THREAD->saved_fpu_context) {
143
            /* Might sleep */
143
            /* Might sleep */
144
            spinlock_unlock(&THREAD->lock);
144
            spinlock_unlock(&THREAD->lock);
145
            spinlock_unlock(&CPU->lock);
145
            spinlock_unlock(&CPU->lock);
146
            THREAD->saved_fpu_context =
146
            THREAD->saved_fpu_context =
147
                (fpu_context_t *) slab_alloc(fpu_context_slab, 0);
147
                (fpu_context_t *) slab_alloc(fpu_context_slab, 0);
148
            /* We may have switched CPUs during slab_alloc */
148
            /* We may have switched CPUs during slab_alloc */
149
            goto restart;
149
            goto restart;
150
        }
150
        }
151
        fpu_init();
151
        fpu_init();
152
        THREAD->fpu_context_exists = 1;
152
        THREAD->fpu_context_exists = 1;
153
    }
153
    }
154
    CPU->fpu_owner = THREAD;
154
    CPU->fpu_owner = THREAD;
155
    THREAD->fpu_context_engaged = 1;
155
    THREAD->fpu_context_engaged = 1;
156
    spinlock_unlock(&THREAD->lock);
156
    spinlock_unlock(&THREAD->lock);
157
 
157
 
158
    spinlock_unlock(&CPU->lock);
158
    spinlock_unlock(&CPU->lock);
159
}
159
}
160
#endif
160
#endif
161
 
161
 
162
/** Initialize scheduler
162
/** Initialize scheduler
163
 *
163
 *
164
 * Initialize kernel scheduler.
164
 * Initialize kernel scheduler.
165
 *
165
 *
166
 */
166
 */
167
void scheduler_init(void)
167
void scheduler_init(void)
168
{
168
{
169
}
169
}
170
 
170
 
171
/** Get thread to be scheduled
171
/** Get thread to be scheduled
172
 *
172
 *
173
 * Get the optimal thread to be scheduled
173
 * Get the optimal thread to be scheduled
174
 * according to thread accounting and scheduler
174
 * according to thread accounting and scheduler
175
 * policy.
175
 * policy.
176
 *
176
 *
177
 * @return Thread to be scheduled.
177
 * @return Thread to be scheduled.
178
 *
178
 *
179
 */
179
 */
180
static thread_t *find_best_thread(void)
180
static thread_t *find_best_thread(void)
181
{
181
{
182
    thread_t *t;
182
    thread_t *t;
183
    runq_t *r;
183
    runq_t *r;
184
    int i;
184
    int i;
185
 
185
 
186
    ASSERT(CPU != NULL);
186
    ASSERT(CPU != NULL);
187
 
187
 
188
loop:
188
loop:
189
    interrupts_enable();
189
    interrupts_enable();
190
   
190
   
191
    if (atomic_get(&CPU->nrdy) == 0) {
191
    if (atomic_get(&CPU->nrdy) == 0) {
192
        /*
192
        /*
193
         * For there was nothing to run, the CPU goes to sleep
193
         * For there was nothing to run, the CPU goes to sleep
194
         * until a hardware interrupt or an IPI comes.
194
         * until a hardware interrupt or an IPI comes.
195
         * This improves energy saving and hyperthreading.
195
         * This improves energy saving and hyperthreading.
196
         */
196
         */
197
 
197
 
198
        /*
198
        /*
199
         * An interrupt might occur right now and wake up a thread.
199
         * An interrupt might occur right now and wake up a thread.
200
         * In such case, the CPU will continue to go to sleep
200
         * In such case, the CPU will continue to go to sleep
201
         * even though there is a runnable thread.
201
         * even though there is a runnable thread.
202
         */
202
         */
203
 
203
 
204
         cpu_sleep();
204
         cpu_sleep();
205
         goto loop;
205
         goto loop;
206
    }
206
    }
207
 
207
 
208
    interrupts_disable();
208
    interrupts_disable();
209
   
209
   
210
    for (i = 0; i < RQ_COUNT; i++) {
210
    for (i = 0; i < RQ_COUNT; i++) {
211
        r = &CPU->rq[i];
211
        r = &CPU->rq[i];
212
        spinlock_lock(&r->lock);
212
        spinlock_lock(&r->lock);
213
        if (r->n == 0) {
213
        if (r->n == 0) {
214
            /*
214
            /*
215
             * If this queue is empty, try a lower-priority queue.
215
             * If this queue is empty, try a lower-priority queue.
216
             */
216
             */
217
            spinlock_unlock(&r->lock);
217
            spinlock_unlock(&r->lock);
218
            continue;
218
            continue;
219
        }
219
        }
220
 
220
 
221
        atomic_dec(&CPU->nrdy);
221
        atomic_dec(&CPU->nrdy);
222
        atomic_dec(&nrdy);
222
        atomic_dec(&nrdy);
223
        r->n--;
223
        r->n--;
224
 
224
 
225
        /*
225
        /*
226
         * Take the first thread from the queue.
226
         * Take the first thread from the queue.
227
         */
227
         */
228
        t = list_get_instance(r->rq_head.next, thread_t, rq_link);
228
        t = list_get_instance(r->rq_head.next, thread_t, rq_link);
229
        list_remove(&t->rq_link);
229
        list_remove(&t->rq_link);
230
 
230
 
231
        spinlock_unlock(&r->lock);
231
        spinlock_unlock(&r->lock);
232
 
232
 
233
        spinlock_lock(&t->lock);
233
        spinlock_lock(&t->lock);
234
        t->cpu = CPU;
234
        t->cpu = CPU;
235
 
235
 
236
        t->ticks = us2ticks((i + 1) * 10000);
236
        t->ticks = us2ticks((i + 1) * 10000);
237
        t->priority = i;    /* correct rq index */
237
        t->priority = i;    /* correct rq index */
238
 
238
 
239
        /*
239
        /*
240
         * Clear the THREAD_FLAG_STOLEN flag so that t can be migrated
240
         * Clear the THREAD_FLAG_STOLEN flag so that t can be migrated
241
         * when load balancing needs emerge.
241
         * when load balancing needs emerge.
242
         */
242
         */
243
        t->flags &= ~THREAD_FLAG_STOLEN;
243
        t->flags &= ~THREAD_FLAG_STOLEN;
244
        spinlock_unlock(&t->lock);
244
        spinlock_unlock(&t->lock);
245
 
245
 
246
        return t;
246
        return t;
247
    }
247
    }
248
    goto loop;
248
    goto loop;
249
 
249
 
250
}
250
}
251
 
251
 
252
/** Prevent rq starvation
252
/** Prevent rq starvation
253
 *
253
 *
254
 * Prevent low priority threads from starving in rq's.
254
 * Prevent low priority threads from starving in rq's.
255
 *
255
 *
256
 * When the function decides to relink rq's, it reconnects
256
 * When the function decides to relink rq's, it reconnects
257
 * respective pointers so that in result threads with 'pri'
257
 * respective pointers so that in result threads with 'pri'
258
 * greater or equal start are moved to a higher-priority queue.
258
 * greater or equal start are moved to a higher-priority queue.
259
 *
259
 *
260
 * @param start Threshold priority.
260
 * @param start Threshold priority.
261
 *
261
 *
262
 */
262
 */
263
static void relink_rq(int start)
263
static void relink_rq(int start)
264
{
264
{
265
    link_t head;
265
    link_t head;
266
    runq_t *r;
266
    runq_t *r;
267
    int i, n;
267
    int i, n;
268
 
268
 
269
    list_initialize(&head);
269
    list_initialize(&head);
270
    spinlock_lock(&CPU->lock);
270
    spinlock_lock(&CPU->lock);
271
    if (CPU->needs_relink > NEEDS_RELINK_MAX) {
271
    if (CPU->needs_relink > NEEDS_RELINK_MAX) {
272
        for (i = start; i < RQ_COUNT - 1; i++) {
272
        for (i = start; i < RQ_COUNT - 1; i++) {
273
            /* remember and empty rq[i + 1] */
273
            /* remember and empty rq[i + 1] */
274
            r = &CPU->rq[i + 1];
274
            r = &CPU->rq[i + 1];
275
            spinlock_lock(&r->lock);
275
            spinlock_lock(&r->lock);
276
            list_concat(&head, &r->rq_head);
276
            list_concat(&head, &r->rq_head);
277
            n = r->n;
277
            n = r->n;
278
            r->n = 0;
278
            r->n = 0;
279
            spinlock_unlock(&r->lock);
279
            spinlock_unlock(&r->lock);
280
       
280
       
281
            /* append rq[i + 1] to rq[i] */
281
            /* append rq[i + 1] to rq[i] */
282
            r = &CPU->rq[i];
282
            r = &CPU->rq[i];
283
            spinlock_lock(&r->lock);
283
            spinlock_lock(&r->lock);
284
            list_concat(&r->rq_head, &head);
284
            list_concat(&r->rq_head, &head);
285
            r->n += n;
285
            r->n += n;
286
            spinlock_unlock(&r->lock);
286
            spinlock_unlock(&r->lock);
287
        }
287
        }
288
        CPU->needs_relink = 0;
288
        CPU->needs_relink = 0;
289
    }
289
    }
290
    spinlock_unlock(&CPU->lock);
290
    spinlock_unlock(&CPU->lock);
291
 
291
 
292
}
292
}
293
 
293
 
294
/** The scheduler
294
/** The scheduler
295
 *
295
 *
296
 * The thread scheduling procedure.
296
 * The thread scheduling procedure.
297
 * Passes control directly to
297
 * Passes control directly to
298
 * scheduler_separated_stack().
298
 * scheduler_separated_stack().
299
 *
299
 *
300
 */
300
 */
301
void scheduler(void)
301
void scheduler(void)
302
{
302
{
303
    volatile ipl_t ipl;
303
    volatile ipl_t ipl;
304
 
304
 
305
    ASSERT(CPU != NULL);
305
    ASSERT(CPU != NULL);
306
 
306
 
307
    ipl = interrupts_disable();
307
    ipl = interrupts_disable();
308
 
308
 
309
    if (atomic_get(&haltstate))
309
    if (atomic_get(&haltstate))
310
        halt();
310
        halt();
311
   
311
   
312
    if (THREAD) {
312
    if (THREAD) {
313
        spinlock_lock(&THREAD->lock);
313
        spinlock_lock(&THREAD->lock);
314
       
314
       
315
        /* Update thread accounting */
315
        /* Update thread accounting */
316
        THREAD->cycles += get_cycle() - THREAD->last_cycle;
316
        THREAD->cycles += get_cycle() - THREAD->last_cycle;
317
       
317
       
318
#ifndef CONFIG_FPU_LAZY
318
#ifndef CONFIG_FPU_LAZY
319
        fpu_context_save(THREAD->saved_fpu_context);
319
        fpu_context_save(THREAD->saved_fpu_context);
320
#endif
320
#endif
321
        if (!context_save(&THREAD->saved_context)) {
321
        if (!context_save(&THREAD->saved_context)) {
322
            /*
322
            /*
323
             * This is the place where threads leave scheduler();
323
             * This is the place where threads leave scheduler();
324
             */
324
             */
325
           
325
           
326
            /* Save current CPU cycle */
326
            /* Save current CPU cycle */
327
            THREAD->last_cycle = get_cycle();
327
            THREAD->last_cycle = get_cycle();
328
           
328
           
329
            spinlock_unlock(&THREAD->lock);
329
            spinlock_unlock(&THREAD->lock);
330
            interrupts_restore(THREAD->saved_context.ipl);
330
            interrupts_restore(THREAD->saved_context.ipl);
331
           
331
           
332
            return;
332
            return;
333
        }
333
        }
334
 
334
 
335
        /*
335
        /*
336
         * Interrupt priority level of preempted thread is recorded
336
         * Interrupt priority level of preempted thread is recorded
337
         * here to facilitate scheduler() invocations from
337
         * here to facilitate scheduler() invocations from
338
         * interrupts_disable()'d code (e.g. waitq_sleep_timeout()).
338
         * interrupts_disable()'d code (e.g. waitq_sleep_timeout()).
339
         */
339
         */
340
        THREAD->saved_context.ipl = ipl;
340
        THREAD->saved_context.ipl = ipl;
341
    }
341
    }
342
 
342
 
343
    /*
343
    /*
344
     * Through the 'THE' structure, we keep track of THREAD, TASK, CPU, VM
344
     * Through the 'THE' structure, we keep track of THREAD, TASK, CPU, VM
345
     * and preemption counter. At this point THE could be coming either
345
     * and preemption counter. At this point THE could be coming either
346
     * from THREAD's or CPU's stack.
346
     * from THREAD's or CPU's stack.
347
     */
347
     */
348
    the_copy(THE, (the_t *) CPU->stack);
348
    the_copy(THE, (the_t *) CPU->stack);
349
 
349
 
350
    /*
350
    /*
351
     * We may not keep the old stack.
351
     * We may not keep the old stack.
352
     * Reason: If we kept the old stack and got blocked, for instance, in
352
     * Reason: If we kept the old stack and got blocked, for instance, in
353
     * find_best_thread(), the old thread could get rescheduled by another
353
     * find_best_thread(), the old thread could get rescheduled by another
354
     * CPU and overwrite the part of its own stack that was also used by
354
     * CPU and overwrite the part of its own stack that was also used by
355
     * the scheduler on this CPU.
355
     * the scheduler on this CPU.
356
     *
356
     *
357
     * Moreover, we have to bypass the compiler-generated POP sequence
357
     * Moreover, we have to bypass the compiler-generated POP sequence
358
     * which is fooled by SP being set to the very top of the stack.
358
     * which is fooled by SP being set to the very top of the stack.
359
     * Therefore the scheduler() function continues in
359
     * Therefore the scheduler() function continues in
360
     * scheduler_separated_stack().
360
     * scheduler_separated_stack().
361
     */
361
     */
362
    context_save(&CPU->saved_context);
362
    context_save(&CPU->saved_context);
363
    context_set(&CPU->saved_context, FADDR(scheduler_separated_stack),
363
    context_set(&CPU->saved_context, FADDR(scheduler_separated_stack),
364
        (uintptr_t) CPU->stack, CPU_STACK_SIZE);
364
        (uintptr_t) CPU->stack, CPU_STACK_SIZE);
365
    context_restore(&CPU->saved_context);
365
    context_restore(&CPU->saved_context);
366
    /* not reached */
366
    /* not reached */
367
}
367
}
368
 
368
 
369
/** Scheduler stack switch wrapper
369
/** Scheduler stack switch wrapper
370
 *
370
 *
371
 * Second part of the scheduler() function
371
 * Second part of the scheduler() function
372
 * using new stack. Handling the actual context
372
 * using new stack. Handling the actual context
373
 * switch to a new thread.
373
 * switch to a new thread.
374
 *
374
 *
375
 * Assume THREAD->lock is held.
375
 * Assume THREAD->lock is held.
376
 */
376
 */
377
void scheduler_separated_stack(void)
377
void scheduler_separated_stack(void)
378
{
378
{
379
    int priority;
379
    int priority;
380
    DEADLOCK_PROBE_INIT(p_joinwq);
380
    DEADLOCK_PROBE_INIT(p_joinwq);
381
 
381
 
382
    ASSERT(CPU != NULL);
382
    ASSERT(CPU != NULL);
383
   
383
   
384
    if (THREAD) {
384
    if (THREAD) {
385
        /* must be run after the switch to scheduler stack */
385
        /* must be run after the switch to scheduler stack */
386
        after_thread_ran();
386
        after_thread_ran();
387
 
387
 
388
        switch (THREAD->state) {
388
        switch (THREAD->state) {
389
        case Running:
389
        case Running:
390
            spinlock_unlock(&THREAD->lock);
390
            spinlock_unlock(&THREAD->lock);
391
            thread_ready(THREAD);
391
            thread_ready(THREAD);
392
            break;
392
            break;
393
 
393
 
394
        case Exiting:
394
        case Exiting:
395
repeat:
395
repeat:
396
            if (THREAD->detached) {
396
            if (THREAD->detached) {
397
                thread_destroy(THREAD);
397
                thread_destroy(THREAD);
398
            } else {
398
            } else {
399
                /*
399
                /*
400
                 * The thread structure is kept allocated until
400
                 * The thread structure is kept allocated until
401
                 * somebody calls thread_detach() on it.
401
                 * somebody calls thread_detach() on it.
402
                 */
402
                 */
403
                if (!spinlock_trylock(&THREAD->join_wq.lock)) {
403
                if (!spinlock_trylock(&THREAD->join_wq.lock)) {
404
                    /*
404
                    /*
405
                     * Avoid deadlock.
405
                     * Avoid deadlock.
406
                     */
406
                     */
407
                    spinlock_unlock(&THREAD->lock);
407
                    spinlock_unlock(&THREAD->lock);
408
                    delay(HZ);
408
                    delay(HZ);
409
                    spinlock_lock(&THREAD->lock);
409
                    spinlock_lock(&THREAD->lock);
410
                    DEADLOCK_PROBE(p_joinwq,
410
                    DEADLOCK_PROBE(p_joinwq,
411
                        DEADLOCK_THRESHOLD);
411
                        DEADLOCK_THRESHOLD);
412
                    goto repeat;
412
                    goto repeat;
413
                }
413
                }
414
                _waitq_wakeup_unsafe(&THREAD->join_wq,
414
                _waitq_wakeup_unsafe(&THREAD->join_wq,
415
                    WAKEUP_FIRST);
415
                    WAKEUP_FIRST);
416
                spinlock_unlock(&THREAD->join_wq.lock);
416
                spinlock_unlock(&THREAD->join_wq.lock);
417
               
417
               
418
                THREAD->state = Lingering;
418
                THREAD->state = Lingering;
419
                spinlock_unlock(&THREAD->lock);
419
                spinlock_unlock(&THREAD->lock);
420
            }
420
            }
421
            break;
421
            break;
422
           
422
           
423
        case Sleeping:
423
        case Sleeping:
424
            /*
424
            /*
425
             * Prefer the thread after it's woken up.
425
             * Prefer the thread after it's woken up.
426
             */
426
             */
427
            THREAD->priority = -1;
427
            THREAD->priority = -1;
428
 
428
 
429
            /*
429
            /*
430
             * We need to release wq->lock which we locked in
430
             * We need to release wq->lock which we locked in
431
             * waitq_sleep(). Address of wq->lock is kept in
431
             * waitq_sleep(). Address of wq->lock is kept in
432
             * THREAD->sleep_queue.
432
             * THREAD->sleep_queue.
433
             */
433
             */
434
            spinlock_unlock(&THREAD->sleep_queue->lock);
434
            spinlock_unlock(&THREAD->sleep_queue->lock);
435
 
435
 
436
            /*
436
            /*
437
             * Check for possible requests for out-of-context
437
             * Check for possible requests for out-of-context
438
             * invocation.
438
             * invocation.
439
             */
439
             */
440
            if (THREAD->call_me) {
440
            if (THREAD->call_me) {
441
                THREAD->call_me(THREAD->call_me_with);
441
                THREAD->call_me(THREAD->call_me_with);
442
                THREAD->call_me = NULL;
442
                THREAD->call_me = NULL;
443
                THREAD->call_me_with = NULL;
443
                THREAD->call_me_with = NULL;
444
            }
444
            }
445
 
445
 
446
            spinlock_unlock(&THREAD->lock);
446
            spinlock_unlock(&THREAD->lock);
447
 
447
 
448
            break;
448
            break;
449
 
449
 
450
        default:
450
        default:
451
            /*
451
            /*
452
             * Entering state is unexpected.
452
             * Entering state is unexpected.
453
             */
453
             */
454
            panic("tid%" PRIu64 ": unexpected state %s\n", THREAD->tid,
454
            panic("tid%" PRIu64 ": unexpected state %s\n",
455
                thread_states[THREAD->state]);
455
                THREAD->tid, thread_states[THREAD->state]);
456
            break;
456
            break;
457
        }
457
        }
458
 
458
 
459
        THREAD = NULL;
459
        THREAD = NULL;
460
    }
460
    }
461
 
461
 
462
    THREAD = find_best_thread();
462
    THREAD = find_best_thread();
463
   
463
   
464
    spinlock_lock(&THREAD->lock);
464
    spinlock_lock(&THREAD->lock);
465
    priority = THREAD->priority;
465
    priority = THREAD->priority;
466
    spinlock_unlock(&THREAD->lock);
466
    spinlock_unlock(&THREAD->lock);
467
 
467
 
468
    relink_rq(priority);       
468
    relink_rq(priority);       
469
 
469
 
470
    /*
470
    /*
471
     * If both the old and the new task are the same, lots of work is
471
     * If both the old and the new task are the same, lots of work is
472
     * avoided.
472
     * avoided.
473
     */
473
     */
474
    if (TASK != THREAD->task) {
474
    if (TASK != THREAD->task) {
475
        as_t *as1 = NULL;
475
        as_t *as1 = NULL;
476
        as_t *as2;
476
        as_t *as2;
477
 
477
 
478
        if (TASK) {
478
        if (TASK) {
479
            spinlock_lock(&TASK->lock);
479
            spinlock_lock(&TASK->lock);
480
            as1 = TASK->as;
480
            as1 = TASK->as;
481
            spinlock_unlock(&TASK->lock);
481
            spinlock_unlock(&TASK->lock);
482
        }
482
        }
483
 
483
 
484
        spinlock_lock(&THREAD->task->lock);
484
        spinlock_lock(&THREAD->task->lock);
485
        as2 = THREAD->task->as;
485
        as2 = THREAD->task->as;
486
        spinlock_unlock(&THREAD->task->lock);
486
        spinlock_unlock(&THREAD->task->lock);
487
       
487
       
488
        /*
488
        /*
489
         * Note that it is possible for two tasks to share one address
489
         * Note that it is possible for two tasks to share one address
490
         * space.
490
         * space.
491
         */
491
         */
492
        if (as1 != as2) {
492
        if (as1 != as2) {
493
            /*
493
            /*
494
             * Both tasks and address spaces are different.
494
             * Both tasks and address spaces are different.
495
             * Replace the old one with the new one.
495
             * Replace the old one with the new one.
496
             */
496
             */
497
            as_switch(as1, as2);
497
            as_switch(as1, as2);
498
        }
498
        }
499
        TASK = THREAD->task;
499
        TASK = THREAD->task;
500
        before_task_runs();
500
        before_task_runs();
501
    }
501
    }
502
 
502
 
503
    spinlock_lock(&THREAD->lock);  
503
    spinlock_lock(&THREAD->lock);  
504
    THREAD->state = Running;
504
    THREAD->state = Running;
505
 
505
 
506
#ifdef SCHEDULER_VERBOSE
506
#ifdef SCHEDULER_VERBOSE
507
    printf("cpu%u: tid %" PRIu64 " (priority=%d, ticks=%" PRIu64 ", nrdy=%ld)\n",
507
    printf("cpu%u: tid %" PRIu64 " (priority=%d, ticks=%" PRIu64
508
        CPU->id, THREAD->tid, THREAD->priority, THREAD->ticks,
508
        ", nrdy=%ld)\n", CPU->id, THREAD->tid, THREAD->priority,
509
        atomic_get(&CPU->nrdy));
509
        THREAD->ticks, atomic_get(&CPU->nrdy));
510
#endif  
510
#endif  
511
 
511
 
512
    /*
512
    /*
513
     * Some architectures provide late kernel PA2KA(identity)
513
     * Some architectures provide late kernel PA2KA(identity)
514
     * mapping in a page fault handler. However, the page fault
514
     * mapping in a page fault handler. However, the page fault
515
     * handler uses the kernel stack of the running thread and
515
     * handler uses the kernel stack of the running thread and
516
     * therefore cannot be used to map it. The kernel stack, if
516
     * therefore cannot be used to map it. The kernel stack, if
517
     * necessary, is to be mapped in before_thread_runs(). This
517
     * necessary, is to be mapped in before_thread_runs(). This
518
     * function must be executed before the switch to the new stack.
518
     * function must be executed before the switch to the new stack.
519
     */
519
     */
520
    before_thread_runs();
520
    before_thread_runs();
521
 
521
 
522
    /*
522
    /*
523
     * Copy the knowledge of CPU, TASK, THREAD and preemption counter to
523
     * Copy the knowledge of CPU, TASK, THREAD and preemption counter to
524
     * thread's stack.
524
     * thread's stack.
525
     */
525
     */
526
    the_copy(THE, (the_t *) THREAD->kstack);
526
    the_copy(THE, (the_t *) THREAD->kstack);
527
   
527
   
528
    context_restore(&THREAD->saved_context);
528
    context_restore(&THREAD->saved_context);
529
    /* not reached */
529
    /* not reached */
530
}
530
}
531
 
531
 
532
#ifdef CONFIG_SMP
532
#ifdef CONFIG_SMP
533
/** Load balancing thread
533
/** Load balancing thread
534
 *
534
 *
535
 * SMP load balancing thread, supervising thread supplies
535
 * SMP load balancing thread, supervising thread supplies
536
 * for the CPU it's wired to.
536
 * for the CPU it's wired to.
537
 *
537
 *
538
 * @param arg Generic thread argument (unused).
538
 * @param arg Generic thread argument (unused).
539
 *
539
 *
540
 */
540
 */
541
void kcpulb(void *arg)
541
void kcpulb(void *arg)
542
{
542
{
543
    thread_t *t;
543
    thread_t *t;
544
    int count, average, j, k = 0;
544
    int count, average, j, k = 0;
545
    unsigned int i;
545
    unsigned int i;
546
    ipl_t ipl;
546
    ipl_t ipl;
547
 
547
 
548
    /*
548
    /*
549
     * Detach kcpulb as nobody will call thread_join_timeout() on it.
549
     * Detach kcpulb as nobody will call thread_join_timeout() on it.
550
     */
550
     */
551
    thread_detach(THREAD);
551
    thread_detach(THREAD);
552
   
552
   
553
loop:
553
loop:
554
    /*
554
    /*
555
     * Work in 1s intervals.
555
     * Work in 1s intervals.
556
     */
556
     */
557
    thread_sleep(1);
557
    thread_sleep(1);
558
 
558
 
559
not_satisfied:
559
not_satisfied:
560
    /*
560
    /*
561
     * Calculate the number of threads that will be migrated/stolen from
561
     * Calculate the number of threads that will be migrated/stolen from
562
     * other CPU's. Note that situation can have changed between two
562
     * other CPU's. Note that situation can have changed between two
563
     * passes. Each time get the most up to date counts.
563
     * passes. Each time get the most up to date counts.
564
     */
564
     */
565
    average = atomic_get(&nrdy) / config.cpu_active + 1;
565
    average = atomic_get(&nrdy) / config.cpu_active + 1;
566
    count = average - atomic_get(&CPU->nrdy);
566
    count = average - atomic_get(&CPU->nrdy);
567
 
567
 
568
    if (count <= 0)
568
    if (count <= 0)
569
        goto satisfied;
569
        goto satisfied;
570
 
570
 
571
    /*
571
    /*
572
     * Searching least priority queues on all CPU's first and most priority
572
     * Searching least priority queues on all CPU's first and most priority
573
     * queues on all CPU's last.
573
     * queues on all CPU's last.
574
     */
574
     */
575
    for (j = RQ_COUNT - 1; j >= 0; j--) {
575
    for (j = RQ_COUNT - 1; j >= 0; j--) {
576
        for (i = 0; i < config.cpu_active; i++) {
576
        for (i = 0; i < config.cpu_active; i++) {
577
            link_t *l;
577
            link_t *l;
578
            runq_t *r;
578
            runq_t *r;
579
            cpu_t *cpu;
579
            cpu_t *cpu;
580
 
580
 
581
            cpu = &cpus[(i + k) % config.cpu_active];
581
            cpu = &cpus[(i + k) % config.cpu_active];
582
 
582
 
583
            /*
583
            /*
584
             * Not interested in ourselves.
584
             * Not interested in ourselves.
585
             * Doesn't require interrupt disabling for kcpulb has
585
             * Doesn't require interrupt disabling for kcpulb has
586
             * THREAD_FLAG_WIRED.
586
             * THREAD_FLAG_WIRED.
587
             */
587
             */
588
            if (CPU == cpu)
588
            if (CPU == cpu)
589
                continue;
589
                continue;
590
            if (atomic_get(&cpu->nrdy) <= average)
590
            if (atomic_get(&cpu->nrdy) <= average)
591
                continue;
591
                continue;
592
 
592
 
593
            ipl = interrupts_disable();
593
            ipl = interrupts_disable();
594
            r = &cpu->rq[j];
594
            r = &cpu->rq[j];
595
            spinlock_lock(&r->lock);
595
            spinlock_lock(&r->lock);
596
            if (r->n == 0) {
596
            if (r->n == 0) {
597
                spinlock_unlock(&r->lock);
597
                spinlock_unlock(&r->lock);
598
                interrupts_restore(ipl);
598
                interrupts_restore(ipl);
599
                continue;
599
                continue;
600
            }
600
            }
601
       
601
       
602
            t = NULL;
602
            t = NULL;
603
            l = r->rq_head.prev;    /* search rq from the back */
603
            l = r->rq_head.prev;    /* search rq from the back */
604
            while (l != &r->rq_head) {
604
            while (l != &r->rq_head) {
605
                t = list_get_instance(l, thread_t, rq_link);
605
                t = list_get_instance(l, thread_t, rq_link);
606
                /*
606
                /*
607
                 * We don't want to steal CPU-wired threads
607
                 * We don't want to steal CPU-wired threads
608
                 * neither threads already stolen. The latter
608
                 * neither threads already stolen. The latter
609
                 * prevents threads from migrating between CPU's
609
                 * prevents threads from migrating between CPU's
610
                 * without ever being run. We don't want to
610
                 * without ever being run. We don't want to
611
                 * steal threads whose FPU context is still in
611
                 * steal threads whose FPU context is still in
612
                 * CPU.
612
                 * CPU.
613
                 */
613
                 */
614
                spinlock_lock(&t->lock);
614
                spinlock_lock(&t->lock);
615
                if ((!(t->flags & (THREAD_FLAG_WIRED |
615
                if ((!(t->flags & (THREAD_FLAG_WIRED |
616
                    THREAD_FLAG_STOLEN))) &&
616
                    THREAD_FLAG_STOLEN))) &&
617
                    (!(t->fpu_context_engaged))) {
617
                    (!(t->fpu_context_engaged))) {
618
                    /*
618
                    /*
619
                     * Remove t from r.
619
                     * Remove t from r.
620
                     */
620
                     */
621
                    spinlock_unlock(&t->lock);
621
                    spinlock_unlock(&t->lock);
622
                   
622
                   
623
                    atomic_dec(&cpu->nrdy);
623
                    atomic_dec(&cpu->nrdy);
624
                    atomic_dec(&nrdy);
624
                    atomic_dec(&nrdy);
625
 
625
 
626
                    r->n--;
626
                    r->n--;
627
                    list_remove(&t->rq_link);
627
                    list_remove(&t->rq_link);
628
 
628
 
629
                    break;
629
                    break;
630
                }
630
                }
631
                spinlock_unlock(&t->lock);
631
                spinlock_unlock(&t->lock);
632
                l = l->prev;
632
                l = l->prev;
633
                t = NULL;
633
                t = NULL;
634
            }
634
            }
635
            spinlock_unlock(&r->lock);
635
            spinlock_unlock(&r->lock);
636
 
636
 
637
            if (t) {
637
            if (t) {
638
                /*
638
                /*
639
                 * Ready t on local CPU
639
                 * Ready t on local CPU
640
                 */
640
                 */
641
                spinlock_lock(&t->lock);
641
                spinlock_lock(&t->lock);
642
#ifdef KCPULB_VERBOSE
642
#ifdef KCPULB_VERBOSE
643
                printf("kcpulb%u: TID %" PRIu64 " -> cpu%u, nrdy=%ld, "
643
                printf("kcpulb%u: TID %" PRIu64 " -> cpu%u, "
644
                    "avg=%ld\n", CPU->id, t->tid, CPU->id,
644
                    "nrdy=%ld, avg=%ld\n", CPU->id, t->tid,
645
                    atomic_get(&CPU->nrdy),
645
                    CPU->id, atomic_get(&CPU->nrdy),
646
                    atomic_get(&nrdy) / config.cpu_active);
646
                    atomic_get(&nrdy) / config.cpu_active);
647
#endif
647
#endif
648
                t->flags |= THREAD_FLAG_STOLEN;
648
                t->flags |= THREAD_FLAG_STOLEN;
649
                t->state = Entering;
649
                t->state = Entering;
650
                spinlock_unlock(&t->lock);
650
                spinlock_unlock(&t->lock);
651
   
651
   
652
                thread_ready(t);
652
                thread_ready(t);
653
 
653
 
654
                interrupts_restore(ipl);
654
                interrupts_restore(ipl);
655
   
655
   
656
                if (--count == 0)
656
                if (--count == 0)
657
                    goto satisfied;
657
                    goto satisfied;
658
                   
658
                   
659
                /*
659
                /*
660
                 * We are not satisfied yet, focus on another
660
                 * We are not satisfied yet, focus on another
661
                 * CPU next time.
661
                 * CPU next time.
662
                 */
662
                 */
663
                k++;
663
                k++;
664
               
664
               
665
                continue;
665
                continue;
666
            }
666
            }
667
            interrupts_restore(ipl);
667
            interrupts_restore(ipl);
668
        }
668
        }
669
    }
669
    }
670
 
670
 
671
    if (atomic_get(&CPU->nrdy)) {
671
    if (atomic_get(&CPU->nrdy)) {
672
        /*
672
        /*
673
         * Be a little bit light-weight and let migrated threads run.
673
         * Be a little bit light-weight and let migrated threads run.
674
         */
674
         */
675
        scheduler();
675
        scheduler();
676
    } else {
676
    } else {
677
        /*
677
        /*
678
         * We failed to migrate a single thread.
678
         * We failed to migrate a single thread.
679
         * Give up this turn.
679
         * Give up this turn.
680
         */
680
         */
681
        goto loop;
681
        goto loop;
682
    }
682
    }
683
       
683
       
684
    goto not_satisfied;
684
    goto not_satisfied;
685
 
685
 
686
satisfied:
686
satisfied:
687
    goto loop;
687
    goto loop;
688
}
688
}
689
 
689
 
690
#endif /* CONFIG_SMP */
690
#endif /* CONFIG_SMP */
691
 
691
 
692
 
692
 
693
/** Print information about threads & scheduler queues */
693
/** Print information about threads & scheduler queues */
694
void sched_print_list(void)
694
void sched_print_list(void)
695
{
695
{
696
    ipl_t ipl;
696
    ipl_t ipl;
697
    unsigned int cpu, i;
697
    unsigned int cpu, i;
698
    runq_t *r;
698
    runq_t *r;
699
    thread_t *t;
699
    thread_t *t;
700
    link_t *cur;
700
    link_t *cur;
701
 
701
 
702
    /* We are going to mess with scheduler structures,
702
    /* We are going to mess with scheduler structures,
703
     * let's not be interrupted */
703
     * let's not be interrupted */
704
    ipl = interrupts_disable();
704
    ipl = interrupts_disable();
705
    for (cpu = 0; cpu < config.cpu_count; cpu++) {
705
    for (cpu = 0; cpu < config.cpu_count; cpu++) {
706
 
706
 
707
        if (!cpus[cpu].active)
707
        if (!cpus[cpu].active)
708
            continue;
708
            continue;
709
 
709
 
710
        spinlock_lock(&cpus[cpu].lock);
710
        spinlock_lock(&cpus[cpu].lock);
711
        printf("cpu%u: address=%p, nrdy=%ld, needs_relink=%" PRIc "\n",
711
        printf("cpu%u: address=%p, nrdy=%ld, needs_relink=%" PRIc "\n",
712
            cpus[cpu].id, &cpus[cpu], atomic_get(&cpus[cpu].nrdy),
712
            cpus[cpu].id, &cpus[cpu], atomic_get(&cpus[cpu].nrdy),
713
            cpus[cpu].needs_relink);
713
            cpus[cpu].needs_relink);
714
       
714
       
715
        for (i = 0; i < RQ_COUNT; i++) {
715
        for (i = 0; i < RQ_COUNT; i++) {
716
            r = &cpus[cpu].rq[i];
716
            r = &cpus[cpu].rq[i];
717
            spinlock_lock(&r->lock);
717
            spinlock_lock(&r->lock);
718
            if (!r->n) {
718
            if (!r->n) {
719
                spinlock_unlock(&r->lock);
719
                spinlock_unlock(&r->lock);
720
                continue;
720
                continue;
721
            }
721
            }
722
            printf("\trq[%u]: ", i);
722
            printf("\trq[%u]: ", i);
723
            for (cur = r->rq_head.next; cur != &r->rq_head;
723
            for (cur = r->rq_head.next; cur != &r->rq_head;
724
                cur = cur->next) {
724
                cur = cur->next) {
725
                t = list_get_instance(cur, thread_t, rq_link);
725
                t = list_get_instance(cur, thread_t, rq_link);
726
                printf("%" PRIu64 "(%s) ", t->tid,
726
                printf("%" PRIu64 "(%s) ", t->tid,
727
                    thread_states[t->state]);
727
                    thread_states[t->state]);
728
            }
728
            }
729
            printf("\n");
729
            printf("\n");
730
            spinlock_unlock(&r->lock);
730
            spinlock_unlock(&r->lock);
731
        }
731
        }
732
        spinlock_unlock(&cpus[cpu].lock);
732
        spinlock_unlock(&cpus[cpu].lock);
733
    }
733
    }
734
   
734
   
735
    interrupts_restore(ipl);
735
    interrupts_restore(ipl);
736
}
736
}
737
 
737
 
738
/** @}
738
/** @}
739
 */
739
 */
740
 
740