aboutsummaryrefslogtreecommitdiffstats
path: root/test/testbmk.c
blob: 5e5b11e7db310a5b9e3bc6dd6dd7ea83ec0af27d (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
/*
    ChibiOS/RT - Copyright (C) 2006,2007,2008,2009,2010,
                 2011,2012 Giovanni Di Sirio.

    This file is part of ChibiOS/RT.

    ChibiOS/RT is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation; either version 3 of the License, or
    (at your option) any later version.

    ChibiOS/RT is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with this program.  If not, see <http://www.gnu.org/licenses/>.
*/

#include "ch.h"
#include "test.h"

/**
 * @page test_benchmarks Kernel Benchmarks
 *
 * File: @ref testbmk.c
 *
 * <h2>Description</h2>
 * This module implements a series of system benchmarks. The benchmarks are
 * useful as a stress test and as a reference when comparing ChibiOS/RT
 * with similar systems.
 *
 * <h2>Objective</h2>
 * Objective of the test module is to provide a performance index for the
 * most critical system subsystems. The performance numbers allow to
 * discover performance regressions between successive ChibiOS/RT releases.
 *
 * <h2>Preconditions</h2>
 * None.
 *
 * <h2>Test Cases</h2>
 * - @subpage test_benchmarks_001
 * - @subpage test_benchmarks_002
 * - @subpage test_benchmarks_003
 * - @subpage test_benchmarks_004
 * - @subpage test_benchmarks_005
 * - @subpage test_benchmarks_006
 * - @subpage test_benchmarks_007
 * - @subpage test_benchmarks_008
 * - @subpage test_benchmarks_009
 * - @subpage test_benchmarks_010
 * - @subpage test_benchmarks_011
 * - @subpage test_benchmarks_012
 * - @subpage test_benchmarks_013
 * .
 * @file testbmk.c Kernel Benchmarks
 * @brief Kernel Benchmarks source file
 * @file testbmk.h
 * @brief Kernel Benchmarks header file
 */

static Semaphore sem1;
#if CH_USE_MUTEXES || defined(__DOXYGEN__)
static Mutex mtx1;
#endif

static msg_t thread1(void *p) {
  Thread *tp;
  msg_t msg;

  (void)p;
  do {
    tp = chMsgWait();
    msg = chMsgGet(tp);
    chMsgRelease(tp, msg);
  } while (msg);
  return 0;
}

#ifdef __GNUC__
__attribute__((noinline))
#endif
static unsigned int msg_loop_test(Thread *tp) {

  uint32_t n = 0;
  test_wait_tick();
  test_start_timer(1000);
  do {
    (void)chMsgSend(tp, 1);
    n++;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  (void)chMsgSend(tp, 0);
  return n;
}

/**
 * @page test_benchmarks_001 Messages performance #1
 *
 * <h2>Description</h2>
 * A message server thread is created with a lower priority than the client
 * thread, the messages throughput per second is measured and the result
 * printed in the output log.
 */

static void bmk1_execute(void) {
  uint32_t n;

  threads[0] = chThdCreateStatic(wa[0], WA_SIZE, chThdGetPriority()-1, thread1, NULL);
  n = msg_loop_test(threads[0]);
  test_wait_threads();
  test_print("--- Score : ");
  test_printn(n);
  test_print(" msgs/S, ");
  test_printn(n << 1);
  test_println(" ctxswc/S");
}

ROMCONST struct testcase testbmk1 = {
  "Benchmark, messages #1",
  NULL,
  NULL,
  bmk1_execute
};

/**
 * @page test_benchmarks_002 Messages performance #2
 *
 * <h2>Description</h2>
 * A message server thread is created with an higher priority than the client
 * thread, the messages throughput per second is measured and the result
 * printed in the output log.
 */

static void bmk2_execute(void) {
  uint32_t n;

  threads[0] = chThdCreateStatic(wa[0], WA_SIZE, chThdGetPriority()+1, thread1, NULL);
  n = msg_loop_test(threads[0]);
  test_wait_threads();
  test_print("--- Score : ");
  test_printn(n);
  test_print(" msgs/S, ");
  test_printn(n << 1);
  test_println(" ctxswc/S");
}

ROMCONST struct testcase testbmk2 = {
  "Benchmark, messages #2",
  NULL,
  NULL,
  bmk2_execute
};

static msg_t thread2(void *p) {

  return (msg_t)p;
}

/**
 * @page test_benchmarks_003 Messages performance #3
 *
 * <h2>Description</h2>
 * A message server thread is created with an higher priority than the client
 * thread, four lower priority threads crowd the ready list, the messages
 * throughput per second is measured while the ready list and the result
 * printed in the output log.
 */

static void bmk3_execute(void) {
  uint32_t n;

  threads[0] = chThdCreateStatic(wa[0], WA_SIZE, chThdGetPriority()+1, thread1, NULL);
  threads[1] = chThdCreateStatic(wa[1], WA_SIZE, chThdGetPriority()-2, thread2, NULL);
  threads[2] = chThdCreateStatic(wa[2], WA_SIZE, chThdGetPriority()-3, thread2, NULL);
  threads[3] = chThdCreateStatic(wa[3], WA_SIZE, chThdGetPriority()-4, thread2, NULL);
  threads[4] = chThdCreateStatic(wa[4], WA_SIZE, chThdGetPriority()-5, thread2, NULL);
  n = msg_loop_test(threads[0]);
  test_wait_threads();
  test_print("--- Score : ");
  test_printn(n);
  test_print(" msgs/S, ");
  test_printn(n << 1);
  test_println(" ctxswc/S");
}

ROMCONST struct testcase testbmk3 = {
  "Benchmark, messages #3",
  NULL,
  NULL,
  bmk3_execute
};

/**
 * @page test_benchmarks_004 Context Switch performance
 *
 * <h2>Description</h2>
 * A thread is created that just performs a @p chSchGoSleepS() into a loop,
 * the thread is awakened as fast is possible by the tester thread.<br>
 * The Context Switch performance is calculated by measuring the number of
 * iterations after a second of continuous operations.
 */

msg_t thread4(void *p) {
  msg_t msg;
  Thread *self = chThdSelf();

  (void)p;
  chSysLock();
  do {
    chSchGoSleepS(THD_STATE_SUSPENDED);
    msg = self->p_u.rdymsg;
  } while (msg == RDY_OK);
  chSysUnlock();
  return 0;
}

static void bmk4_execute(void) {
  Thread *tp;
  uint32_t n;

  tp = threads[0] = chThdCreateStatic(wa[0], WA_SIZE, chThdGetPriority()+1, thread4, NULL);
  n = 0;
  test_wait_tick();
  test_start_timer(1000);
  do {
    chSysLock();
    chSchWakeupS(tp, RDY_OK);
    chSchWakeupS(tp, RDY_OK);
    chSchWakeupS(tp, RDY_OK);
    chSchWakeupS(tp, RDY_OK);
    chSysUnlock();
    n += 4;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  chSysLock();
  chSchWakeupS(tp, RDY_TIMEOUT);
  chSysUnlock();

  test_wait_threads();
  test_print("--- Score : ");
  test_printn(n * 2);
  test_println(" ctxswc/S");
}

ROMCONST struct testcase testbmk4 = {
  "Benchmark, context switch",
  NULL,
  NULL,
  bmk4_execute
};

/**
 * @page test_benchmarks_005 Threads performance, full cycle
 *
 * <h2>Description</h2>
 * Threads are continuously created and terminated into a loop. A full
 * @p chThdCreateStatic() / @p chThdExit() / @p chThdWait() cycle is performed
 * in each iteration.<br>
 * The performance is calculated by measuring the number of iterations after
 * a second of continuous operations.
 */

static void bmk5_execute(void) {

  uint32_t n = 0;
  void *wap = wa[0];
  tprio_t prio = chThdGetPriority() - 1;
  test_wait_tick();
  test_start_timer(1000);
  do {
    chThdWait(chThdCreateStatic(wap, WA_SIZE, prio, thread2, NULL));
    n++;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  test_print("--- Score : ");
  test_printn(n);
  test_println(" threads/S");
}

ROMCONST struct testcase testbmk5 = {
  "Benchmark, threads, full cycle",
  NULL,
  NULL,
  bmk5_execute
};

/**
 * @page test_benchmarks_006 Threads performance, create/exit only
 *
 * <h2>Description</h2>
 * Threads are continuously created and terminated into a loop. A partial
 * @p chThdCreateStatic() / @p chThdExit() cycle is performed in each
 * iteration, the @p chThdWait() is not necessary because the thread is
 * created at an higher priority so there is no need to wait for it to
 * terminate.<br>
 * The performance is calculated by measuring the number of iterations after
 * a second of continuous operations.
 */

static void bmk6_execute(void) {

  uint32_t n = 0;
  void *wap = wa[0];
  tprio_t prio = chThdGetPriority() + 1;
  test_wait_tick();
  test_start_timer(1000);
  do {
    chThdCreateStatic(wap, WA_SIZE, prio, thread2, NULL);
    n++;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  test_print("--- Score : ");
  test_printn(n);
  test_println(" threads/S");
}

ROMCONST struct testcase testbmk6 = {
  "Benchmark, threads, create only",
  NULL,
  NULL,
  bmk6_execute
};

/**
 * @page test_benchmarks_007 Mass reschedule performance
 *
 * <h2>Description</h2>
 * Five threads are created and atomically rescheduled by resetting the
 * semaphore where they are waiting on. The operation is performed into a
 * continuous loop.<br>
 * The performance is calculated by measuring the number of iterations after
 * a second of continuous operations.
 */

static msg_t thread3(void *p) {

  (void)p;
  while (!chThdShouldTerminate())
    chSemWait(&sem1);
  return 0;
}

static void bmk7_setup(void) {

  chSemInit(&sem1, 0);
}

static void bmk7_execute(void) {
  uint32_t n;

  threads[0] = chThdCreateStatic(wa[0], WA_SIZE, chThdGetPriority()+5, thread3, NULL);
  threads[1] = chThdCreateStatic(wa[1], WA_SIZE, chThdGetPriority()+4, thread3, NULL);
  threads[2] = chThdCreateStatic(wa[2], WA_SIZE, chThdGetPriority()+3, thread3, NULL);
  threads[3] = chThdCreateStatic(wa[3], WA_SIZE, chThdGetPriority()+2, thread3, NULL);
  threads[4] = chThdCreateStatic(wa[4], WA_SIZE, chThdGetPriority()+1, thread3, NULL);

  n = 0;
  test_wait_tick();
  test_start_timer(1000);
  do {
    chSemReset(&sem1, 0);
    n++;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  test_terminate_threads();
  chSemReset(&sem1, 0);
  test_wait_threads();

  test_print("--- Score : ");
  test_printn(n);
  test_print(" reschedules/S, ");
  test_printn(n * 6);
  test_println(" ctxswc/S");
}

ROMCONST struct testcase testbmk7 = {
  "Benchmark, mass reschedule, 5 threads",
  bmk7_setup,
  NULL,
  bmk7_execute
};

/**
 * @page test_benchmarks_008 I/O Round-Robin voluntary reschedule.
 *
 * <h2>Description</h2>
 * Five threads are created at equal priority, each thread just increases a
 * variable and yields.<br>
 * The performance is calculated by measuring the number of iterations after
 * a second of continuous operations.
 */

static msg_t thread8(void *p) {

  do {
    chThdYield();
    chThdYield();
    chThdYield();
    chThdYield();
    (*(uint32_t *)p) += 4;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while(!chThdShouldTerminate());
  return 0;
}

static void bmk8_execute(void) {
  uint32_t n;

  n = 0;
  test_wait_tick();

  threads[0] = chThdCreateStatic(wa[0], WA_SIZE, chThdGetPriority()-1, thread8, (void *)&n);
  threads[1] = chThdCreateStatic(wa[1], WA_SIZE, chThdGetPriority()-1, thread8, (void *)&n);
  threads[2] = chThdCreateStatic(wa[2], WA_SIZE, chThdGetPriority()-1, thread8, (void *)&n);
  threads[3] = chThdCreateStatic(wa[3], WA_SIZE, chThdGetPriority()-1, thread8, (void *)&n);
  threads[4] = chThdCreateStatic(wa[4], WA_SIZE, chThdGetPriority()-1, thread8, (void *)&n);

  chThdSleepSeconds(1);
  test_terminate_threads();
  test_wait_threads();

  test_print("--- Score : ");
  test_printn(n);
  test_println(" ctxswc/S");
}

ROMCONST struct testcase testbmk8 = {
  "Benchmark, round robin context switching",
  NULL,
  NULL,
  bmk8_execute
};

/**
 * @page test_benchmarks_009 I/O Queues throughput
 *
 * <h2>Description</h2>
 * Four bytes are written and then read from an @p InputQueue into a continuous
 * loop.<br>
 * The performance is calculated by measuring the number of iterations after
 * a second of continuous operations.
 */

static void bmk9_execute(void) {
  uint32_t n;
  static uint8_t ib[16];
  static InputQueue iq;

  chIQInit(&iq, ib, sizeof(ib), NULL, NULL);
  n = 0;
  test_wait_tick();
  test_start_timer(1000);
  do {
    chSysLock();
    chIQPutI(&iq, 0);
    chIQPutI(&iq, 1);
    chIQPutI(&iq, 2);
    chIQPutI(&iq, 3);
    chSysUnlock();
    (void)chIQGet(&iq);
    (void)chIQGet(&iq);
    (void)chIQGet(&iq);
    (void)chIQGet(&iq);
    n++;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  test_print("--- Score : ");
  test_printn(n * 4);
  test_println(" bytes/S");
}

ROMCONST struct testcase testbmk9 = {
  "Benchmark, I/O Queues throughput",
  NULL,
  NULL,
  bmk9_execute
};

/**
 * @page test_benchmarks_010 Virtual Timers set/reset performance
 *
 * <h2>Description</h2>
 * A virtual timer is set and immediately reset into a continuous loop.<br>
 * The performance is calculated by measuring the number of iterations after
 * a second of continuous operations.
 */

static void tmo(void *param) {(void)param;}

static void bmk10_execute(void) {
  static VirtualTimer vt1, vt2;
  uint32_t n = 0;

  test_wait_tick();
  test_start_timer(1000);
  do {
    chSysLock();
    chVTSetI(&vt1, 1, tmo, NULL);
    chVTSetI(&vt2, 10000, tmo, NULL);
    chVTResetI(&vt1);
    chVTResetI(&vt2);
    chSysUnlock();
    n++;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  test_print("--- Score : ");
  test_printn(n * 2);
  test_println(" timers/S");
}

ROMCONST struct testcase testbmk10 = {
  "Benchmark, virtual timers set/reset",
  NULL,
  NULL,
  bmk10_execute
};

/**
 * @page test_benchmarks_011 Semaphores wait/signal performance
 *
 * <h2>Description</h2>
 * A counting semaphore is taken/released into a continuous loop, no Context
 * Switch happens because the counter is always non negative.<br>
 * The performance is calculated by measuring the number of iterations after
 * a second of continuous operations.
 */

static void bmk11_setup(void) {

  chSemInit(&sem1, 1);
}

static void bmk11_execute(void) {
  uint32_t n = 0;

  test_wait_tick();
  test_start_timer(1000);
  do {
    chSemWait(&sem1);
    chSemSignal(&sem1);
    chSemWait(&sem1);
    chSemSignal(&sem1);
    chSemWait(&sem1);
    chSemSignal(&sem1);
    chSemWait(&sem1);
    chSemSignal(&sem1);
    n++;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  test_print("--- Score : ");
  test_printn(n * 4);
  test_println(" wait+signal/S");
}

ROMCONST struct testcase testbmk11 = {
  "Benchmark, semaphores wait/signal",
  bmk11_setup,
  NULL,
  bmk11_execute
};

#if CH_USE_MUTEXES || defined(__DOXYGEN__)
/**
 * @page test_benchmarks_012 Mutexes lock/unlock performance
 *
 * <h2>Description</h2>
 * A mutex is locked/unlocked into a continuous loop, no Context Switch happens
 * because there are no other threads asking for the mutex.<br>
 * The performance is calculated by measuring the number of iterations after
 * a second of continuous operations.
 */

static void bmk12_setup(void) {

  chMtxInit(&mtx1);
}

static void bmk12_execute(void) {
  uint32_t n = 0;

  test_wait_tick();
  test_start_timer(1000);
  do {
    chMtxLock(&mtx1);
    chMtxUnlock();
    chMtxLock(&mtx1);
    chMtxUnlock();
    chMtxLock(&mtx1);
    chMtxUnlock();
    chMtxLock(&mtx1);
    chMtxUnlock();
    n++;
#if defined(SIMULATOR)
    ChkIntSources();
#endif
  } while (!test_timer_done);
  test_print("--- Score : ");
  test_printn(n * 4);
  test_println(" lock+unlock/S");
}

ROMCONST struct testcase testbmk12 = {
  "Benchmark, mutexes lock/unlock",
  bmk12_setup,
  NULL,
  bmk12_execute
};
#endif

/**
 * @page test_benchmarks_013 RAM Footprint
 *
 * <h2>Description</h2>
 * The memory size of the various kernel objects is printed.
 */

static void bmk13_execute(void) {

  test_print("--- System: ");
  test_printn(sizeof(ReadyList) + sizeof(VTList) +
              PORT_IDLE_THREAD_STACK_SIZE +
              (sizeof(Thread) + sizeof(struct intctx) +
               sizeof(struct extctx) +
               PORT_INT_REQUIRED_STACK) * 2);
  test_println(" bytes");
  test_print("--- Thread: ");
  test_printn(sizeof(Thread));
  test_println(" bytes");
  test_print("--- Timer : ");
  test_printn(sizeof(VirtualTimer));
  test_println(" bytes");
  test_print("--- Semaph: ");
  test_printn(sizeof(Semaphore));
  test_println(" bytes");
#if CH_USE_EVENTS || defined(__DOXYGEN__)
  test_print("--- EventS: ");
  test_printn(sizeof(EventSource));
  test_println(" bytes");
  test_print("--- EventL: ");
  test_printn(sizeof(EventListener));
  test_println(" bytes");
#endif
#if CH_USE_MUTEXES || defined(__DOXYGEN__)
  test_print("--- Mutex : ");
  test_printn(sizeof(Mutex));
  test_println(" bytes");
#endif
#if CH_USE_CONDVARS || defined(__DOXYGEN__)
  test_print("--- CondV.: ");
  test_printn(sizeof(CondVar));
  test_println(" bytes");
#endif
#if CH_USE_QUEUES || defined(__DOXYGEN__)
  test_print("--- Queue : ");
  test_printn(sizeof(GenericQueue));
  test_println(" bytes");
#endif
#if CH_USE_MAILBOXES || defined(__DOXYGEN__)
  test_print("--- MailB.: ");
  test_printn(sizeof(Mailbox));
  test_println(" bytes");
#endif
}

ROMCONST struct testcase testbmk13 = {
  "Benchmark, RAM footprint",
  NULL,
  NULL,
  bmk13_execute
};

/**
 * @brief   Test sequence for benchmarks.
 */
ROMCONST struct testcase * ROMCONST patternbmk[] = {
#if !TEST_NO_BENCHMARKS
  &testbmk1,
  &testbmk2,
  &testbmk3,
  &testbmk4,
  &testbmk5,
  &testbmk6,
  &testbmk7,
  &testbmk8,
  &testbmk9,
  &testbmk10,
  &testbmk11,
#if CH_USE_MUTEXES || defined(__DOXYGEN__)
  &testbmk12,
#endif
  &testbmk13,
#endif
  NULL
};