aboutsummaryrefslogtreecommitdiff
path: root/math/libm-test-support.c
blob: cb78daa14b9c2772adf7031b0a76595206a986c4 (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
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
/* Support code for testing libm functions (compiled once per type).
   Copyright (C) 1997-2018 Free Software Foundation, Inc.
   This file is part of the GNU C Library.

   The GNU C Library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
   License as published by the Free Software Foundation; either
   version 2.1 of the License, or (at your option) any later version.

   The GNU C Library 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
   Lesser General Public License for more details.

   You should have received a copy of the GNU Lesser General Public
   License along with the GNU C Library; if not, see
   <http://www.gnu.org/licenses/>.  */

/* Part of testsuite for libm.

   libm-test-support.c contains functions shared by tests of different
   libm functions and types; it is compiled once per type.
   libm-test-driver.c defines the main function, and various variables
   that are used to configure the code in libm-test-support.c for
   different types and for variants such as testing inline functions.

   The tests of individual functions are in .inc files processed by
   gen-libm-test.py, with the resulting files included together with
   libm-test-driver.c.

   The per-type headers included both before libm-test-support.c and
   for the tests of individual functions must define the following
   macros:

   FUNC(function): Convert general function name (like cos) to name
   with correct suffix (e.g. cosl or cosf).

   FLOAT: Floating-point type to test.

   BUILD_COMPLEX(real, imag): Create a complex number by calling a
   macro such as CMPLX.

   PREFIX: The prefix for <float.h> macros for the type (e.g. LDBL,
   DBL, or FLT).

   TYPE_STR: The name of the type as used in ulps files, as a string.

   ULP_IDX, ULP_I_IDX: The array indexes for ulps values for this
   function.

   LIT: Append the correct suffix to a literal.

   LITM: Append the correct suffix to an M_* macro name.

   FTOSTR: A function similar in type to strfromf which converts a
   FLOAT to a string.

   snan_value_MACRO: The macro such as SNAN for a signaling NaN for
   the type.

*/

/* Parameter handling is primitive in the moment:
   --verbose=[0..3] for different levels of output:
   0: only error count
   1: basic report on failed tests (default)
   2: full report on all tests
   -v for full output (equals --verbose=3)
   -u for generation of an ULPs file
 */

/* "Philosophy":

   This suite tests some aspects of the correct implementation of
   mathematical functions in libm.  Some simple, specific parameters
   are tested for correctness but there's no exhaustive
   testing.  Handling of specific inputs (e.g. infinity, not-a-number)
   is also tested.  Correct handling of exceptions is checked
   against.  These implemented tests should check all cases that are
   specified in ISO C99.

   NaN values: The payload of NaNs is set in inputs for functions
   where it is significant, and is examined in the outputs of some
   functions.

   Inline functions: Inlining functions should give an improvement in
   speed - but not in precission.  The inlined functions return
   reasonable values for a reasonable range of input values.  The
   result is not necessarily correct for all values and exceptions are
   not correctly raised in all cases.  Problematic input and return
   values are infinity, not-a-number and minus zero.  This suite
   therefore does not check these specific inputs and the exception
   handling for inlined mathematical functions - just the "reasonable"
   values are checked.

   Beware: The tests might fail for any of the following reasons:
   - Tests are wrong
   - Functions are wrong
   - Floating Point Unit not working properly
   - Compiler has errors

   With e.g. gcc 2.7.2.2 the test for cexp fails because of a compiler error.


   To Do: All parameter should be numbers that can be represented as
   exact floating point values.  Currently some values cannot be
   represented exactly and therefore the result is not the expected
   result.  For this we will use 36 digits so that numbers can be
   represented exactly.  */

#include "libm-test-support.h"

#include <argp.h>
#include <errno.h>
#include <string.h>

/* This header defines func_ulps, func_real_ulps and func_imag_ulps
   arrays.  */
#include "libm-test-ulps.h"

/* Maximum character buffer to store a stringitized FLOAT value.  */
#define FSTR_MAX (128)

#define ulps_file_name "ULPs"	/* Name of the ULPs file.  */
static FILE *ulps_file;		/* File to document difference.  */
static int output_ulps;		/* Should ulps printed?  */
static char *output_dir;	/* Directory where generated files will be written.  */

static int noErrors;	/* number of errors */
static int noTests;	/* number of tests (without testing exceptions) */
static int noExcTests;	/* number of tests for exception flags */
static int noErrnoTests;/* number of tests for errno values */

static int verbose;
static int output_max_error;	/* Should the maximal errors printed?  */
static int output_points;	/* Should the single function results printed?  */
static int ignore_max_ulp;	/* Should we ignore max_ulp?  */
static int test_ibm128;		/* Is argument or result IBM long double?  */

static FLOAT max_error, real_max_error, imag_max_error;

static FLOAT prev_max_error, prev_real_max_error, prev_imag_max_error;

static FLOAT max_valid_error;

/* Sufficient numbers of digits to represent any floating-point value
   unambiguously (for any choice of the number of bits in the first
   hex digit, in the case of TYPE_HEX_DIG).  When used with printf
   formats where the precision counts only digits after the point, 1
   is subtracted from these values. */
#define TYPE_DECIMAL_DIG __CONCATX (PREFIX, _DECIMAL_DIG)
#define TYPE_HEX_DIG ((MANT_DIG + 6) / 4)

/* Converts VALUE (a floating-point number) to string and writes it to DEST.
   PRECISION specifies the number of fractional digits that should be printed.
   CONVERSION is the conversion specifier, such as in printf, e.g. 'f' or 'a'.
   The output is prepended with an empty space if VALUE is non-negative.  */
static void
fmt_ftostr (char *dest, size_t size, int precision, const char *conversion,
	    FLOAT value)
{
  char format[64];
  char *ptr_format;
  int ret;

  /* Generate the format string.  */
  ptr_format = stpcpy (format, "%.");
  ret = sprintf (ptr_format, "%d", precision);
  ptr_format += ret;
  ptr_format = stpcpy (ptr_format, conversion);

  /* Add a space to the beginning of the output string, if the floating-point
     number is non-negative.  This mimics the behavior of the space (' ') flag
     in snprintf, which is not available on strfrom.  */
  if (! signbit (value))
    {
      *dest = ' ';
      dest++;
      size--;
    }

  /* Call the float to string conversion function, e.g.: strfromd.  */
  FTOSTR (dest, size, format, value);
}

/* Compare KEY (a string, with the name of a function) with ULP (a
   pointer to a struct ulp_data structure), returning a value less
   than, equal to or greater than zero for use in bsearch.  */

static int
compare_ulp_data (const void *key, const void *ulp)
{
  const char *keystr = key;
  const struct ulp_data *ulpdat = ulp;
  return strcmp (keystr, ulpdat->name);
}

static const int ulp_i_idx = ULP_I_IDX;
static const int ulp_idx = ULP_IDX;

/* Return the ulps for NAME in array DATA with NMEMB elements, or 0 if
   no ulps listed.  */

static FLOAT
find_ulps (const char *name, const struct ulp_data *data, size_t nmemb)
{
  const struct ulp_data *entry = bsearch (name, data, nmemb, sizeof (*data),
					  compare_ulp_data);
  if (entry == NULL)
    return 0;
  else
    return entry->max_ulp[(flag_test_inline ? ulp_i_idx : ulp_idx)];
}

void
init_max_error (const char *name, int exact, int testing_ibm128)
{
  max_error = 0;
  real_max_error = 0;
  imag_max_error = 0;
  test_ibm128 = testing_ibm128;
  prev_max_error = find_ulps (name, func_ulps,
			      sizeof (func_ulps) / sizeof (func_ulps[0]));
  prev_real_max_error = find_ulps (name, func_real_ulps,
				   (sizeof (func_real_ulps)
				    / sizeof (func_real_ulps[0])));
  prev_imag_max_error = find_ulps (name, func_imag_ulps,
				   (sizeof (func_imag_ulps)
				    / sizeof (func_imag_ulps[0])));
  if (testing_ibm128)
    /* The documented accuracy of IBM long double division is 3ulp
       (see libgcc/config/rs6000/ibm-ldouble-format), so do not
       require better accuracy for libm functions that are exactly
       defined for other formats.  */
    max_valid_error = exact ? 3 : 16;
  else
    max_valid_error = exact ? 0 : 9;
  prev_max_error = (prev_max_error <= max_valid_error
		    ? prev_max_error
		    : max_valid_error);
  prev_real_max_error = (prev_real_max_error <= max_valid_error
			 ? prev_real_max_error
			 : max_valid_error);
  prev_imag_max_error = (prev_imag_max_error <= max_valid_error
			 ? prev_imag_max_error
			 : max_valid_error);
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;
}

static void
set_max_error (FLOAT current, FLOAT *curr_max_error)
{
  if (current > *curr_max_error && current <= max_valid_error)
    *curr_max_error = current;
}


/* Print a FLOAT.  */
static void
print_float (FLOAT f)
{
  /* As printf doesn't differ between a sNaN and a qNaN, do this manually.  */
  if (issignaling (f))
    printf ("sNaN\n");
  else if (isnan (f))
    printf ("qNaN\n");
  else
    {
      char fstrn[FSTR_MAX], fstrx[FSTR_MAX];
      fmt_ftostr (fstrn, FSTR_MAX, TYPE_DECIMAL_DIG - 1, "e", f);
      fmt_ftostr (fstrx, FSTR_MAX, TYPE_HEX_DIG - 1, "a", f);
      printf ("%s  %s\n", fstrn, fstrx);
    }
}

/* Should the message print to screen?  This depends on the verbose flag,
   and the test status.  */
static int
print_screen (int ok)
{
  if (output_points
      && (verbose > 1
	  || (verbose == 1 && ok == 0)))
    return 1;
  return 0;
}


/* Should the message print to screen?  This depends on the verbose flag,
   and the test status.  */
static int
print_screen_max_error (int ok)
{
  if (output_max_error
      && (verbose > 1
	  || ((verbose == 1) && (ok == 0))))
    return 1;
  return 0;
}

/* Update statistic counters.  */
static void
update_stats (int ok)
{
  ++noTests;
  if (!ok)
    ++noErrors;
}

static void
print_function_ulps (const char *function_name, FLOAT ulp)
{
  if (output_ulps)
    {
      char ustrn[FSTR_MAX];
      FTOSTR (ustrn, FSTR_MAX, "%.0f", FUNC (ceil) (ulp));
      fprintf (ulps_file, "Function: \"%s\":\n", function_name);
      fprintf (ulps_file, "%s: %s\n", qtype_str, ustrn);
    }
}


static void
print_complex_function_ulps (const char *function_name, FLOAT real_ulp,
			     FLOAT imag_ulp)
{
  if (output_ulps)
    {
      char fstrn[FSTR_MAX];
      if (real_ulp != 0.0)
	{
	  FTOSTR (fstrn, FSTR_MAX, "%.0f", FUNC (ceil) (real_ulp));
	  fprintf (ulps_file, "Function: Real part of \"%s\":\n", function_name);
	  fprintf (ulps_file, "%s: %s\n", qtype_str, fstrn);
	}
      if (imag_ulp != 0.0)
	{
	  FTOSTR (fstrn, FSTR_MAX, "%.0f", FUNC (ceil) (imag_ulp));
	  fprintf (ulps_file, "Function: Imaginary part of \"%s\":\n", function_name);
	  fprintf (ulps_file, "%s: %s\n", qtype_str, fstrn);
	}


    }
}



/* Test if Floating-Point stack hasn't changed */
static void
fpstack_test (const char *test_name)
{
#if defined (__i386__) || defined (__x86_64__)
  static int old_stack;
  int sw;

  asm ("fnstsw" : "=a" (sw));
  sw >>= 11;
  sw &= 7;

  if (sw != old_stack)
    {
      printf ("FP-Stack wrong after test %s (%d, should be %d)\n",
	      test_name, sw, old_stack);
      ++noErrors;
      old_stack = sw;
    }
#endif
}


void
print_max_error (const char *func_name)
{
  int ok = 0;

  if (max_error == 0.0 || (max_error <= prev_max_error && !ignore_max_ulp))
    {
      ok = 1;
    }

  if (!ok)
    print_function_ulps (func_name, max_error);


  if (print_screen_max_error (ok))
    {
      char mestr[FSTR_MAX], pmestr[FSTR_MAX];
      FTOSTR (mestr, FSTR_MAX, "%.0f", FUNC (ceil) (max_error));
      FTOSTR (pmestr, FSTR_MAX, "%.0f", FUNC (ceil) (prev_max_error));
      printf ("Maximal error of `%s'\n", func_name);
      printf (" is      : %s ulp\n", mestr);
      printf (" accepted: %s ulp\n", pmestr);
    }

  update_stats (ok);
}


void
print_complex_max_error (const char *func_name)
{
  int real_ok = 0, imag_ok = 0, ok;

  if (real_max_error == 0
      || (real_max_error <= prev_real_max_error && !ignore_max_ulp))
    {
      real_ok = 1;
    }

  if (imag_max_error == 0
      || (imag_max_error <= prev_imag_max_error && !ignore_max_ulp))
    {
      imag_ok = 1;
    }

  ok = real_ok && imag_ok;

  if (!ok)
    print_complex_function_ulps (func_name,
				 real_ok ? 0 : real_max_error,
				 imag_ok ? 0 : imag_max_error);

  if (print_screen_max_error (ok))
    {
      char rmestr[FSTR_MAX], prmestr[FSTR_MAX];
      char imestr[FSTR_MAX], pimestr[FSTR_MAX];
      FTOSTR (rmestr, FSTR_MAX, "%.0f", FUNC (ceil) (real_max_error));
      FTOSTR (prmestr, FSTR_MAX, "%.0f", FUNC (ceil) (prev_real_max_error));
      FTOSTR (imestr, FSTR_MAX, "%.0f", FUNC (ceil) (imag_max_error));
      FTOSTR (pimestr, FSTR_MAX, "%.0f", FUNC (ceil) (prev_imag_max_error));
      printf ("Maximal error of real part of: %s\n", func_name);
      printf (" is      : %s ulp\n", rmestr);
      printf (" accepted: %s ulp\n", prmestr);
      printf ("Maximal error of imaginary part of: %s\n", func_name);
      printf (" is      : %s ulp\n", imestr);
      printf (" accepted: %s ulp\n", pimestr);
    }

  update_stats (ok);
}


#if FE_ALL_EXCEPT
/* Test whether a given exception was raised.  */
static void
test_single_exception (const char *test_name,
		       int exception,
		       int exc_flag,
		       int fe_flag,
		       const char *flag_name)
{
  int ok = 1;
  if (exception & exc_flag)
    {
      if (fetestexcept (fe_flag))
	{
	  if (print_screen (1))
	    printf ("Pass: %s: Exception \"%s\" set\n", test_name, flag_name);
	}
      else
	{
	  ok = 0;
	  if (print_screen (0))
	    printf ("Failure: %s: Exception \"%s\" not set\n",
		    test_name, flag_name);
	}
    }
  else
    {
      if (fetestexcept (fe_flag))
	{
	  ok = 0;
	  if (print_screen (0))
	    printf ("Failure: %s: Exception \"%s\" set\n",
		    test_name, flag_name);
	}
      else
	{
	  if (print_screen (1))
	    printf ("%s: Exception \"%s\" not set\n", test_name,
		    flag_name);
	}
    }
  if (!ok)
    ++noErrors;
}
#endif

/* Test whether exceptions given by EXCEPTION are raised.  Ignore thereby
   allowed but not required exceptions.
*/
static void
test_exceptions (const char *test_name, int exception)
{
  if (flag_test_exceptions && EXCEPTION_TESTS (FLOAT))
    {
      ++noExcTests;
#ifdef FE_DIVBYZERO
      if ((exception & DIVIDE_BY_ZERO_EXCEPTION_OK) == 0)
	test_single_exception (test_name, exception,
			       DIVIDE_BY_ZERO_EXCEPTION, FE_DIVBYZERO,
			       "Divide by zero");
#endif
#ifdef FE_INVALID
      if ((exception & INVALID_EXCEPTION_OK) == 0)
	test_single_exception (test_name, exception,
			       INVALID_EXCEPTION, FE_INVALID,
			       "Invalid operation");
#endif
#ifdef FE_OVERFLOW
      if ((exception & OVERFLOW_EXCEPTION_OK) == 0)
	test_single_exception (test_name, exception, OVERFLOW_EXCEPTION,
			       FE_OVERFLOW, "Overflow");
#endif
      /* Spurious "underflow" and "inexact" exceptions are always
	 allowed for IBM long double, in line with the underlying
	 arithmetic.  */
#ifdef FE_UNDERFLOW
      if ((exception & UNDERFLOW_EXCEPTION_OK) == 0
	  && !(test_ibm128
	       && (exception & UNDERFLOW_EXCEPTION) == 0))
	test_single_exception (test_name, exception, UNDERFLOW_EXCEPTION,
			       FE_UNDERFLOW, "Underflow");
#endif
#ifdef FE_INEXACT
      if ((exception & (INEXACT_EXCEPTION | NO_INEXACT_EXCEPTION)) != 0
	  && !(test_ibm128
	       && (exception & NO_INEXACT_EXCEPTION) != 0))
	test_single_exception (test_name, exception, INEXACT_EXCEPTION,
			       FE_INEXACT, "Inexact");
#endif
    }
  feclearexcept (FE_ALL_EXCEPT);
}

/* Test whether errno for TEST_NAME, set to ERRNO_VALUE, has value
   EXPECTED_VALUE (description EXPECTED_NAME).  */
static void
test_single_errno (const char *test_name, int errno_value,
		   int expected_value, const char *expected_name)
{
  if (errno_value == expected_value)
    {
      if (print_screen (1))
	printf ("Pass: %s: errno set to %d (%s)\n", test_name, errno_value,
		expected_name);
    }
  else
    {
      ++noErrors;
      if (print_screen (0))
	printf ("Failure: %s: errno set to %d, expected %d (%s)\n",
		test_name, errno_value, expected_value, expected_name);
    }
}

/* Test whether errno (value ERRNO_VALUE) has been for TEST_NAME set
   as required by EXCEPTIONS.  */
static void
test_errno (const char *test_name, int errno_value, int exceptions)
{
  if (flag_test_errno)
    {
      ++noErrnoTests;
      if (exceptions & ERRNO_UNCHANGED)
	test_single_errno (test_name, errno_value, 0, "unchanged");
      if (exceptions & ERRNO_EDOM)
	test_single_errno (test_name, errno_value, EDOM, "EDOM");
      if (exceptions & ERRNO_ERANGE)
	test_single_errno (test_name, errno_value, ERANGE, "ERANGE");
    }
}

/* Returns the number of ulps that GIVEN is away from EXPECTED.  */
#define ULPDIFF(given, expected) \
	(FUNC(fabs) ((given) - (expected)) / ulp (expected))

/* Returns the size of an ulp for VALUE.  */
static FLOAT
ulp (FLOAT value)
{
  FLOAT ulp;

  switch (fpclassify (value))
    {
      case FP_ZERO:
	/* We compute the distance to the next FP which is the same as the
	   value of the smallest subnormal number. Previously we used
	   2^-(MANT_DIG - 1) which is too large a value to be useful. Note that we
	   can't use ilogb(0), since that isn't a valid thing to do. As a point
	   of comparison Java's ulp returns the next normal value e.g.
	   2^(1 - MAX_EXP) for ulp(0), but that is not what we want for
	   glibc.  */
	/* Fall through...  */
      case FP_SUBNORMAL:
        /* The next closest subnormal value is a constant distance away.  */
	ulp = FUNC(ldexp) (1.0, MIN_EXP - MANT_DIG);
	break;

      case FP_NORMAL:
	ulp = FUNC(ldexp) (1.0, FUNC(ilogb) (value) - MANT_DIG + 1);
	break;

      default:
	/* It should never happen. */
	abort ();
	break;
    }
  return ulp;
}

static void
check_float_internal (const char *test_name, FLOAT computed, FLOAT expected,
		      int exceptions,
		      FLOAT *curr_max_error, FLOAT max_ulp)
{
  int ok = 0;
  int print_diff = 0;
  FLOAT diff = 0;
  FLOAT ulps = 0;
  int errno_value = errno;

  test_exceptions (test_name, exceptions);
  test_errno (test_name, errno_value, exceptions);
  if (exceptions & IGNORE_RESULT)
    goto out;
  if (issignaling (computed) && issignaling (expected))
    {
      if ((exceptions & TEST_NAN_SIGN) != 0
	  && signbit (computed) != signbit (expected))
	{
	  ok = 0;
	  printf ("signaling NaN has wrong sign.\n");
	}
      else if ((exceptions & TEST_NAN_PAYLOAD) != 0
	       && (FUNC (getpayload) (&computed)
		   != FUNC (getpayload) (&expected)))
	{
	  ok = 0;
	  printf ("signaling NaN has wrong payload.\n");
	}
      else
	ok = 1;
    }
  else if (issignaling (computed) || issignaling (expected))
    ok = 0;
  else if (isnan (computed) && isnan (expected))
    {
      if ((exceptions & TEST_NAN_SIGN) != 0
	  && signbit (computed) != signbit (expected))
	{
	  ok = 0;
	  printf ("quiet NaN has wrong sign.\n");
	}
      else if ((exceptions & TEST_NAN_PAYLOAD) != 0
	       && (FUNC (getpayload) (&computed)
		   != FUNC (getpayload) (&expected)))
	{
	  ok = 0;
	  printf ("quiet NaN has wrong payload.\n");
	}
      else
	ok = 1;
    }
  else if (isinf (computed) && isinf (expected))
    {
      /* Test for sign of infinities.  */
      if ((exceptions & IGNORE_ZERO_INF_SIGN) == 0
	  && signbit (computed) != signbit (expected))
	{
	  ok = 0;
	  printf ("infinity has wrong sign.\n");
	}
      else
	ok = 1;
    }
  /* Don't calculate ULPs for infinities or any kind of NaNs.  */
  else if (isinf (computed) || isnan (computed)
	   || isinf (expected) || isnan (expected))
    ok = 0;
  else
    {
      diff = FUNC(fabs) (computed - expected);
      ulps = ULPDIFF (computed, expected);
      set_max_error (ulps, curr_max_error);
      print_diff = 1;
      if ((exceptions & IGNORE_ZERO_INF_SIGN) == 0
	  && computed == 0.0 && expected == 0.0
	  && signbit(computed) != signbit (expected))
	ok = 0;
      else if (ulps <= max_ulp && !ignore_max_ulp)
	ok = 1;
      else
	ok = 0;
    }
  if (print_screen (ok))
    {
      if (!ok)
	printf ("Failure: ");
      printf ("Test: %s\n", test_name);
      printf ("Result:\n");
      printf (" is:         ");
      print_float (computed);
      printf (" should be:  ");
      print_float (expected);
      if (print_diff)
	{
	  char dstrn[FSTR_MAX], dstrx[FSTR_MAX];
	  char ustrn[FSTR_MAX], mustrn[FSTR_MAX];
	  fmt_ftostr (dstrn, FSTR_MAX, TYPE_DECIMAL_DIG - 1, "e", diff);
	  fmt_ftostr (dstrx, FSTR_MAX, TYPE_HEX_DIG - 1, "a", diff);
	  fmt_ftostr (ustrn, FSTR_MAX, 4, "f", ulps);
	  fmt_ftostr (mustrn, FSTR_MAX, 4, "f", max_ulp);
	  printf (" difference: %s  %s\n", dstrn, dstrx);
	  printf (" ulp       : %s\n", ustrn);
	  printf (" max.ulp   : %s\n", mustrn);
	}
    }
  update_stats (ok);

 out:
  fpstack_test (test_name);
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;
}


void
check_float (const char *test_name, FLOAT computed, FLOAT expected,
	     int exceptions)
{
  check_float_internal (test_name, computed, expected,
			exceptions, &max_error, prev_max_error);
}


void
check_complex (const char *test_name, CFLOAT computed,
	       CFLOAT expected,
	       int exception)
{
  FLOAT part_comp, part_exp;
  char *str;

  if (asprintf (&str, "Real part of: %s", test_name) == -1)
    abort ();

  part_comp = __real__ computed;
  part_exp = __real__ expected;

  check_float_internal (str, part_comp, part_exp,
			exception, &real_max_error, prev_real_max_error);
  free (str);

  if (asprintf (&str, "Imaginary part of: %s", test_name) == -1)
    abort ();

  part_comp = __imag__ computed;
  part_exp = __imag__ expected;

  /* Don't check again for exceptions or errno, just pass through the
     other relevant flags.  */
  check_float_internal (str, part_comp, part_exp,
			exception & (IGNORE_ZERO_INF_SIGN
				     | TEST_NAN_SIGN
				     | IGNORE_RESULT),
			&imag_max_error, prev_imag_max_error);
  free (str);
}


/* Check that computed and expected values are equal (int values).  */
void
check_int (const char *test_name, int computed, int expected,
	   int exceptions)
{
  int ok = 0;
  int errno_value = errno;

  test_exceptions (test_name, exceptions);
  test_errno (test_name, errno_value, exceptions);
  if (exceptions & IGNORE_RESULT)
    goto out;
  noTests++;
  if (computed == expected)
    ok = 1;

  if (print_screen (ok))
    {
      if (!ok)
	printf ("Failure: ");
      printf ("Test: %s\n", test_name);
      printf ("Result:\n");
      printf (" is:         %d\n", computed);
      printf (" should be:  %d\n", expected);
    }

  update_stats (ok);
 out:
  fpstack_test (test_name);
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;
}


/* Check that computed and expected values are equal (long int values).  */
void
check_long (const char *test_name, long int computed, long int expected,
	    int exceptions)
{
  int ok = 0;
  int errno_value = errno;

  test_exceptions (test_name, exceptions);
  test_errno (test_name, errno_value, exceptions);
  if (exceptions & IGNORE_RESULT)
    goto out;
  noTests++;
  if (computed == expected)
    ok = 1;

  if (print_screen (ok))
    {
      if (!ok)
	printf ("Failure: ");
      printf ("Test: %s\n", test_name);
      printf ("Result:\n");
      printf (" is:         %ld\n", computed);
      printf (" should be:  %ld\n", expected);
    }

  update_stats (ok);
 out:
  fpstack_test (test_name);
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;
}


/* Check that computed value is true/false.  */
void
check_bool (const char *test_name, int computed, int expected,
	    int exceptions)
{
  int ok = 0;
  int errno_value = errno;

  test_exceptions (test_name, exceptions);
  test_errno (test_name, errno_value, exceptions);
  if (exceptions & IGNORE_RESULT)
    goto out;
  noTests++;
  if ((computed == 0) == (expected == 0))
    ok = 1;

  if (print_screen (ok))
    {
      if (!ok)
	printf ("Failure: ");
      printf ("Test: %s\n", test_name);
      printf ("Result:\n");
      printf (" is:         %d\n", computed);
      printf (" should be:  %d\n", expected);
    }

  update_stats (ok);
 out:
  fpstack_test (test_name);
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;
}


/* check that computed and expected values are equal (long int values) */
void
check_longlong (const char *test_name, long long int computed,
		long long int expected,
		int exceptions)
{
  int ok = 0;
  int errno_value = errno;

  test_exceptions (test_name, exceptions);
  test_errno (test_name, errno_value, exceptions);
  if (exceptions & IGNORE_RESULT)
    goto out;
  noTests++;
  if (computed == expected)
    ok = 1;

  if (print_screen (ok))
    {
      if (!ok)
	printf ("Failure:");
      printf ("Test: %s\n", test_name);
      printf ("Result:\n");
      printf (" is:         %lld\n", computed);
      printf (" should be:  %lld\n", expected);
    }

  update_stats (ok);
 out:
  fpstack_test (test_name);
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;
}


/* Check that computed and expected values are equal (intmax_t values).  */
void
check_intmax_t (const char *test_name, intmax_t computed,
		intmax_t expected, int exceptions)
{
  int ok = 0;
  int errno_value = errno;

  test_exceptions (test_name, exceptions);
  test_errno (test_name, errno_value, exceptions);
  if (exceptions & IGNORE_RESULT)
    goto out;
  noTests++;
  if (computed == expected)
    ok = 1;

  if (print_screen (ok))
    {
      if (!ok)
	printf ("Failure:");
      printf ("Test: %s\n", test_name);
      printf ("Result:\n");
      printf (" is:         %jd\n", computed);
      printf (" should be:  %jd\n", expected);
    }

  update_stats (ok);
 out:
  fpstack_test (test_name);
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;
}


/* Check that computed and expected values are equal (uintmax_t values).  */
void
check_uintmax_t (const char *test_name, uintmax_t computed,
		 uintmax_t expected, int exceptions)
{
  int ok = 0;
  int errno_value = errno;

  test_exceptions (test_name, exceptions);
  test_errno (test_name, errno_value, exceptions);
  if (exceptions & IGNORE_RESULT)
    goto out;
  noTests++;
  if (computed == expected)
    ok = 1;

  if (print_screen (ok))
    {
      if (!ok)
	printf ("Failure:");
      printf ("Test: %s\n", test_name);
      printf ("Result:\n");
      printf (" is:         %ju\n", computed);
      printf (" should be:  %ju\n", expected);
    }

  update_stats (ok);
 out:
  fpstack_test (test_name);
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;
}

/* Return whether a test with flags EXCEPTIONS should be run.  */
int
enable_test (int exceptions)
{
  if (exceptions & XFAIL_TEST)
    return 0;
  if (flag_test_inline && (exceptions & NO_TEST_INLINE))
    return 0;
  if (flag_test_finite && (exceptions & NON_FINITE) != 0)
    return 0;
  if ((!SNAN_TESTS (FLOAT) || !snan_tests_arg)
      && (exceptions & TEST_SNAN) != 0)
    return 0;
  if (flag_test_mathvec && (exceptions & NO_TEST_MATHVEC) != 0)
    return 0;

  return 1;
}

static void
initialize (void)
{
  fpstack_test ("start *init*");

  /* Clear all exceptions.  From now on we must not get random exceptions.  */
  feclearexcept (FE_ALL_EXCEPT);
  errno = 0;

  /* Test to make sure we start correctly.  */
  fpstack_test ("end *init*");
}

/* Definitions of arguments for argp functions.  */
static const struct argp_option options[] =
{
  { "verbose", 'v', "NUMBER", 0, "Level of verbosity (0..3)"},
  { "ulps-file", 'u', NULL, 0, "Output ulps to file ULPs"},
  { "no-max-error", 'f', NULL, 0,
    "Don't output maximal errors of functions"},
  { "no-points", 'p', NULL, 0,
    "Don't output results of functions invocations"},
  { "ignore-max-ulp", 'i', "yes/no", 0,
    "Ignore given maximal errors"},
  { "output-dir", 'o', "DIR", 0,
    "Directory where generated files will be placed"},
  { NULL, 0, NULL, 0, NULL }
};

/* Prototype for option handler.  */
static error_t parse_opt (int key, char *arg, struct argp_state *state);

/* Data structure to communicate with argp functions.  */
static struct argp argp =
{
  options, parse_opt, NULL, doc,
};


/* Handle program arguments.  */
static error_t
parse_opt (int key, char *arg, struct argp_state *state)
{
  switch (key)
    {
    case 'f':
      output_max_error = 0;
      break;
    case 'i':
      if (strcmp (arg, "yes") == 0)
	ignore_max_ulp = 1;
      else if (strcmp (arg, "no") == 0)
	ignore_max_ulp = 0;
      break;
    case 'o':
      output_dir = (char *) malloc (strlen (arg) + 1);
      if (output_dir != NULL)
	strcpy (output_dir, arg);
      else
        return errno;
      break;
    case 'p':
      output_points = 0;
      break;
    case 'u':
      output_ulps = 1;
      break;
    case 'v':
      if (optarg)
	verbose = (unsigned int) strtoul (optarg, NULL, 0);
      else
	verbose = 3;
      break;
    default:
      return ARGP_ERR_UNKNOWN;
    }
  return 0;
}

/* Verify that our ulp () implementation is behaving as expected
   or abort.  */
static void
check_ulp (void)
{
   FLOAT ulps, ulpx, value;
   int i;
   /* Check ulp of zero is a subnormal value...  */
   ulps = ulp (0x0.0p0);
   if (fpclassify (ulps) != FP_SUBNORMAL)
     {
       fprintf (stderr, "ulp (0x0.0p0) is not FP_SUBNORMAL!\n");
       exit (EXIT_FAILURE);
     }
   /* Check that the ulp of one is a normal value... */
   ulps = ulp (LIT(1.0));
   if (fpclassify (ulps) != FP_NORMAL)
     {
       fprintf (stderr, "ulp (1.0L) is not FP_NORMAL\n");
       exit (EXIT_FAILURE);
     }

   /* Compute the next subnormal value using nextafter to validate ulp.
      We allow +/- 1 ulp around the represented value.  */
   value = FUNC(nextafter) (0, 1);
   ulps = ULPDIFF (value, 0);
   ulpx = ulp (LIT(1.0));
   if (ulps < (LIT(1.0) - ulpx) || ulps > (LIT(1.0) + ulpx))
     {
       fprintf (stderr, "Value outside of 1 +/- 1ulp.\n");
       exit (EXIT_FAILURE);
     }
   /* Compute the nearest representable number from 10 towards 20.
      The result is 10 + 1ulp.  We use this to check the ulp function.
      We allow +/- 1 ulp around the represented value.  */
   value = FUNC(nextafter) (10, 20);
   ulps = ULPDIFF (value, 10);
   ulpx = ulp (LIT(1.0));
   if (ulps < (LIT(1.0) - ulpx) || ulps > (LIT(1.0) + ulpx))
     {
       fprintf (stderr, "Value outside of 1 +/- 1ulp.\n");
       exit (EXIT_FAILURE);
     }
   /* This gives one more ulp.  */
   value = FUNC(nextafter) (value, 20);
   ulps = ULPDIFF (value, 10);
   ulpx = ulp (LIT(2.0));
   if (ulps < (LIT(2.0) - ulpx) || ulps > (LIT(2.0) + ulpx))
     {
       fprintf (stderr, "Value outside of 2 +/- 1ulp.\n");
       exit (EXIT_FAILURE);
     }
   /* And now calculate 100 ulp.  */
   for (i = 2; i < 100; i++)
     value = FUNC(nextafter) (value, 20);
   ulps = ULPDIFF (value, 10);
   ulpx = ulp (LIT(100.0));
   if (ulps < (LIT(100.0) - ulpx) || ulps > (LIT(100.0) + ulpx))
     {
       fprintf (stderr, "Value outside of 100 +/- 1ulp.\n");
       exit (EXIT_FAILURE);
     }
}

/* Do all initialization for a test run with arguments given by ARGC
   and ARGV.  */
void
libm_test_init (int argc, char **argv)
{
  int remaining;
  char *ulps_file_path;
  size_t dir_len = 0;

  verbose = 1;
  output_ulps = 0;
  output_max_error = 1;
  output_points = 1;
  output_dir = NULL;
  /* XXX set to 0 for releases.  */
  ignore_max_ulp = 0;

  /* Parse and process arguments.  */
  argp_parse (&argp, argc, argv, 0, &remaining, NULL);

  if (remaining != argc)
    {
      fprintf (stderr, "wrong number of arguments");
      argp_help (&argp, stdout, ARGP_HELP_SEE, program_invocation_short_name);
      exit (EXIT_FAILURE);
    }

  if (output_ulps)
    {
      if (output_dir != NULL)
	dir_len = strlen (output_dir);
      ulps_file_path = (char *) malloc (dir_len + strlen (ulps_file_name) + 1);
      if (ulps_file_path == NULL)
        {
	  perror ("can't allocate path for `ULPs' file: ");
	  exit (1);
        }
      sprintf (ulps_file_path, "%s%s", output_dir == NULL ? "" : output_dir, ulps_file_name);
      ulps_file = fopen (ulps_file_path, "a");
      if (ulps_file == NULL)
	{
	  perror ("can't open file `ULPs' for writing: ");
	  exit (1);
	}
    }


  initialize ();
  fputs (test_msg, stdout);

  check_ulp ();
}

/* Process the test results, returning the exit status.  */
int
libm_test_finish (void)
{
  if (output_ulps)
    fclose (ulps_file);

  printf ("\nTest suite completed:\n");
  printf ("  %d test cases plus %d tests for exception flags and\n"
	  "    %d tests for errno executed.\n",
	  noTests, noExcTests, noErrnoTests);
  if (noErrors)
    {
      printf ("  %d errors occurred.\n", noErrors);
      return 1;
    }
  printf ("  All tests passed successfully.\n");

  return 0;
}