summaryrefslogtreecommitdiff
path: root/math
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>1997-05-29 12:06:58 +0000
committerUlrich Drepper <drepper@redhat.com>1997-05-29 12:06:58 +0000
commit76b87c039ba8d20add4f52ba43f3471fd92e210b (patch)
tree08c6a1bc32f699056a97f2683bc859cb0d7f6871 /math
parent06bdbaa0a614c256b34214fde7c395f9e0a6206c (diff)
Update.
1997-05-29 12:48 Ulrich Drepper <drepper@cygnus.com> * io/ftw.c: Complete rewrite. Add implementation of `nftw'. * io/ftw.h: Update for new implementation and XPG4.2. * login/Makefile: Update for UTMP daemon implementation. Update resolver code to bind-4.9.6-T1A. * resolv/Banner: Update. * nss/digits_dots.c: Adapt text address matching to T1A. * nss/nss_files/files-hosts.c: Always use inet_pton. * resolv/base64.c (b64_pton): Follow T1A but don't use this code since it would lead to warnings. * resolv/gethnamaddr.c (getanswer): Test host name for maximal length at several places. * resolv/inet_net_pton.c (inet_net_pton_ipv4): Correct typo in comment. * resolv/res_comp.c (dn_expand): Check for overflow. (dn_comp): Likewise. * resolv/res_debug.c (precsize_aton): Better implementation. * resolv/res_init.c (res_init): Make `buf' of size MAXDNAME. * resolv/res_send.c (res_send): Check for overflow in descriptor set. * resolv/nss_dns/dns-host.c (getanswer_r): Test host name for maximal length at several places. 1997-05-29 12:51 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp-private.h (struct utfuncs): Add one more parameter to updwtmp function. Declare all three function jump tables. * login/utmp.h: Declare __utmpname. * login/getutent_r.c: Remove db backend and provide support for utmpd backend. * login/login.c: Use `updwtmp' function insteead of writing the record ourself. * login/logwtmp.c: Move `updwtmp' function to... * login/updwtmp.c: ...here. New file. * login/utmp_db.h: Removed. * login/utmp_file.c: Add updwtmp function to write to file. * login/utmp_daemon.c: New file. Daemon backend. * login/utmpname.c: New file. Implementation of utmpname function. * login/utmpdump.c: New file. Tool to dump utmp-like files. * login/utmpd/connection.c: New file. * login/utmpd/database.c: New file. * login/utmpd/error.c: New file. * login/utmpd/request.c: New file. * login/utmpd/utmpd-private.h: New file. * login/utmpd/utmpd.c: New file. * login/utmpd/utmpd.h: New file. * login/utmpd/xtmp.c: New file. * login/utmpd/xtmp.h: New file. 1997-05-29 12:28 Jim Meyering <meyering@eng.ascend.com> * time/strftime.c: Correct/normalize indentation in cpp directives. 1997-05-28 20:43 Philip Blundell <pjb27@cam.ac.uk> * nis/nis_error.c: Include <string.h> to fix warning. * nis/nis_print.c: Likewise. * nis/nss_nisplus/nisplus-hosts.c: Arg 3 of map_v4v6_hostent is int* not size_t*. 1997-05-28 21:56 Andreas Jaeger <aj@arthur.rhein-neckar.de> * math/cmathcalls.h: Correct typo in comment. * inet/netinet/icmp6.h: Include <netinet/in.h> for in6_addr. * sysdeps/unix/sysv/linux/netinet/ip_fw.h: Include <net/if.h> for IFNAMSIZ. * sysdeps/unix/sysv/linux/net/ppp_defs.h: Include <time.h> for time_t. * login/pty.h: Include <ioctl-types.h> for definition of struct winsize. * misc/regexp.h (compile): Correct typo. * argp/argp.h: Put extern before __const in defintion of argp_program_bug_address. 1997-05-29 00:20 Ulrich Drepper <drepper@cygnus.com> * sysdeps/wordsize-32/inttypes.h: Correct names of unsigned fast and least types. Correct names of ?INT_FAST*_{MIN,MAX} macros. * sysdeps/wordsize-64/inttypes.h: Likewise. Reported by Andreas Jaeger <aj@arthur.rhein-neckar.de>. 1997-05-28 22:51 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/Makefile (make-ioctls-CFLAGS): Use generic ttydefaults.h file instead of non-existing version in termios/sys. Reported by Zack Weinberg <zack@rabi.phys.columbia.edu>. * time/strptime.c (strptime_internal, case 'Y'): Restrict year number to four digits and to representable range for 4 byte time_t values. Patch by H.J. Lu <hjl@lucon.org>. 1997-05-28 18:19 Philip Blundell <pjb27@cam.ac.uk> * posix/execl.c: Include <alloca.h> to avoid warning. 1997-05-27 18:19 Andreas Jaeger <aj@arthur.rhein-neckar.de> * math/libm-test.c: Implement testing of inlined functions, make output nicer, update comments. * math/test-idouble.c: New file. Frontend for double tests of inlined functions. * math/test-ildoubl.c: New file. Frontend for long double tests of inlined functions. * math/test-ifloat.c: New file. Frontend for float tests of inlined functions. * math/test-longdouble.c: Rename to... * math/test-ldouble.c: ...this. * math/Makefile: Add rules for new test programs, change rules for renaming of longdouble test. 1997-05-20 15:50 H.J. Lu <hjl@gnu.ai.mit.edu> * sunrpc/rpc/svc.h (__dispatch_fn_t): New. (svc_register): Use __dispatch_fn_t in prototype. 1997-05-28 17:02 Ulrich Drepper <drepper@cygnus.com> * sysdeps/generic/bzero.c (bzero): Fix typo. Patch by Witek Wnuk <spider@pest.waw.ids.edu.pl>. 1997-05-27 12:00 Andreas Jaeger <aj@arthur.rhein-neckar.de> * sysdeps/generic/vtimes.c: Use ISO C declaration style. * sysdeps/unix/bsd/ualarm.c: Include <unistd.h> for prototype. * sysdeps/generic/memccpy.c: Include <string.h> for prototype. * signal/tst-signal.c (handler): Correct function declaration to avoid warning. * stdlib/testsort.c (compare): Likewise. * string/tester.c: Likewise. 1997-05-27 14:16 Miles Bader <miles@gnu.ai.mit.edu> * argp-help.c (argp_args_usage): Supply correct argp to filter_doc. 1997-05-27 17:51 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * db/hash/extern.h, db/hash/hash.c, db/hash/hash.h, db/hash/hash_log2.c: Rename __log2 to __hash_log2 to avoid clash with libm. 1997-05-27 14:47 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/m68k/fpu/e_atan2.c: Fix missing negate. Use __m81_test instead of explicit comparisons. 1997-05-26 18:36 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * inet/netinet/icmp6.h: Remove use of <asm/bitops.h> which has no place in a generic header and is no user include file.
Diffstat (limited to 'math')
-rw-r--r--math/Makefile12
-rw-r--r--math/cmathcalls.h2
-rw-r--r--math/libm-test.c307
-rw-r--r--math/test-idouble.c33
-rw-r--r--math/test-ifloat.c32
-rw-r--r--math/test-ildoubl.c33
-rw-r--r--math/test-ldouble.c (renamed from math/test-longdouble.c)0
7 files changed, 344 insertions, 75 deletions
diff --git a/math/Makefile b/math/Makefile
index c2f3274877..e0c295b123 100644
--- a/math/Makefile
+++ b/math/Makefile
@@ -74,19 +74,23 @@ long-c-yes = $(calls:=l)
distribute += $(long-c-yes:=.c)
# Rules for the test suite.
-tests = test-float test-double $(test-longdouble-$(long-double-fcts))
+tests = test-float test-double $(test-longdouble-$(long-double-fcts)) \
+ test-ifloat test-idouble
# We do the `long double' tests only if this data type is available and
# distrinct from `double'.
#
# XXX This test is disabled for now since the functions are too buggy.
-#test-longdouble-yes = test-longdouble
+#test-longdouble-yes = test-ldouble test-ildoubl
CFLAGS-test-float.c = -fno-inline
CFLAGS-test-double.c = -fno-inline
-CFLAGS-test-longdouble.c = -fno-inline
+CFLAGS-test-ldouble.c = -fno-inline
+LDLIBS-test-ifloat = libm
+LDLIBS-test-idouble = libm
+LDLIBS-test-ildoubl = libm
LDLIBS-test-float = libm
LDLIBS-test-double = libm
-LDLIBS-test-longdouble = libm
+LDLIBS-test-ldouble = libm
distribute += libm-test.c
diff --git a/math/cmathcalls.h b/math/cmathcalls.h
index 73a1c2a5da..1f74f56e0b 100644
--- a/math/cmathcalls.h
+++ b/math/cmathcalls.h
@@ -18,7 +18,7 @@
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
-/* NOTE: Because of the special way this file is used by <math.h>, this
+/* NOTE: Because of the special way this file is used by <complex.h>, this
file must NOT be protected from multiple inclusion as header files
usually are.
diff --git a/math/libm-test.c b/math/libm-test.c
index 1f64d1ad24..866313e963 100644
--- a/math/libm-test.c
+++ b/math/libm-test.c
@@ -1,6 +1,6 @@
/* Copyright (C) 1997 Free Software Foundation, Inc.
This file is part of the GNU C Library.
- Contributed by Andreas Jaeger <aj@arthur.pfalz.de>, 1997.
+ Contributed by Andreas Jaeger <aj@arthur.rhein-neckar.de>, 1997.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Library General Public License as
@@ -39,30 +39,67 @@
*/
/* This program isn't finished yet.
- It has tests for acos, acosh, asin, asinh, atan, atan2, atanh,
+ It has tests for:
+ acos, acosh, asin, asinh, atan, atan2, atanh,
cbrt, ceil, copysign, cos, cosh, exp, exp2, expm1,
fabs, fdim, floor, fmin, fmax, fpclassify,
frexp, hypot, ilogb, ldexp,
log, log10, log1p, log2, logb, modf, nextafter,
- pow, scalb, scalbn, sin, sinh, sqrt, tan, tanh, trunc.
- Tests for the other libm-functions will come later.
+ pow, rint, rinttol, rinttoll, round, roundtol, roundtoll,
+ scalb, scalbn, sin, sinh, sqrt, tan, tanh, trunc
+
+ and for the following complex math functions:
+ cacos, cacosh, casin, casinh, catan, catanh,
+ ccos, ccosh, cexp, clog, cpow, csin, csinh, csqrt, ctanh.
+
+ At the moment the following functions aren't tested:
+ cabs, carg, conj, cproj, cimag, creal, drem,
+ erf, erfc, gamma, lgamma,
+ j0, j1, jn, y0, y1, yn,
+ nearbyint, remainder, remquo, signbit, significant, sincos.
The routines using random variables are still under construction. I don't
like it the way it's working now and will change it.
- Exception handling has not been implemented so far so don't get fooled
- that these tests pass.
-
Parameter handling is primitive in the moment:
- --verbose=[0..3] for different levels of output:
+ --verbose=[0..4] for different levels of output:
0: only error count
- 1: basic report on failed tests
+ 1: basic report on failed tests (default)
2: full report on failed tests
- 3: full report on failed and passed tests (default)
- -v for full output (equals --verbose=3)
+ 3: full report on failed and passed tests
+ 4: additional report on exceptions
+ -v for full output (equals --verbose=4)
-s,--silent outputs only the error count (equals --verbose=0)
*/
+/* "Philosophy":
+
+ This suite tests the correct implementation of mathematical
+ functions in libm. Some simple, specific parameters are tested for
+ correctness. 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 C 9X.
+
+ 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.
+*/
+
#ifndef _GNU_SOURCE
# define _GNU_SOURCE
#endif
@@ -75,11 +112,8 @@
#include <errno.h>
#include <stdlib.h>
#include <stdio.h>
-#include <time.h>
#include <getopt.h>
-/* TEST_EXCEPTION: tests if an exception as occured */
-/* for the moment: does nothing */
/* Possible exceptions */
#define NO_EXCEPTION 0x0
#define INVALID_EXCEPTION 0x1
@@ -167,11 +201,39 @@ random_greater (MATHTYPE min_value)
}
/* Get a random value x with x < max_value. */
+#ifndef TEST_INLINE
static MATHTYPE
random_less (MATHTYPE max_value)
{
return random_value (-1e6, max_value);
}
+#endif
+
+
+static void
+output_new_test (const char *test_name)
+{
+ if (verbose > 2)
+ printf ("\nTesting: %s\n", test_name);
+}
+
+
+static void
+output_pass_value (void)
+{
+ if (verbose > 2)
+ printf ("Pass: Value Ok.\n");
+}
+
+
+static void
+output_fail_value (const char * test_name)
+{
+ if (verbose > 0 && verbose < 3)
+ printf ("Fail: %s\n", test_name);
+ if (verbose >= 3)
+ printf ("Fail:\n");
+}
/* Test whether a given exception was raised. */
@@ -182,18 +244,22 @@ test_single_exception (const char *test_name,
fexcept_t fe_flag,
const char *flag_name)
{
+#ifndef TEST_INLINE
if (exception & exc_flag)
{
if (fetestexcept (fe_flag))
{
- if (verbose > 2)
- printf ("Pass: %s:\nException \"%s\" set\n", test_name, flag_name);
+ if (verbose > 3)
+ printf ("Pass: Exception \"%s\" set\n", flag_name);
}
else
{
- if (verbose)
- printf ("Fail: %s:\nException \"%s\" not set\n",
+ if (verbose && verbose < 3)
+ printf ("Fail: %s: Exception \"%s\" not set\n",
test_name, flag_name);
+ if (verbose >= 3)
+ printf ("Fail: Exception \"%s\" not set\n",
+ flag_name);
++noErrors;
}
}
@@ -201,18 +267,22 @@ test_single_exception (const char *test_name,
{
if (fetestexcept (fe_flag))
{
- if (verbose)
- printf ("Fail: %s:\nException \"%s\" set\n",
+ if (verbose && verbose < 3)
+ printf ("Fail: %s: Exception \"%s\" set\n",
test_name, flag_name);
+ if (verbose >= 3)
+ printf ("Fail: Exception \"%s\" set\n",
+ flag_name);
++noErrors;
}
else
{
- if (verbose > 2)
- printf ("Pass: %s:\nException \"%s\" not set\n",
- test_name, flag_name);
+ if (verbose > 3)
+ printf ("Pass: Exception \"%s\" not set\n",
+ flag_name);
}
}
+#endif
}
@@ -256,6 +326,8 @@ test_exceptions (const char *test_name, short int exception)
static int
check_equal (MATHTYPE computed, MATHTYPE supplied, MATHTYPE eps, MATHTYPE * diff)
{
+ int ret_value;
+
/* Both plus Infinity or both minus infinity. */
if (ISINF (computed) && (ISINF (computed) == ISINF (supplied)))
return 1;
@@ -265,25 +337,28 @@ check_equal (MATHTYPE computed, MATHTYPE supplied, MATHTYPE eps, MATHTYPE * diff
*diff = FUNC(fabs) (computed - supplied);
- if (*diff <= eps && (signbit (computed) == signbit (supplied) || eps != 0.0))
- return 1;
- return 0;
+ ret_value = (*diff <= eps &&
+ (signbit (computed) == signbit (supplied) || eps != 0.0));
+
+ /* Make sure the subtraction/comparsion have no influence on the exceptions. */
+ feclearexcept (FE_ALL_EXCEPT);
+
+ return ret_value;
}
+
static void
output_result_bool (const char *test_name, int result)
{
if (result)
{
- if (verbose > 2)
- printf ("Pass: %s\n", test_name);
+ output_pass_value ();
}
else
{
- if (verbose)
- printf ("Fail: %s\n", test_name);
+ output_fail_value (test_name);
++noErrors;
}
@@ -297,13 +372,11 @@ output_isvalue (const char *test_name, int result,
{
if (result)
{
- if (verbose > 2)
- printf ("Pass: %s\n", test_name);
+ output_pass_value ();
}
else
{
- if (verbose)
- printf ("Fail: %s\n", test_name);
+ output_fail_value (test_name);
if (verbose > 1)
printf (" Value: %.20" PRINTF_EXPR "\n", value);
noErrors++;
@@ -319,13 +392,11 @@ output_isvalue_ext (const char *test_name, int result,
{
if (result)
{
- if (verbose > 2)
- printf ("Pass: %s\n", test_name);
+ output_pass_value ();
}
else
{
- if (verbose)
- printf ("Fail: %s\n", test_name);
+ output_fail_value (test_name);
if (verbose > 1)
{
printf (" Value: %.20" PRINTF_EXPR "\n", value);
@@ -346,13 +417,11 @@ output_result (const char *test_name, int result,
{
if (result)
{
- if (verbose > 2)
- printf ("Pass: %s\n", test_name);
+ output_pass_value ();
}
else
{
- if (verbose)
- printf ("Fail: %s\n", test_name);
+ output_fail_value (test_name);
if (verbose > 1 && print_values)
{
printf ("Result:\n");
@@ -377,13 +446,11 @@ output_result_ext (const char *test_name, int result,
{
if (result)
{
- if (verbose > 2)
- printf ("Pass: %s\n", test_name);
+ output_pass_value ();
}
else
{
- if (verbose)
- printf ("Fail: %s\n", test_name);
+ output_fail_value (test_name);
if (verbose > 1 && print_values)
{
printf ("Result:\n");
@@ -399,13 +466,16 @@ output_result_ext (const char *test_name, int result,
fpstack_test (test_name);
}
-
+/*
+ check that computed and expected values are the same
+ */
static void
check (const char *test_name, MATHTYPE computed, MATHTYPE expected)
{
MATHTYPE diff;
int result;
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
result = check_equal (computed, expected, 0, &diff);
output_result (test_name, result,
@@ -413,6 +483,10 @@ check (const char *test_name, MATHTYPE computed, MATHTYPE expected)
}
+/*
+ check that computed and expected values are the same,
+ outputs the parameter to the function
+ */
static void
check_ext (const char *test_name, MATHTYPE computed, MATHTYPE expected,
MATHTYPE parameter)
@@ -420,6 +494,7 @@ check_ext (const char *test_name, MATHTYPE computed, MATHTYPE expected,
MATHTYPE diff;
int result;
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
result = check_equal (computed, expected, 0, &diff);
output_result_ext (test_name, result,
@@ -427,6 +502,10 @@ check_ext (const char *test_name, MATHTYPE computed, MATHTYPE expected,
}
+/*
+ check that computed and expected values are the same and
+ checks also for exception flags
+ */
static void
check_exc (const char *test_name, MATHTYPE computed, MATHTYPE expected,
short exception)
@@ -434,13 +513,16 @@ check_exc (const char *test_name, MATHTYPE computed, MATHTYPE expected,
MATHTYPE diff;
int result;
+ output_new_test (test_name);
test_exceptions (test_name, exception);
result = check_equal (computed, expected, 0, &diff);
output_result (test_name, result,
computed, expected, diff, PRINT, PRINT);
}
-
+/*
+ check that computed and expected values are close enough
+ */
static void
check_eps (const char *test_name, MATHTYPE computed, MATHTYPE expected,
MATHTYPE epsilon)
@@ -448,38 +530,43 @@ check_eps (const char *test_name, MATHTYPE computed, MATHTYPE expected,
MATHTYPE diff;
int result;
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
result = check_equal (computed, expected, epsilon, &diff);
output_result (test_name, result,
computed, expected, diff, PRINT, PRINT);
}
-
+/*
+ check a boolean condition
+ */
static void
check_bool (const char *test_name, int computed)
{
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
output_result_bool (test_name, computed);
}
-
+/*
+ check that computed and expected values are equal (long int values)
+ */
static void
check_long (const char *test_name, long int computed, long int expected)
{
long int diff = computed - expected;
int result = diff == 0;
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
if (result)
{
- if (verbose > 2)
- printf ("Pass: %s\n", test_name);
+ output_pass_value ();
}
else
{
- if (verbose)
- printf ("Fail: %s\n", test_name);
+ output_fail_value (test_name);
if (verbose > 1)
{
printf ("Result:\n");
@@ -492,7 +579,9 @@ check_long (const char *test_name, long int computed, long int expected)
fpstack_test (test_name);
}
-
+/*
+ check that computed and expected values are equal (long long int values)
+ */
static void
check_longlong (const char *test_name, long long int computed,
long long int expected)
@@ -500,17 +589,16 @@ check_longlong (const char *test_name, long long int computed,
long long int diff = computed - expected;
int result = diff == 0;
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
if (result)
{
- if (verbose > 2)
- printf ("Pass: %s\n", test_name);
+ output_pass_value ();
}
else
{
- if (verbose)
- printf ("Fail: %s\n", test_name);
+ output_fail_value (test_name);
if (verbose > 1)
{
printf ("Result:\n");
@@ -523,46 +611,64 @@ check_longlong (const char *test_name, long long int computed,
fpstack_test (test_name);
}
-
+/*
+ check that computed value is not-a-number
+ */
static void
check_isnan (const char *test_name, MATHTYPE computed)
{
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
output_isvalue (test_name, isnan (computed), computed);
}
+/*
+ check that computed value is not-a-number and test for exceptions
+ */
static void
check_isnan_exc (const char *test_name, MATHTYPE computed,
short exception)
{
+ output_new_test (test_name);
test_exceptions (test_name, exception);
output_isvalue (test_name, isnan (computed), computed);
}
+/*
+ check that computed value is not-a-number and test for exceptions
+ */
static void
check_isnan_maybe_exc (const char *test_name, MATHTYPE computed,
short exception)
{
+ output_new_test (test_name);
test_not_exception (test_name, exception);
output_isvalue (test_name, isnan (computed), computed);
}
+/*
+ check that computed value is not-a-number and supply parameter
+ */
+#ifndef TEST_INLINE
static void
check_isnan_ext (const char *test_name, MATHTYPE computed,
MATHTYPE parameter)
{
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
output_isvalue_ext (test_name, isnan (computed), computed, parameter);
}
+#endif
/* Tests if computed is +Inf */
static void
check_isinfp (const char *test_name, MATHTYPE computed)
{
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
output_isvalue (test_name, (ISINF (computed) == +1), computed);
}
@@ -572,6 +678,7 @@ static void
check_isinfp_ext (const char *test_name, MATHTYPE computed,
MATHTYPE parameter)
{
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
output_isvalue_ext (test_name, (ISINF (computed) == +1), computed, parameter);
}
@@ -582,6 +689,7 @@ static void
check_isinfp_exc (const char *test_name, MATHTYPE computed,
int exception)
{
+ output_new_test (test_name);
test_exceptions (test_name, exception);
output_isvalue (test_name, (ISINF (computed) == +1), computed);
}
@@ -590,18 +698,22 @@ check_isinfp_exc (const char *test_name, MATHTYPE computed,
static void
check_isinfn (const char *test_name, MATHTYPE computed)
{
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
output_isvalue (test_name, (ISINF (computed) == -1), computed);
}
+#ifndef TEST_INLINE
static void
check_isinfn_ext (const char *test_name, MATHTYPE computed,
MATHTYPE parameter)
{
+ output_new_test (test_name);
test_exceptions (test_name, NO_EXCEPTION);
output_isvalue_ext (test_name, (ISINF (computed) == -1), computed, parameter);
}
+#endif
/* Tests if computed is -Inf */
@@ -609,6 +721,7 @@ static void
check_isinfn_exc (const char *test_name, MATHTYPE computed,
int exception)
{
+ output_new_test (test_name);
test_exceptions (test_name, exception);
output_isvalue (test_name, (ISINF (computed) == -1), computed);
}
@@ -621,48 +734,59 @@ check_isinfn_exc (const char *test_name, MATHTYPE computed,
static void
acos_test (void)
{
+#ifndef TEST_INLINE
MATHTYPE x;
- check ("acos (1) == 0", FUNC(acos) (1), 0);
-
x = random_greater (1);
check_isnan_exc ("acos (x) == NaN plus invalid exception for |x| > 1",
FUNC(acos) (x),
INVALID_EXCEPTION);
+#endif
+
+ check ("acos (1) == 0", FUNC(acos) (1), 0);
}
static void
acosh_test (void)
{
+#ifndef TEST_INLINE
MATHTYPE x;
- check ("acosh(1) == 0", FUNC(acosh) (1), 0);
check_isinfp ("acosh(+inf) == +inf", FUNC(acosh) (plus_infty));
x = random_less (1);
check_isnan_exc ("acosh(x) == NaN plus invalid exception if x < 1",
FUNC(acosh) (x), INVALID_EXCEPTION);
+#endif
+
+ check ("acosh(1) == 0", FUNC(acosh) (1), 0);
}
static void
asin_test (void)
{
+#ifndef TEST_INLINE
MATHTYPE x;
- check ("asin (0) == 0", FUNC(asin) (0), 0);
x = random_greater (1);
check_isnan_exc ("asin x == NaN plus invalid exception for |x| > 1",
FUNC(asin) (x),
INVALID_EXCEPTION);
+#endif
+
+ check ("asin (0) == 0", FUNC(asin) (0), 0);
}
+
static void
asinh_test (void)
{
check ("asinh(+0) == +0", FUNC(asinh) (0), 0);
+#ifndef TEST_INLINE
check ("asinh(-0) == -0", FUNC(asinh) (minus_zero), minus_zero);
+#endif
}
@@ -674,7 +798,6 @@ atan_test (void)
check ("atan (+inf) == pi/2", FUNC(atan) (plus_infty), M_PI_2);
check ("atan (-inf) == -pi/2", FUNC(atan) (minus_infty), -M_PI_2);
-
}
static void
@@ -734,11 +857,14 @@ atanh_test (void)
{
check ("atanh(+0) == +0", FUNC(atanh) (0), 0);
+#ifndef TEST_INLINE
check ("atanh(-0) == -0", FUNC(atanh) (minus_zero), minus_zero);
+
check_isinfp_exc ("atanh(+1) == +inf plus divide-by-zero exception",
FUNC(atanh) (1), DIVIDE_BY_ZERO_EXCEPTION);
check_isinfn_exc ("atanh(-1) == -inf plus divide-by-zero exception",
FUNC(atanh) (-1), DIVIDE_BY_ZERO_EXCEPTION);
+#endif
}
@@ -748,10 +874,11 @@ cbrt_test (void)
check ("cbrt (+0) == +0", FUNC(cbrt) (0.0), 0.0);
check ("cbrt (-0) == -0", FUNC(cbrt) (minus_zero), minus_zero);
+#ifndef TEST_INLINE
check_isinfp ("cbrt (+inf) == +inf", FUNC(cbrt) (plus_infty));
check_isinfn ("cbrt (-inf) == -inf", FUNC(cbrt) (minus_infty));
check_isnan ("cbrt (NaN) == NaN", FUNC(cbrt) (nan_value));
-
+#endif
check_eps ("cbrt (8) == 2", FUNC(cbrt) (8), 2, CHOOSE (5e-17L, 0, 0));
check_eps ("cbrt (-27) == -3", FUNC(cbrt) (-27.0), -3.0,
CHOOSE (3e-16L, 0, 0));
@@ -797,8 +924,10 @@ cosh_test (void)
check ("cosh (+0) == 1", FUNC(cosh) (0), 1);
check ("cosh (-0) == 1", FUNC(cosh) (minus_zero), 1);
+#ifndef TEST_INLINE
check_isinfp ("cosh (+inf) == +inf", FUNC(cosh) (plus_infty));
check_isinfp ("cosh (-inf) == +inf", FUNC(cosh) (minus_infty));
+#endif
}
@@ -808,9 +937,10 @@ exp_test (void)
check ("exp (+0) == 1", FUNC(exp) (0), 1);
check ("exp (-0) == 1", FUNC(exp) (minus_zero), 1);
+#ifndef TEST_INLINE
check_isinfp ("exp (+inf) == +inf", FUNC(exp) (plus_infty));
check ("exp (-inf) == 0", FUNC(exp) (minus_infty), 0);
-
+#endif
check_eps ("exp (1) == e", FUNC(exp) (1), M_E, CHOOSE (4e-18L, 0, 0));
}
@@ -1207,10 +1337,13 @@ static void
sinh_test (void)
{
check ("sinh (+0) == +0", FUNC(sinh) (0), 0);
+
+#ifndef TEST_INLINE
check ("sinh (-0) == -0", FUNC(sinh) (minus_zero), minus_zero);
check_isinfp ("sinh (+inf) == +inf", FUNC(sinh) (plus_infty));
check_isinfn ("sinh (-inf) == -inf", FUNC(sinh) (minus_infty));
+#endif
}
@@ -1233,10 +1366,12 @@ static void
tanh_test (void)
{
check ("tanh (+0) == +0", FUNC(tanh) (0), 0);
+#ifndef TEST_INLINE
check ("tanh (-0) == -0", FUNC(tanh) (minus_zero), minus_zero);
check ("tanh (+inf) == +1", FUNC(tanh) (plus_infty), 1);
check ("tanh (-inf) == -1", FUNC(tanh) (minus_infty), -1);
+#endif
}
@@ -1311,6 +1446,7 @@ pow_test (void)
check ("pow (NaN, +0) == 1", FUNC(pow) (nan_value, 0), 1);
check ("pow (NaN, -0) == 1", FUNC(pow) (nan_value, minus_zero), 1);
+#ifndef TEST_INLINE
check_isinfp ("pow (+1.1, +inf) == +inf", FUNC(pow) (1.1, plus_infty));
check_isinfp ("pow (+inf, +inf) == +inf", FUNC(pow) (plus_infty, plus_infty));
check_isinfp ("pow (-1.1, +inf) == +inf", FUNC(pow) (-1.1, plus_infty));
@@ -1408,14 +1544,19 @@ pow_test (void)
FUNC(pow) (minus_zero, -2), DIVIDE_BY_ZERO_EXCEPTION);
check_isinfp_exc ("pow (-0, -11.1) == +inf plus divide-by-zero exception",
FUNC(pow) (minus_zero, -11.1), DIVIDE_BY_ZERO_EXCEPTION);
+#endif
check ("pow (+0, 1) == +0", FUNC(pow) (0, 1), 0);
check ("pow (+0, 11) == +0", FUNC(pow) (0, 11), 0);
+#ifndef TEST_INLINE
check ("pow (-0, 1) == -0", FUNC(pow) (minus_zero, 1), minus_zero);
check ("pow (-0, 11) == -0", FUNC(pow) (minus_zero, 11), minus_zero);
+#endif
check ("pow (+0, 2) == +0", FUNC(pow) (0, 2), 0);
check ("pow (+0, 11.1) == +0", FUNC(pow) (0, 11.1), 0);
+
+#ifndef TEST_INLINE
check ("pow (-0, 2) == +0", FUNC(pow) (minus_zero, 2), 0);
check ("pow (-0, 11.1) == +0", FUNC(pow) (minus_zero, 11.1), 0);
@@ -1425,11 +1566,11 @@ pow_test (void)
x = random_value (-1.0, 1.0);
check_ext ("pow (x, +inf) == +0 for |x| < 1",
- FUNC(pow) (x, plus_infty), 0.0, x);
+ FUNC(pow) (x, plus_infty), 0.0, x);
x = random_greater (1.0);
check_ext ("pow (x, -inf) == +0 for |x| > 1",
- FUNC(pow) (x, minus_infty), 0.0, x);
+ FUNC(pow) (x, minus_infty), 0.0, x);
x = random_value (-1.0, 1.0);
check_isinfp_ext ("pow (x, -inf) == +inf for |x| < 1",
@@ -1458,13 +1599,16 @@ pow_test (void)
x = ((rand () % 1000000) + 1) * -2.0; /* Get random even integer < 0 */
check_ext ("pow (-inf, y) == +0 for y < 0 and not an odd integer",
FUNC(pow) (minus_infty, x), 0.0, x);
+#endif
x = (rand () % 1000000) * 2.0 + 1; /* Get random odd integer > 0 */
check_ext ("pow (+0, y) == +0 for y an odd integer > 0",
FUNC(pow) (0.0, x), 0.0, x);
+#ifndef TEST_INLINE
x = (rand () % 1000000) * 2.0 + 1; /* Get random odd integer > 0 */
check_ext ("pow (-0, y) == -0 for y an odd integer > 0",
FUNC(pow) (minus_zero, x), minus_zero, x);
+#endif
x = ((rand () % 1000000) + 1) * 2.0; /* Get random even integer > 1 */
check_ext ("pow (+0, y) == +0 for y > 0 and not an odd integer",
@@ -1682,6 +1826,7 @@ sqrt_test (void)
x = random_value (0, 10000);
check_ext ("sqrt (x*x) == x", FUNC(sqrt) (x*x), x, x);
check ("sqrt (4) == 2", FUNC(sqrt) (4), 2);
+
}
@@ -1741,11 +1886,18 @@ cexp_test (void)
check ("real(cexp(-inf - 0i)) = 0", __real__ result, 0);
check ("imag(cexp(-inf - 0i)) = -0", __imag__ result, minus_zero);
+
result = FUNC(cexp) (BUILD_COMPLEX (0.0, plus_infty));
check_isnan_exc ("real(cexp(0 + i inf)) = NaN plus invalid exception",
__real__ result, INVALID_EXCEPTION);
check_isnan ("imag(cexp(0 + i inf)) = NaN plus invalid exception",
__imag__ result);
+
+#if defined __GNUC__ && __GNUC__ <= 2 && __GNUC_MINOR <= 7
+ if (verbose)
+ printf ("The following test for cexp might fail due to a gcc compiler error!\n");
+#endif
+
result = FUNC(cexp) (BUILD_COMPLEX (minus_zero, plus_infty));
check_isnan_exc ("real(cexp(-0 + i inf)) = NaN plus invalid exception",
__real__ result, INVALID_EXCEPTION);
@@ -3757,6 +3909,15 @@ static void
cpow_test (void)
{
__complex__ MATHTYPE result;
+
+ result = FUNC (cpow) (BUILD_COMPLEX (1, 0), BUILD_COMPLEX (0, 0));
+ check ("real(cpow (1 + i0), (0 + i0)) = 0", __real__ result, 1);
+ check ("imag(cpow (1 + i0), (0 + i0)) = 0", __imag__ result, 0);
+
+ result = FUNC (cpow) (BUILD_COMPLEX (2, 0), BUILD_COMPLEX (10, 0));
+ check ("real(cpow (2 + i0), (10 + i0)) = 1024", __real__ result, 1024);
+ check ("imag(cpow (2 + i0), (10 + i0)) = 0", __imag__ result, 0);
+
}
@@ -3887,6 +4048,7 @@ inverse_func_pair_test (const char *test_name,
b = inverse (a);
(void) &b;
+ output_new_test (test_name);
result = check_equal (b, x, epsilon, &difference);
output_result (test_name, result,
b, x, difference, PRINT, PRINT);
@@ -3941,6 +4103,7 @@ identities1_test (MATHTYPE x, MATHTYPE epsilon)
res3 = res1 * res1 + res2 * res2;
(void) &res3;
+ output_new_test ("sin^2 + cos^2 == 1");
result = check_equal (res3, 1.0, epsilon, &diff);
output_result_ext ("sin^2 + cos^2 == 1", result,
res3, 1.0, diff, x, PRINT, PRINT);
@@ -3963,6 +4126,7 @@ identities2_test (MATHTYPE x, MATHTYPE epsilon)
res4 = res1 / res2;
(void) &res4;
+ output_new_test ("sin/cos == tan");
result = check_equal (res4, res3, epsilon, &diff);
output_result_ext ("sin/cos == tan", result,
res4, res3, diff, x, PRINT, PRINT);
@@ -3983,6 +4147,7 @@ identities3_test (MATHTYPE x, MATHTYPE epsilon)
res3 = res2 * res2 - res1 * res1;
(void) &res3;
+ output_new_test ("cosh^2 - sinh^2 == 1");
result = check_equal (res3, 1.0, epsilon, &diff);
output_result_ext ("cosh^2 - sinh^2 == 1", result,
res3, 1.0, diff, x, PRINT, PRINT);
@@ -4142,7 +4307,7 @@ parse_options (int argc, char *argv[])
if (optarg)
verbose = (unsigned int) strtoul (optarg, NULL, 0);
else
- verbose = 3;
+ verbose = 4;
break;
case 's':
verbose = 0;
@@ -4156,10 +4321,12 @@ parse_options (int argc, char *argv[])
int
main (int argc, char *argv[])
{
+
parse_options (argc, argv);
initialize ();
printf (TEST_MSG);
+
basic_tests ();
acos_test ();
diff --git a/math/test-idouble.c b/math/test-idouble.c
new file mode 100644
index 0000000000..614f10feb4
--- /dev/null
+++ b/math/test-idouble.c
@@ -0,0 +1,33 @@
+/* Copyright (C) 1997 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Andreas Jaeger <aj@arthur.rhein-neckar.de>, 1997.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Library General Public License as
+ published by the Free Software Foundation; either version 2 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
+ Library General Public License for more details.
+
+
+ You should have received a copy of the GNU Library General Public
+ License along with the GNU C Library; see the file COPYING.LIB. If not,
+ write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ Boston, MA 02111-1307, USA. */
+
+#define FUNC(function) function
+#define MATHTYPE double
+#define TEST_MSG "testing double (inline functions)\n"
+#define MATHCONST(x) x
+#define CHOOSE(Clongdouble,Cdouble,Cfloat) Cdouble
+#define PRINTF_EXPR "e"
+#define TEST_INLINE
+
+#ifdef __NO_MATH_INLINES
+# undef __NO_MATH_INLINES
+#endif
+
+#include "libm-test.c"
diff --git a/math/test-ifloat.c b/math/test-ifloat.c
new file mode 100644
index 0000000000..05e5de62ff
--- /dev/null
+++ b/math/test-ifloat.c
@@ -0,0 +1,32 @@
+/* Copyright (C) 1997 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Andreas Jaeger <aj@arthur.rhein-neckar.de>, 1997.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Library General Public License as
+ published by the Free Software Foundation; either version 2 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
+ Library General Public License for more details.
+
+ You should have received a copy of the GNU Library General Public
+ License along with the GNU C Library; see the file COPYING.LIB. If not,
+ write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ Boston, MA 02111-1307, USA. */
+
+#define FUNC(function) function ## f
+#define MATHTYPE float
+#define TEST_MSG "testing float (inline functions)\n"
+#define MATHCONST(x) x
+#define CHOOSE(Clongdouble,Cdouble,Cfloat) Cfloat
+#define PRINTF_EXPR "e"
+#define TEST_INLINE
+
+#ifdef __NO_MATH_INLINES
+# undef __NO_MATH_INLINES
+#endif
+
+#include "libm-test.c"
diff --git a/math/test-ildoubl.c b/math/test-ildoubl.c
new file mode 100644
index 0000000000..791cb95cf3
--- /dev/null
+++ b/math/test-ildoubl.c
@@ -0,0 +1,33 @@
+/* Copyright (C) 1997 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Andreas Jaeger <aj@arthur.rhein-neckar.de>, 1997.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Library General Public License as
+ published by the Free Software Foundation; either version 2 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
+ Library General Public License for more details.
+
+
+ You should have received a copy of the GNU Library General Public
+ License along with the GNU C Library; see the file COPYING.LIB. If not,
+ write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ Boston, MA 02111-1307, USA. */
+
+#define FUNC(function) function##l
+#define MATHTYPE long double
+#define TEST_MSG "testing long double (inline functions)\n"
+#define MATHCONST(x) x##L
+#define CHOOSE(Clongdouble,Cdouble,Cfloat) Clongdouble
+#define PRINTF_EXPR "Le"
+#define TEST_INLINE
+
+#ifdef __NO_MATH_INLINES
+# undef __NO_MATH_INLINES
+#endif
+
+#include "libm-test.c"
diff --git a/math/test-longdouble.c b/math/test-ldouble.c
index eb9b57ae96..eb9b57ae96 100644
--- a/math/test-longdouble.c
+++ b/math/test-ldouble.c