mirror of
https://sourceware.org/git/glibc.git
synced 2024-11-21 20:40:05 +00:00
1b97a9f23b
Add a FAIL test failure helper analogous to FAIL_RET, that does not cause the current function to return, providing a standardized way to report a test failure with a message supplied while permitting the caller to continue executing, for further reporting, cleaning up, etc. Update existing test cases that provide a conflicting definition of FAIL by removing the local FAIL definition and then as follows: - tst-fortify-syslog: provide a meaningful message in addition to the file name already added by <support/check.h>; 'support_record_failure' is already called by 'support_print_failure_impl' invoked by the new FAIL test failure helper. - tst-ctype: no update to FAIL calls required, with the name of the file and the line number within of the failure site additionally included by the new FAIL test failure helper, and error counting plus count reporting upon test program termination also already provided by 'support_record_failure' and 'support_report_failure' respectively, called by 'support_print_failure_impl' and 'adjust_exit_status' also respectively. However in a number of places 'printf' is called and the error count adjusted by hand, so update these places to make use of FAIL instead. And last but not least adjust the final summary just to report completion, with any error count following as reported by the test driver. - test-tgmath2: no update to FAIL calls required, with the name of the file of the failure site additionally included by the new FAIL test failure helper. Also there is no need to track the return status by hand as any call to FAIL will eventually cause the test case to return an unsuccesful exit status regardless of the return status from the test function, via a call to 'adjust_exit_status' made by the test driver. Reviewed-by: DJ Delorie <dj@redhat.com>
595 lines
15 KiB
C
595 lines
15 KiB
C
/* Test compilation of tgmath macros.
|
|
Copyright (C) 2007-2024 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
|
|
<https://www.gnu.org/licenses/>. */
|
|
|
|
#ifndef HAVE_MAIN
|
|
#include <float.h>
|
|
#include <math.h>
|
|
#include <complex.h>
|
|
#include <stdio.h>
|
|
#include <string.h>
|
|
#include <tgmath.h>
|
|
|
|
#include <support/check.h>
|
|
|
|
//#define DEBUG
|
|
|
|
typedef complex float cfloat;
|
|
typedef complex double cdouble;
|
|
#if LDBL_MANT_DIG > DBL_MANT_DIG
|
|
typedef long double ldouble;
|
|
typedef complex long double cldouble;
|
|
#else
|
|
typedef double ldouble;
|
|
typedef complex double cldouble;
|
|
#endif
|
|
|
|
float vfloat1, vfloat2, vfloat3;
|
|
double vdouble1, vdouble2, vdouble3;
|
|
ldouble vldouble1, vldouble2, vldouble3;
|
|
cfloat vcfloat1, vcfloat2, vcfloat3;
|
|
cdouble vcdouble1, vcdouble2, vcdouble3;
|
|
cldouble vcldouble1, vcldouble2, vcldouble4;
|
|
int vint1, vint2, vint3;
|
|
long int vlong1, vlong2, vlong3;
|
|
long long int vllong1, vllong2, vllong3;
|
|
const float Vfloat1 = 1, Vfloat2 = 2, Vfloat3 = 3;
|
|
const double Vdouble1 = 1, Vdouble2 = 2, Vdouble3 = 3;
|
|
const ldouble Vldouble1 = 1, Vldouble2 = 2, Vldouble3 = 3;
|
|
const cfloat Vcfloat1 = 1, Vcfloat2 = 2, Vcfloat3 = 3;
|
|
const cdouble Vcdouble1 = 1, Vcdouble2 = 2, Vcdouble3 = 3;
|
|
const cldouble Vcldouble1 = 1, Vcldouble2 = 2, Vcldouble4 = 3;
|
|
const int Vint1 = 1, Vint2 = 2, Vint3 = 3;
|
|
const long int Vlong1 = 1, Vlong2 = 2, Vlong3 = 3;
|
|
const long long int Vllong1 = 1, Vllong2 = 2, Vllong3 = 3;
|
|
enum
|
|
{
|
|
Tfloat = 0,
|
|
Tcfloat,
|
|
Tdouble,
|
|
Tcdouble,
|
|
#if LDBL_MANT_DIG > DBL_MANT_DIG
|
|
Tldouble,
|
|
Tcldouble,
|
|
#else
|
|
Tldouble = Tdouble,
|
|
Tcldouble = Tcdouble,
|
|
#endif
|
|
Tlast
|
|
};
|
|
enum
|
|
{
|
|
C_cos = 0,
|
|
C_fabs,
|
|
C_cabs,
|
|
C_conj,
|
|
C_expm1,
|
|
C_lrint,
|
|
C_ldexp,
|
|
C_atan2,
|
|
C_remquo,
|
|
C_pow,
|
|
C_fma,
|
|
C_last
|
|
};
|
|
int count;
|
|
int counts[Tlast][C_last];
|
|
|
|
#define TEST_TYPE_ONLY(expr, rettype) \
|
|
do \
|
|
{ \
|
|
__typeof__ (expr) texpr = 0; \
|
|
__typeof__ (rettype) ttype = 0, *ptype; \
|
|
if (sizeof (expr) != sizeof (rettype)) \
|
|
FAIL ("type"); \
|
|
if (__alignof__ (expr) != __alignof__ (rettype)) \
|
|
FAIL ("type"); \
|
|
__asm ("" : "=r" (ptype) : "0" (&ttype), "r" (&texpr)); \
|
|
if (&texpr == ptype) \
|
|
FAIL ("type"); \
|
|
} \
|
|
while (0)
|
|
#define TEST2(expr, type, rettype, fn) \
|
|
do \
|
|
{ \
|
|
__typeof__ (expr) texpr = 0; \
|
|
TEST_TYPE_ONLY (expr, rettype); \
|
|
if (count != 0) \
|
|
FAIL ("internal error"); \
|
|
if (counts[T##type][C_##fn] != 0) \
|
|
FAIL ("internal error"); \
|
|
texpr = expr; \
|
|
__asm __volatile ("" : : "r" (&texpr)); \
|
|
if (count != 1 || counts[T##type][C_##fn] != 1) \
|
|
{ \
|
|
FAIL ("wrong function called, "#fn" ("#type")"); \
|
|
memset (counts, 0, sizeof (counts)); \
|
|
} \
|
|
count = 0; \
|
|
counts[T##type][C_##fn] = 0; \
|
|
} \
|
|
while (0)
|
|
#define TEST(expr, type, fn) TEST2(expr, type, type, fn)
|
|
|
|
int
|
|
test_cos (const int Vint4, const long long int Vllong4)
|
|
{
|
|
TEST (cos (vfloat1), float, cos);
|
|
TEST (cos (vdouble1), double, cos);
|
|
TEST (cos (vldouble1), ldouble, cos);
|
|
TEST (cos (vint1), double, cos);
|
|
TEST (cos (vllong1), double, cos);
|
|
TEST (cos (vcfloat1), cfloat, cos);
|
|
TEST (cos (vcdouble1), cdouble, cos);
|
|
TEST (cos (vcldouble1), cldouble, cos);
|
|
TEST (cos (Vfloat1), float, cos);
|
|
TEST (cos (Vdouble1), double, cos);
|
|
TEST (cos (Vldouble1), ldouble, cos);
|
|
TEST (cos (Vint1), double, cos);
|
|
TEST (cos (Vllong1), double, cos);
|
|
TEST (cos (Vcfloat1), cfloat, cos);
|
|
TEST (cos (Vcdouble1), cdouble, cos);
|
|
TEST (cos (Vcldouble1), cldouble, cos);
|
|
|
|
return 0;
|
|
}
|
|
|
|
int
|
|
test_fabs (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
|
|
TEST (fabs (vfloat1), float, fabs);
|
|
TEST (fabs (vdouble1), double, fabs);
|
|
TEST (fabs (vldouble1), ldouble, fabs);
|
|
TEST (fabs (vint1), double, fabs);
|
|
TEST (fabs (vllong1), double, fabs);
|
|
TEST (fabs (vcfloat1), float, cabs);
|
|
TEST (fabs (vcdouble1), double, cabs);
|
|
TEST (fabs (vcldouble1), ldouble, cabs);
|
|
TEST (fabs (Vfloat1), float, fabs);
|
|
TEST (fabs (Vdouble1), double, fabs);
|
|
TEST (fabs (Vldouble2), ldouble, fabs);
|
|
#ifndef __OPTIMIZE__
|
|
/* GCC is too smart to optimize these out. */
|
|
TEST (fabs (Vint1), double, fabs);
|
|
TEST (fabs (Vllong1), double, fabs);
|
|
#else
|
|
TEST_TYPE_ONLY (fabs (vllong1), double);
|
|
TEST_TYPE_ONLY (fabs (vllong1), double);
|
|
#endif
|
|
TEST (fabs (Vint4), double, fabs);
|
|
TEST (fabs (Vllong4), double, fabs);
|
|
TEST (fabs (Vcfloat1), float, cabs);
|
|
TEST (fabs (Vcdouble1), double, cabs);
|
|
TEST (fabs (Vcldouble1), ldouble, cabs);
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_conj (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
TEST (conj (vfloat1), cfloat, conj);
|
|
TEST (conj (vdouble1), cdouble, conj);
|
|
TEST (conj (vldouble1), cldouble, conj);
|
|
TEST (conj (vint1), cdouble, conj);
|
|
TEST (conj (vllong1), cdouble, conj);
|
|
TEST (conj (vcfloat1), cfloat, conj);
|
|
TEST (conj (vcdouble1), cdouble, conj);
|
|
TEST (conj (vcldouble1), cldouble, conj);
|
|
TEST (conj (Vfloat1), cfloat, conj);
|
|
TEST (conj (Vdouble1), cdouble, conj);
|
|
TEST (conj (Vldouble1), cldouble, conj);
|
|
TEST (conj (Vint1), cdouble, conj);
|
|
TEST (conj (Vllong1), cdouble, conj);
|
|
TEST (conj (Vcfloat1), cfloat, conj);
|
|
TEST (conj (Vcdouble1), cdouble, conj);
|
|
TEST (conj (Vcldouble1), cldouble, conj);
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_expm1 (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
|
|
TEST (expm1 (vfloat1), float, expm1);
|
|
TEST (expm1 (vdouble1), double, expm1);
|
|
TEST (expm1 (vldouble1), ldouble, expm1);
|
|
TEST (expm1 (vint1), double, expm1);
|
|
TEST (expm1 (vllong1), double, expm1);
|
|
TEST (expm1 (Vfloat1), float, expm1);
|
|
TEST (expm1 (Vdouble1), double, expm1);
|
|
TEST (expm1 (Vldouble1), ldouble, expm1);
|
|
TEST (expm1 (Vint1), double, expm1);
|
|
TEST (expm1 (Vllong1), double, expm1);
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_lrint (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
TEST2 (lrint (vfloat1), float, long int, lrint);
|
|
TEST2 (lrint (vdouble1), double, long int, lrint);
|
|
TEST2 (lrint (vldouble1), ldouble, long int, lrint);
|
|
TEST2 (lrint (vint1), double, long int, lrint);
|
|
TEST2 (lrint (vllong1), double, long int, lrint);
|
|
TEST2 (lrint (Vfloat1), float, long int, lrint);
|
|
TEST2 (lrint (Vdouble1), double, long int, lrint);
|
|
TEST2 (lrint (Vldouble1), ldouble, long int, lrint);
|
|
TEST2 (lrint (Vint1), double, long int, lrint);
|
|
TEST2 (lrint (Vllong1), double, long int, lrint);
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_ldexp (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
|
|
TEST (ldexp (vfloat1, 6), float, ldexp);
|
|
TEST (ldexp (vdouble1, 6), double, ldexp);
|
|
TEST (ldexp (vldouble1, 6), ldouble, ldexp);
|
|
TEST (ldexp (vint1, 6), double, ldexp);
|
|
TEST (ldexp (vllong1, 6), double, ldexp);
|
|
TEST (ldexp (Vfloat1, 6), float, ldexp);
|
|
TEST (ldexp (Vdouble1, 6), double, ldexp);
|
|
TEST (ldexp (Vldouble1, 6), ldouble, ldexp);
|
|
TEST (ldexp (Vint1, 6), double, ldexp);
|
|
TEST (ldexp (Vllong1, 6), double, ldexp);
|
|
|
|
return result;
|
|
}
|
|
|
|
#define FIRST(x, y) (y, x)
|
|
#define SECOND(x, y) (x, y)
|
|
#define NON_LDBL_TEST(fn, argm, arg, type, fnt) \
|
|
TEST (fn argm (arg, vfloat1), type, fnt); \
|
|
TEST (fn argm (arg, vdouble1), type, fnt); \
|
|
TEST (fn argm (arg, vint1), type, fnt); \
|
|
TEST (fn argm (arg, vllong1), type, fnt); \
|
|
TEST (fn argm (arg, Vfloat1), type, fnt); \
|
|
TEST (fn argm (arg, Vdouble1), type, fnt); \
|
|
TEST (fn argm (arg, Vint1), type, fnt); \
|
|
TEST (fn argm (arg, Vllong1), type, fnt);
|
|
#define NON_LDBL_CTEST(fn, argm, arg, type, fnt) \
|
|
NON_LDBL_TEST(fn, argm, arg, type, fnt); \
|
|
TEST (fn argm (arg, vcfloat1), type, fnt); \
|
|
TEST (fn argm (arg, vcdouble1), type, fnt); \
|
|
TEST (fn argm (arg, Vcfloat1), type, fnt); \
|
|
TEST (fn argm (arg, Vcdouble1), type, fnt);
|
|
#define BINARY_TEST(fn, fnt) \
|
|
TEST (fn (vfloat1, vfloat2), float, fnt); \
|
|
TEST (fn (Vfloat1, vfloat2), float, fnt); \
|
|
TEST (fn (vfloat1, Vfloat2), float, fnt); \
|
|
TEST (fn (Vfloat1, Vfloat2), float, fnt); \
|
|
TEST (fn (vldouble1, vldouble2), ldouble, fnt); \
|
|
TEST (fn (Vldouble1, vldouble2), ldouble, fnt); \
|
|
TEST (fn (vldouble1, Vldouble2), ldouble, fnt); \
|
|
TEST (fn (Vldouble1, Vldouble2), ldouble, fnt); \
|
|
NON_LDBL_TEST (fn, FIRST, vldouble2, ldouble, fnt); \
|
|
NON_LDBL_TEST (fn, SECOND, vldouble2, ldouble, fnt); \
|
|
NON_LDBL_TEST (fn, FIRST, Vldouble2, ldouble, fnt); \
|
|
NON_LDBL_TEST (fn, SECOND, Vldouble2, ldouble, fnt); \
|
|
NON_LDBL_TEST (fn, FIRST, vdouble2, double, fnt); \
|
|
NON_LDBL_TEST (fn, SECOND, vdouble2, double, fnt); \
|
|
NON_LDBL_TEST (fn, FIRST, Vdouble2, double, fnt); \
|
|
NON_LDBL_TEST (fn, SECOND, Vdouble2, double, fnt); \
|
|
NON_LDBL_TEST (fn, FIRST, vint2, double, fnt); \
|
|
NON_LDBL_TEST (fn, SECOND, vint2, double, fnt); \
|
|
NON_LDBL_TEST (fn, FIRST, Vint2, double, fnt); \
|
|
NON_LDBL_TEST (fn, SECOND, Vint2, double, fnt); \
|
|
NON_LDBL_TEST (fn, FIRST, vllong2, double, fnt); \
|
|
NON_LDBL_TEST (fn, SECOND, vllong2, double, fnt); \
|
|
NON_LDBL_TEST (fn, FIRST, Vllong2, double, fnt); \
|
|
NON_LDBL_TEST (fn, SECOND, Vllong2, double, fnt);
|
|
#define BINARY_CTEST(fn, fnt) \
|
|
BINARY_TEST (fn, fnt); \
|
|
TEST (fn (vcfloat1, vfloat2), cfloat, fnt); \
|
|
TEST (fn (Vcfloat1, vfloat2), cfloat, fnt); \
|
|
TEST (fn (vcfloat1, Vfloat2), cfloat, fnt); \
|
|
TEST (fn (Vcfloat1, Vfloat2), cfloat, fnt); \
|
|
TEST (fn (vcldouble1, vldouble2), cldouble, fnt); \
|
|
TEST (fn (Vcldouble1, vldouble2), cldouble, fnt); \
|
|
TEST (fn (vcldouble1, Vldouble2), cldouble, fnt); \
|
|
TEST (fn (Vcldouble1, Vldouble2), cldouble, fnt); \
|
|
TEST (fn (vcfloat1, vfloat2), cfloat, fnt); \
|
|
TEST (fn (Vcfloat1, vfloat2), cfloat, fnt); \
|
|
TEST (fn (vcfloat1, Vfloat2), cfloat, fnt); \
|
|
TEST (fn (Vcfloat1, Vfloat2), cfloat, fnt); \
|
|
TEST (fn (vcldouble1, vldouble2), cldouble, fnt); \
|
|
TEST (fn (Vcldouble1, vldouble2), cldouble, fnt); \
|
|
TEST (fn (vcldouble1, Vldouble2), cldouble, fnt); \
|
|
TEST (fn (Vcldouble1, Vldouble2), cldouble, fnt); \
|
|
TEST (fn (vcfloat1, vcfloat2), cfloat, fnt); \
|
|
TEST (fn (Vcfloat1, vcfloat2), cfloat, fnt); \
|
|
TEST (fn (vcfloat1, Vcfloat2), cfloat, fnt); \
|
|
TEST (fn (Vcfloat1, Vcfloat2), cfloat, fnt); \
|
|
TEST (fn (vcldouble1, vcldouble2), cldouble, fnt); \
|
|
TEST (fn (Vcldouble1, vcldouble2), cldouble, fnt); \
|
|
TEST (fn (vcldouble1, Vcldouble2), cldouble, fnt); \
|
|
TEST (fn (Vcldouble1, Vcldouble2), cldouble, fnt); \
|
|
NON_LDBL_CTEST (fn, FIRST, vcldouble2, cldouble, fnt); \
|
|
NON_LDBL_CTEST (fn, SECOND, vcldouble2, cldouble, fnt); \
|
|
NON_LDBL_CTEST (fn, FIRST, Vcldouble2, cldouble, fnt); \
|
|
NON_LDBL_CTEST (fn, SECOND, Vcldouble2, cldouble, fnt); \
|
|
NON_LDBL_CTEST (fn, FIRST, vcdouble2, cdouble, fnt); \
|
|
NON_LDBL_CTEST (fn, SECOND, vcdouble2, cdouble, fnt); \
|
|
NON_LDBL_CTEST (fn, FIRST, Vcdouble2, cdouble, fnt); \
|
|
NON_LDBL_CTEST (fn, SECOND, Vcdouble2, cdouble, fnt);
|
|
|
|
int
|
|
test_atan2 (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
|
|
BINARY_TEST (atan2, atan2);
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_remquo (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
int quo = 0;
|
|
|
|
#define my_remquo(x, y) remquo (x, y, &quo)
|
|
BINARY_TEST (my_remquo, remquo);
|
|
#undef my_remquo
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_pow (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
|
|
BINARY_CTEST (pow, pow);
|
|
|
|
return result;
|
|
}
|
|
|
|
/* Testing all arguments of fma would be just too expensive,
|
|
so test just some. */
|
|
|
|
int
|
|
test_fma_1 (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
|
|
#define my_fma(x, y) fma (x, y, vfloat3)
|
|
BINARY_TEST (my_fma, fma);
|
|
#undef my_fma
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_fma_2 (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
|
|
#define my_fma(x, y) fma (x, vfloat3, y)
|
|
BINARY_TEST (my_fma, fma);
|
|
#undef my_fma
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_fma_3 (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
|
|
#define my_fma(x, y) fma (Vfloat3, x, y)
|
|
BINARY_TEST (my_fma, fma);
|
|
#undef my_fma
|
|
|
|
return result;
|
|
}
|
|
|
|
int
|
|
test_fma_4 (const int Vint4, const long long int Vllong4)
|
|
{
|
|
int result = 0;
|
|
TEST (fma (vdouble1, Vdouble2, vllong3), double, fma);
|
|
TEST (fma (vint1, Vint2, vint3), double, fma);
|
|
TEST (fma (Vldouble1, vldouble2, Vldouble3), ldouble, fma);
|
|
TEST (fma (vldouble1, vint2, Vdouble3), ldouble, fma);
|
|
|
|
return result;
|
|
}
|
|
|
|
static int
|
|
do_test (void)
|
|
{
|
|
int result;
|
|
|
|
result = test_cos (vint1, vllong1);
|
|
result |= test_fabs (vint1, vllong1);
|
|
result |= test_conj (vint1, vllong1);
|
|
result |= test_expm1 (vint1, vllong1);
|
|
result |= test_lrint (vint1, vllong1);
|
|
result |= test_ldexp (vint1, vllong1);
|
|
result |= test_atan2 (vint1, vllong1);
|
|
result |= test_remquo (vint1, vllong1);
|
|
result |= test_pow (vint1, vllong1);
|
|
result |= test_fma_1 (vint1, vllong1);
|
|
result |= test_fma_2 (vint1, vllong1);
|
|
result |= test_fma_3 (vint1, vllong1);
|
|
result |= test_fma_4 (vint1, vllong1);
|
|
|
|
return result;
|
|
}
|
|
|
|
/* Now generate the three functions. */
|
|
#define HAVE_MAIN
|
|
|
|
#define F(name) name
|
|
#define TYPE double
|
|
#define CTYPE cdouble
|
|
#define T Tdouble
|
|
#define C Tcdouble
|
|
#include "test-tgmath2.c"
|
|
|
|
#define F(name) name##f
|
|
#define TYPE float
|
|
#define CTYPE cfloat
|
|
#define T Tfloat
|
|
#define C Tcfloat
|
|
#include "test-tgmath2.c"
|
|
|
|
#if LDBL_MANT_DIG > DBL_MANT_DIG
|
|
#define F(name) name##l
|
|
#define TYPE ldouble
|
|
#define CTYPE cldouble
|
|
#define T Tldouble
|
|
#define C Tcldouble
|
|
#include "test-tgmath2.c"
|
|
#endif
|
|
|
|
#define TEST_FUNCTION do_test ()
|
|
#include "../test-skeleton.c"
|
|
|
|
#else
|
|
|
|
#ifdef DEBUG
|
|
#define P() puts (__FUNCTION__); count++
|
|
#else
|
|
#define P() count++;
|
|
#endif
|
|
|
|
TYPE
|
|
(F(cos)) (TYPE x)
|
|
{
|
|
counts[T][C_cos]++;
|
|
P ();
|
|
return x;
|
|
}
|
|
|
|
CTYPE
|
|
(F(ccos)) (CTYPE x)
|
|
{
|
|
counts[C][C_cos]++;
|
|
P ();
|
|
return x;
|
|
}
|
|
|
|
TYPE
|
|
(F(fabs)) (TYPE x)
|
|
{
|
|
counts[T][C_fabs]++;
|
|
P ();
|
|
return x;
|
|
}
|
|
|
|
TYPE
|
|
(F(cabs)) (CTYPE x)
|
|
{
|
|
counts[T][C_cabs]++;
|
|
P ();
|
|
return x;
|
|
}
|
|
|
|
CTYPE
|
|
(F(conj)) (CTYPE x)
|
|
{
|
|
counts[C][C_conj]++;
|
|
P ();
|
|
return x;
|
|
}
|
|
|
|
TYPE
|
|
(F(expm1)) (TYPE x)
|
|
{
|
|
counts[T][C_expm1]++;
|
|
P ();
|
|
return x;
|
|
}
|
|
|
|
long int
|
|
(F(lrint)) (TYPE x)
|
|
{
|
|
counts[T][C_lrint]++;
|
|
P ();
|
|
return x;
|
|
}
|
|
|
|
TYPE
|
|
(F(ldexp)) (TYPE x, int y)
|
|
{
|
|
counts[T][C_ldexp]++;
|
|
P ();
|
|
return x + y;
|
|
}
|
|
|
|
TYPE
|
|
(F(atan2)) (TYPE x, TYPE y)
|
|
{
|
|
counts[T][C_atan2]++;
|
|
P ();
|
|
return x + y;
|
|
}
|
|
|
|
TYPE
|
|
(F(remquo)) (TYPE x, TYPE y, int *z)
|
|
{
|
|
counts[T][C_remquo]++;
|
|
P ();
|
|
return x + y + *z;
|
|
}
|
|
|
|
TYPE
|
|
(F(pow)) (TYPE x, TYPE y)
|
|
{
|
|
counts[T][C_pow]++;
|
|
P ();
|
|
return x + y;
|
|
}
|
|
|
|
CTYPE
|
|
(F(cpow)) (CTYPE x, CTYPE y)
|
|
{
|
|
counts[C][C_pow]++;
|
|
P ();
|
|
return x + y;
|
|
}
|
|
|
|
TYPE
|
|
(F(fma)) (TYPE x, TYPE y, TYPE z)
|
|
{
|
|
counts[T][C_fma]++;
|
|
P ();
|
|
return x + y + z;
|
|
}
|
|
|
|
#undef F
|
|
#undef TYPE
|
|
#undef CTYPE
|
|
#undef T
|
|
#undef C
|
|
#undef P
|
|
#endif
|