glibc/stdlib/tst-strtod-round-skeleton.c

317 lines
9.9 KiB
C
Raw Normal View History

/* Test for correct rounding of results of strtod and related
functions.
Copyright (C) 2012-2018 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/>. */
/* Defining _LIBC_TEST ensures long double math functions are
declared in the headers. */
#define _LIBC_TEST 1
#define __STDC_WANT_IEC_60559_TYPES_EXT__
#include <fenv.h>
#include <float.h>
#include <math.h>
#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <math-tests.h>
#include "tst-strtod.h"
/* Non-standard macros expected to be externally defined:
L_(str): Pastes the appropriate modifier to a string literal str.
FNPFX: Expands to the correct prefix for the strtod equivalent
of type CHAR. (e.g str or wcs).
CHAR: Expands to the string type being tested (e.g wchar_t or char).
STRM: Expands to a string literal suitable for printing CHAR* via
printf (e.g "%s" or "%ls"). */
#define _CONCAT(a, b) a ## b
#define CONCAT(a, b) _CONCAT (a, b)
#define STRTO(x) CONCAT (CONCAT (FNPFX, to), x)
#if LDBL_MANT_DIG == 106 && LDBL_MAX_EXP == 1024
/* This is a stupid hack for IBM long double. This test ignores
inexact values for long double due to the limitations of the
format. This ensures rounding tests are ignored. */
# undef ROUNDING_TESTS_long_double
# define ROUNDING_TESTS_long_double(x) 0
#endif
/* Generator to create an FTYPE member variabled named FSUF
used to populate struct member variables. */
#define FTYPE_MEMBER(FSUF, FTYPE, FTOSTR, LSUF, CSUF) \
FTYPE FSUF;
/* Likewise, but each member is of type bool. */
#define BOOL_MEMBER(FSUF, FTYPE, FTOSTR, LSUF, CSUF) \
bool FSUF;
#define STRUCT_FOREACH_FLOAT_FTYPE GEN_TEST_STRTOD_FOREACH (FTYPE_MEMBER)
#define STRUCT_FOREACH_FLOAT_BOOL GEN_TEST_STRTOD_FOREACH (BOOL_MEMBER)
/* Define the long double choose (CHOOSE_ld) macro
to select the appropriate generated long double
value from the generated test data. */
#if LDBL_MANT_DIG == 53 && LDBL_MAX_EXP == 1024
/* This is for the long double == double format. */
# define CHOOSE_ld(f,d,...) d
#elif LDBL_MANT_DIG == 64 && LDBL_MAX_EXP == 16384 && LDBL_MIN_EXP == -16381
/* This is for the Intel extended float format. */
# define CHOOSE_ld(f,d,ld64i,...) ld64i
#elif LDBL_MANT_DIG == 64 && LDBL_MAX_EXP == 16384 && LDBL_MIN_EXP == -16382
/* This is for the Motorola extended float format. */
# define CHOOSE_ld(f,d,ld64i,ld64m,...) ld64m
#elif LDBL_MANT_DIG == 106 && LDBL_MAX_EXP == 1024
/* This is for the IBM extended double format. */
# define CHOOSE_ld(f,d,ld64i,ld64m,ld106,...) ld106
#elif LDBL_MANT_DIG == 113 && LDBL_MAX_EXP == 16384
/* This is for the IEEE binary128 format. */
# define CHOOSE_ld(f,d,ld64i,ld64m,ld106,ld113,...) ld113
#else
# error "unknown long double format"
#endif
/* Add type specific choosing macros below. */
#define CHOOSE_f(f,...) f
#define CHOOSE_f32(f,...) f
#define CHOOSE_d(f,d,...) d
#define CHOOSE_f64(f,d,...) d
#define CHOOSE_f32x(f,d,...) d
float128: Add strtof128, wcstof128, and related functions. The implementations are contained with sysdeps/ieee754/float128 as they are only built when _Float128 is enabled within libc/m. * include/gmp.h (__mpn_construct_float128): New declaration. * include/stdlib.h: Include bits/floatn.h for _Float128 tests. (__strtof128_l): New declaration. (__strtof128_nan): Likewise. (__wcstof128_nan): Likewise. (__strtof128_internal): Likewise. (____strtof128_l_internal): Likewise. * include/wchar.h: Include bits/floatn.h for _Float128 tests. (__wcstof128_l): New declaration. (__wcstof128_internal): Likewise. * stdlib/Makefile (bug-strtod2): Link libm too. * stdlib/stdlib.h (strtof128): New declaration. (strtof128_l): Likewise. * stdlib/tst-strtod-nan-locale-main.c: Updated to use tst-strtod.h macros to ensure float128 gets tested too. * stdlib/tst-strtod-round-skeleton.c (CHOOSE_f128): New macro. * stdlib/tst-strtod.h: Include bits/floatn.h for _Float128 tests. (IF_FLOAT128): New macro. (GEN_TEST_STRTOD): Update to optionally include _Float128 in the tests. (STRTOD_TEST_FOREACH): Likewise. * sysdeps/ieee754/float128/Makefile: Insert new strtof128 and wcstof128 functions into libc. * sysdeps/ieee754/float128/Versions: Add exports for the above new functions. * sysdeps/ieee754/float128/mpn2float128.c: New file. * sysdeps/ieee754/float128/strtod_nan_float128.h: New file. * sysdeps/ieee754/float128/strtof128.c: New file. * sysdeps/ieee754/float128/strtof128_l.c: New file. * sysdeps/ieee754/float128/strtof128_nan.c: New file. * sysdeps/ieee754/float128/wcstof128.c: New file. * sysdeps/ieee754/float128/wcstof128_l.c: New file. * sysdeps/ieee754/float128/wcstof128_nan.c: New fike. * wcsmbs/Makefile: (CFLAGS-wcstof128.c): Append strtox-CFLAGS. (CFLAGS-wcstof128_l): Likewise. * wcsmbs/wchar.h: Include bits/floatn.h for _Float128 tests. (wcstof128): New declaration. (wcstof128_l): Likewise.
2016-07-25 20:10:00 +02:00
#define CHOOSE_f128(f,d,ld64i,ld64m,ld106,ld113,...) ld113
/* long double is special, and handled above. _Float16 would require
updates to the generator to generate appropriate expectations, and
updates to the test inputs to cover difficult rounding cases for
_Float16. */
#if __HAVE_FLOAT64X
# if FLT64X_MANT_DIG == 113 && FLT64X_MAX_EXP == 16384
# define CHOOSE_f64x(f,d,ld64i,ld64m,ld106,ld113,...) ld113
# elif (FLT64X_MANT_DIG == 64 \
&& FLT64X_MAX_EXP == 16384 \
&& FLT64X_MIN_EXP == -16381)
# define CHOOSE_f64x(f,d,ld64i,...) ld64i
# else
# error "unknown _Float64x format"
# endif
#endif
/* Selector for expected result field of a given type. */
#define _ENTRY(FSUF, FTYPE, FTOSTR, LSUF, CSUF, ...) \
CONCAT (CHOOSE_ ## FSUF (__VA_ARGS__), LSUF),
#define ENTRY(...) \
GEN_TEST_STRTOD_FOREACH (_ENTRY, __VA_ARGS__)
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
/* Selector for boolean exact tag of expected results and that for
overflow. */
#define _XNTRY(FSUF, FTYPE, FTOSTR, LSUF, CSUF, ...) \
CHOOSE_ ## FSUF (__VA_ARGS__),
#define XNTRY(...) \
GEN_TEST_STRTOD_FOREACH (_XNTRY, __VA_ARGS__)
/* This is hacky way around the seemingly unavoidable macro
expansion of the INFINITY or HUGE_VAL like macros in the
above. It is assumed the compiler will implicitly convert
the infinity correctly. */
#define INF INFINITY + 0.0
/* This macro is used in conjunction with the output from the
gen-tst-strtod-round utility to select the appropriately
rounded long double value for a given format. */
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
#define TEST(s, \
fx, fd, fdo, fn, fno, fz, fzo, fu, fuo, \
dx, dd, ddo, dn, dno, dz, dzo, du, duo, \
ld64ix, ld64id, ld64ido, ld64in, ld64ino, \
ld64iz, ld64izo, ld64iu, ld64iuo, \
ld64mx, ld64md, ld64mdo, ld64mn, ld64mno, \
ld64mz, ld64mzo, ld64mu, ld64muo, \
ld106x, ld106d, ld106do, ld106n, ld106no, \
ld106z, ld106zo, ld106u, ld106uo, \
ld113x, ld113d, ld113do, ld113n, ld113no, \
ld113z, ld113zo, ld113u, ld113uo) \
{ \
L_ (s), \
{ XNTRY (fx, dx, ld64ix, ld64mx, ld106x, ld113x) }, \
{ \
{ ENTRY (fn, dn, ld64in, ld64mn, ld106n, ld113n) }, \
{ ENTRY (fd, dd, ld64id, ld64md, ld106d, ld113d) }, \
{ ENTRY (fz, dz, ld64iz, ld64mz, ld106z, ld113z) }, \
{ ENTRY (fu, du, ld64iu, ld64mu, ld106u, ld113u) } \
}, \
{ \
{ XNTRY (fno, dno, ld64ino, ld64mno, ld106no, ld113no) }, \
{ XNTRY (fdo, ddo, ld64ido, ld64mdo, ld106do, ld113do) }, \
{ XNTRY (fzo, dzo, ld64izo, ld64mzo, ld106zo, ld113zo) }, \
{ XNTRY (fuo, duo, ld64iuo, ld64muo, ld106uo, ld113uo) } \
} \
}
struct test_exactness
{
STRUCT_FOREACH_FLOAT_BOOL
};
struct test_results
{
STRUCT_FOREACH_FLOAT_FTYPE
};
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
struct test_overflow
{
STRUCT_FOREACH_FLOAT_BOOL
};
struct test {
const CHAR *s;
struct test_exactness exact;
struct test_results r[4];
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
struct test_overflow o[4];
};
/* Include the generated test data. */
#include "tst-strtod-round-data.h"
#define STRX(x) #x
#define STR(x) STRX (x)
#define FNPFXS STR (FNPFX)
#ifndef FE_INEXACT
# define FE_INEXACT 0
#endif
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
#ifndef FE_OVERFLOW
# define FE_OVERFLOW 0
#endif
#define GEN_ONE_TEST(FSUF, FTYPE, FTOSTR, LSUF, CSUF) \
{ \
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
feclearexcept (FE_ALL_EXCEPT); \
FTYPE f = STRTO (FSUF) (s, NULL); \
if (f != expected->FSUF \
|| (copysign ## CSUF) (1.0 ## LSUF, f) \
!= (copysign ## CSUF) (1.0 ## LSUF, expected->FSUF)) \
{ \
char efstr[FSTRLENMAX]; \
char fstr[FSTRLENMAX]; \
FTOSTR (efstr, FSTRLENMAX, "%a", expected->FSUF); \
FTOSTR (fstr, FSTRLENMAX, "%a", f); \
printf (FNPFXS "to" #FSUF " (" STRM ") returned %s not " \
"%s (%s)\n", s, fstr, efstr, mode_name); \
if (ROUNDING_TESTS (FTYPE, rnd_mode) || exact->FSUF) \
result = 1; \
else \
printf ("ignoring this inexact result\n"); \
} \
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
else \
{ \
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
if (FE_INEXACT != 0) \
{ \
bool inexact_raised = fetestexcept (FE_INEXACT) != 0; \
if (inexact_raised != !exact->FSUF) \
{ \
printf (FNPFXS "to" #FSUF \
" (" STRM ") inexact %d " \
"not %d\n", s, inexact_raised, \
!exact->FSUF); \
if (EXCEPTION_TESTS (FTYPE)) \
result = 1; \
else \
printf ("ignoring this exception error\n"); \
} \
} \
if (FE_OVERFLOW != 0) \
{ \
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
bool overflow_raised \
= fetestexcept (FE_OVERFLOW) != 0; \
if (overflow_raised != overflow->FSUF) \
{ \
printf (FNPFXS "to" #FSUF \
" (" STRM ") overflow %d " \
"not %d\n", s, overflow_raised, \
overflow->FSUF); \
if (EXCEPTION_TESTS (FTYPE)) \
result = 1; \
else \
printf ("ignoring this exception error\n"); \
} \
} \
} \
}
static int
test_in_one_mode (const CHAR *s, const struct test_results *expected,
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
const struct test_exactness *exact,
const struct test_overflow *overflow,
const char *mode_name, int rnd_mode)
{
int result = 0;
GEN_TEST_STRTOD_FOREACH (GEN_ONE_TEST)
return result;
}
static const struct fetestmodes
{
const char *mode_name;
int rnd_mode;
int rnd_i; /* Corresponding index into r array of struct test. */
} modes[] = {
{ "default rounding mode", FE_TONEAREST, 0 },
#ifdef FE_DOWNWARD
{ "FE_DOWNWARD", FE_DOWNWARD, 1 },
#endif
#ifdef FE_TOWARDZERO
{ "FE_TOWARDZERO", FE_TOWARDZERO, 2 },
#endif
#ifdef FE_UPWARD
{ "FE_UPWARD", FE_UPWARD, 3 },
#endif
{}
};
static int
do_test (void)
{
int save_round_mode __attribute__ ((unused)) = fegetround ();
int result = 0;
for (size_t i = 0; i < sizeof (tests) / sizeof (tests[0]); i++)
{
result |= test_in_one_mode (tests[i].s, &tests[i].r[modes[0].rnd_i],
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
&tests[i].exact, &tests[i].o[modes[0].rnd_i],
modes[0].mode_name, modes[0].rnd_mode);
for (const struct fetestmodes *m = &modes[1]; m->mode_name != NULL; m++)
{
if (!fesetround (m->rnd_mode))
{
result |= test_in_one_mode (tests[i].s, &tests[i].r[m->rnd_i],
Fix strtod overflow detection (bug 23279). As shown by bug 23279, strtod's round_and_return has an off-by-one error in its overflow detection, only counting an exponent greater than MAX_EXP as overflowing when an exponent of MAX_EXP also means overflow (recall the ISO C definition of DBL_MAX_EXP etc. is based on a floating-point model where 2^exp is multiplied by a value in the interval [0.5, 1), so 2^MAX_EXP is not representable). For decimal arguments to strtod, a separate overflow check in the main implementation covers the case where the integer part of the argument (truncated to the nearest integer towards zero) has more than MAX_EXP bits, meaning that this issue in round_and_return only affects cases (arguments with absolute value strictly between the maximum representable value and 2^MAX_EXP) where overflow depends on the rounding mode; in such cases, the returned value would still have been correct on overflow but without the overflow exception being raised or errno being set to ERANGE. For hex float arguments, however, other cases can arise, as shown in bug 23279, where a value with exponent already set to MAX_EXP is passed into round_and_return and a result can wrongly end up being NaN, or infinity instead of the largest finite value. This patch fixes the off-by-one error, adds testing of overflow exceptions to the tst-strtod-round framework, and adds tests of these issues. Tested for x86_64. Also ran the tst-strtod-round tests for powerpc to make sure the new tests didn't introduce any new failures for IBM long double. [BZ #23279] * stdlib/strtod_l.c (round_and_return): Handle an exponent of MAX_EXP as overflowing. * stdlib/gen-tst-strtod-round.c (string_to_fp): Clear MPFR overflow flag. (round_str): Output also whether result overflows in each rounding mode. * stdlib/tst-strtod-round-data: Add more tests. * stdlib/tst-strtod-round-data.h: Regenerated. * stdlib/tst-strtod-round-skeleton.c (_XNTRY): Update comment. (TEST): Handle extra arguments for overflow flags. (struct test_overflow): New type. [!FE_OVERFLOW] (FE_OVERFLOW): Define to 0. (GEN_ONE_TEST): Clear all exceptions. Test overflow flag. (test_in_one_mode): Take argument with overflow information. (do_test): Update calls to test_in_one_mode.
2018-06-13 18:06:14 +02:00
&tests[i].exact,
&tests[i].o[m->rnd_i], m->mode_name,
m->rnd_mode);
fesetround (save_round_mode);
}
}
}
return result;
}
#define TEST_FUNCTION do_test ()
#include "../test-skeleton.c"