Joseph Myers 59a63cca11 Fix nexttoward overflow in non-default rounding modes (bug 19059). ... ISO C requires overflowing results from nexttoward to be the appropriate infinity independent of the rounding mode, but some implementations use a rounding-mode-dependent result (this is the same issue as was fixed for nextafter in bug 16677). This patch fixes the problem by making the nexttoward implementations discard the result from the floating-point computation that forced an overflow exception and then return the infinity previously computed with integer arithmetic. Tested for x86_64, x86, mips64 and powerpc. [BZ #19059] * math/s_nexttowardf.c (__nexttowardf): Do not return value from overflowing computation. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise. * math/libm-test.inc (nexttoward_test_data): Add more tests.		2015-10-02 17:11:13 +00:00
..
cabs.c
cabsl.c
carg.c
cargl.c
cimag.c
cimagl.c
configure	rename configure.in to configure.ac	2013-10-30 17:32:08 +10:00
configure.ac	rename configure.in to configure.ac	2013-10-30 17:32:08 +10:00
conj.c
conjl.c
creal.c
creall.c
Makefile
math_ldbl_opt.c
math_ldbl_opt.h	Remove --disable-versioning.	2013-09-04 15:25:42 +00:00
nldbl-acos.c
nldbl-acosh.c
nldbl-asin.c
nldbl-asinh.c
nldbl-asprintf_chk.c
nldbl-asprintf.c
nldbl-atan2.c
nldbl-atan.c
nldbl-atanh.c
nldbl-cabs.c
nldbl-cacos.c
nldbl-cacosh.c
nldbl-carg.c
nldbl-casin.c
nldbl-casinh.c
nldbl-catan.c
nldbl-catanh.c
nldbl-cbrt.c
nldbl-ccos.c
nldbl-ccosh.c
nldbl-ceil.c
nldbl-cexp.c
nldbl-cimag.c
nldbl-clog10.c
nldbl-clog.c
nldbl-compat.c	Move bits/libc-lock.h and bits/libc-lockP.h out of bits/ (bug 14912).	2015-09-08 21:11:03 +00:00
nldbl-compat.h	Update copyright dates with scripts/update-copyrights.	2015-01-02 16:29:47 +00:00
nldbl-conj.c
nldbl-copysign.c	Use __copysign rather than copysign.	2015-04-22 12:07:56 +00:00
nldbl-cos.c
nldbl-cosh.c
nldbl-cpow.c
nldbl-cproj.c
nldbl-creal.c
nldbl-csin.c
nldbl-csinh.c
nldbl-csqrt.c
nldbl-ctan.c
nldbl-ctanh.c
nldbl-dprintf_chk.c
nldbl-dprintf.c
nldbl-drem.c
nldbl-erf.c
nldbl-erfc.c
nldbl-exp2.c
nldbl-exp10.c
nldbl-exp.c
nldbl-expm1.c
nldbl-fabs.c
nldbl-fdim.c
nldbl-finite.c	Replace finite with isfinite.	2015-06-03 16:35:44 +01:00
nldbl-floor.c
nldbl-fma.c
nldbl-fmax.c
nldbl-fmin.c
nldbl-fmod.c
nldbl-fprintf_chk.c
nldbl-fprintf.c
nldbl-frexp.c
nldbl-fscanf.c
nldbl-fwprintf_chk.c
nldbl-fwprintf.c
nldbl-fwscanf.c
nldbl-gamma.c
nldbl-hypot.c
nldbl-ilogb.c
nldbl-iovfscanf.c
nldbl-isinf.c	This patch renames all uses of __isinf*, __isnan*, __finite* and __signbit* to use standard C99 macros. This has no effect on generated code.	2015-06-03 15:41:36 +01:00
nldbl-isnan.c	This patch renames all uses of __isinf*, __isnan*, __finite* and __signbit* to use standard C99 macros. This has no effect on generated code.	2015-06-03 15:41:36 +01:00
nldbl-isoc99_fscanf.c
nldbl-isoc99_fwscanf.c
nldbl-isoc99_scanf.c
nldbl-isoc99_sscanf.c
nldbl-isoc99_swscanf.c
nldbl-isoc99_vfscanf.c
nldbl-isoc99_vfwscanf.c
nldbl-isoc99_vscanf.c
nldbl-isoc99_vsscanf.c
nldbl-isoc99_vswscanf.c
nldbl-isoc99_vwscanf.c
nldbl-isoc99_wscanf.c
nldbl-j0.c
nldbl-j1.c
nldbl-jn.c
nldbl-ldexp.c
nldbl-lgamma_r.c
nldbl-lgamma.c
nldbl-llrint.c
nldbl-llround.c
nldbl-log1p.c
nldbl-log2.c
nldbl-log10.c
nldbl-log.c
nldbl-logb.c
nldbl-lrint.c
nldbl-lround.c
nldbl-modf.c
nldbl-nan.c
nldbl-nearbyint.c
nldbl-nextafter.c
nldbl-nexttoward.c
nldbl-nexttowardf.c
nldbl-obstack_printf_chk.c
nldbl-obstack_printf.c
nldbl-obstack_vprintf_chk.c
nldbl-obstack_vprintf.c
nldbl-pow10.c
nldbl-pow.c
nldbl-printf_chk.c
nldbl-printf_fp.c
nldbl-printf_size.c
nldbl-printf.c
nldbl-qecvt_r.c
nldbl-qecvt.c
nldbl-qfcvt_r.c
nldbl-qfcvt.c
nldbl-qgcvt.c
nldbl-remainder.c
nldbl-remquo.c
nldbl-rint.c
nldbl-round.c
nldbl-scalb.c
nldbl-scalbln.c
nldbl-scalbn.c
nldbl-scanf.c
nldbl-signbit.c	This patch renames all uses of __isinf*, __isnan*, __finite* and __signbit* to use standard C99 macros. This has no effect on generated code.	2015-06-03 15:41:36 +01:00
nldbl-significand.c
nldbl-sin.c
nldbl-sincos.c
nldbl-sinh.c
nldbl-snprintf_chk.c
nldbl-snprintf.c
nldbl-sprintf_chk.c
nldbl-sprintf.c
nldbl-sqrt.c
nldbl-sscanf.c
nldbl-strfmon_l.c
nldbl-strfmon.c
nldbl-strtold_l.c
nldbl-strtold.c
nldbl-strtoldint.c
nldbl-swprintf_chk.c
nldbl-swprintf.c
nldbl-swscanf.c
nldbl-syslog_chk.c
nldbl-syslog.c
nldbl-tan.c
nldbl-tanh.c
nldbl-tgamma.c
nldbl-trunc.c
nldbl-vasprintf_chk.c
nldbl-vasprintf.c
nldbl-vdprintf_chk.c
nldbl-vdprintf.c
nldbl-vfprintf_chk.c
nldbl-vfprintf.c
nldbl-vfscanf.c
nldbl-vfwprintf_chk.c
nldbl-vfwprintf.c
nldbl-vfwscanf.c
nldbl-vprintf_chk.c
nldbl-vprintf.c
nldbl-vscanf.c
nldbl-vsnprintf_chk.c
nldbl-vsnprintf.c
nldbl-vsprintf_chk.c
nldbl-vsprintf.c
nldbl-vsscanf.c
nldbl-vswprintf_chk.c
nldbl-vswprintf.c
nldbl-vswscanf.c
nldbl-vsyslog_chk.c
nldbl-vsyslog.c
nldbl-vwprintf_chk.c
nldbl-vwprintf.c
nldbl-vwscanf.c
nldbl-wcstold_l.c
nldbl-wcstold.c
nldbl-wcstoldint.c
nldbl-wprintf_chk.c
nldbl-wprintf.c
nldbl-wscanf.c
nldbl-y0.c
nldbl-y1.c
nldbl-yn.c
s_asinh.c
s_atan.c
s_cacos.c
s_cacosh.c
s_cacoshl.c
s_cacosl.c
s_casin.c
s_casinh.c
s_casinhl.c
s_casinl.c
s_catan.c
s_catanh.c
s_catanhl.c
s_catanl.c
s_cbrt.c
s_ccos.c
s_ccosh.c
s_ccoshl.c
s_ccosl.c
s_ceil.c
s_cexp.c
s_cexpl.c
s_clog10.c
s_clog10l.c
s_clog.c
s_clogl.c
s_copysign.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_cpow.c
s_cpowl.c
s_cproj.c
s_cprojl.c
s_csin.c
s_csinh.c
s_csinhl.c
s_csinl.c
s_csqrt.c
s_csqrtl.c
s_ctan.c
s_ctanh.c
s_ctanhl.c
s_ctanl.c
s_erf.c
s_expm1.c
s_fabs.c
s_fdim.c
s_fdiml.c
s_finite.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_floor.c
s_fma.c
s_fmal.c
s_fmax.c
s_fmaxl.c
s_fmin.c
s_fminl.c
s_frexp.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_isinf.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_isnan.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_ldexp.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_ldexpl.c	Make scalbn set errno (bug 6803).	2015-09-16 21:11:00 +00:00
s_llrint.c
s_llround.c
s_log1p.c
s_logb.c
s_lrint.c
s_lround.c
s_modf.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_nan.c
s_nanl.c
s_nearbyint.c
s_nextafter.c
s_nexttowardfd.c	Fix nexttoward overflow in non-default rounding modes (bug 19059).	2015-10-02 17:11:13 +00:00
s_remquo.c
s_rint.c
s_round.c
s_scalbln.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_scalbn.c	Remove IS_IN_libm	2014-11-24 11:41:47 +05:30
s_significand.c
s_significandl.c
s_sin.c	Fixed powerpc64 build.	2015-06-19 20:42:34 +03:00
s_sincos.c
s_tan.c
s_tanh.c
s_trunc.c
Versions
w_acos.c
w_acosh.c
w_acoshl.c
w_acosl.c
w_asin.c
w_asinl.c
w_atan2.c
w_atan2l.c
w_atanh.c
w_atanhl.c
w_cosh.c
w_coshl.c
w_drem.c
w_dreml.c
w_exp10.c
w_exp10l.c
w_exp.c
w_fmod.c
w_fmodl.c
w_hypot.c
w_hypotl.c
w_ilogb.c
w_ilogbl.c
w_j0.c
w_j0l.c
w_j1.c
w_j1l.c
w_jn.c
w_jnl.c
w_lgamma_r.c
w_lgamma.c
w_lgammal_r.c
w_lgammal.c
w_log2.c
w_log2l.c
w_log10.c
w_log10l.c
w_log.c
w_logl.c
w_pow.c
w_powl.c
w_remainder.c
w_remainderl.c
w_scalb.c
w_scalbl.c
w_sinh.c
w_sinhl.c
w_sqrt.c
w_sqrtl.c
w_tgamma.c
w_tgammal.c