AuroraMiddleware/glibc
1
0
Fork 0
mirror of https://sourceware.org/git/glibc.git synced 2024-12-26 12:41:05 +00:00
Code Issues Packages Projects Releases Wiki Activity
047512374a
glibc/math/auto-libm-test-in

8474 lines
225 KiB
Plaintext
Raw Normal View History

Start generating libm tests automatically with MPFR.
2013-11-29 16:27:55 +00:00
# libm test inputs for gen-auto-libm-tests.c.
Update copyright dates with scripts/update-copyrights I used these shell commands: ../glibc/scripts/update-copyrights $PWD/../gnulib/build-aux/update-copyright (cd ../glibc && git commit -am"[this commit message]") and then ignored the output, which consisted lines saying "FOO: warning: copyright statement not found" for each of 7061 files FOO. I then removed trailing white space from math/tgmath.h, support/tst-support-open-dev-null-range.c, and sysdeps/x86_64/multiarch/strlen-vec.S, to work around the following obscure pre-commit check failure diagnostics from Savannah. I don't know why I run into these diagnostics whereas others evidently do not. remote: *** 912-#endif remote: *** 913: remote: *** 914- remote: *** error: lines with trailing whitespace found ... remote: *** error: sysdeps/unix/sysv/linux/statx_cp.c: trailing lines
2022-01-01 18:54:23 +00:00
# Copyright (C) 1997-2022 Free Software Foundation, Inc.
Start generating libm tests automatically with MPFR.
2013-11-29 16:27:55 +00:00
# This file is part of the GNU C Library.
#
# The GNU C Library is free software; you can redistribute it and/or
# modify it under the terms of the GNU Lesser General Public
# License as published by the Free Software Foundation; either
# version 2.1 of the License, or (at your option) any later version.
#
# The GNU C Library is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
# Lesser General Public License for more details.
#
# You should have received a copy of the GNU Lesser General Public
# License along with the GNU C Library; if not, see
Prefer https to http for gnu.org and fsf.org URLs Also, change sources.redhat.com to sourceware.org. This patch was automatically generated by running the following shell script, which uses GNU sed, and which avoids modifying files imported from upstream: sed -ri ' s,(http|ftp)(://(.*\.)?(gnu|fsf|sourceware)\.org($|[^.]|\.[^a-z])),https\2,g s,(http|ftp)(://(.*\.)?)sources\.redhat\.com($|[^.]|\.[^a-z]),https\2sourceware.org\4,g ' \ $(find $(git ls-files) -prune -type f \ ! -name '*.po' \ ! -name 'ChangeLog*' \ ! -path COPYING ! -path COPYING.LIB \ ! -path manual/fdl-1.3.texi ! -path manual/lgpl-2.1.texi \ ! -path manual/texinfo.tex ! -path scripts/config.guess \ ! -path scripts/config.sub ! -path scripts/install-sh \ ! -path scripts/mkinstalldirs ! -path scripts/move-if-change \ ! -path INSTALL ! -path locale/programs/charmap-kw.h \ ! -path po/libc.pot ! -path sysdeps/gnu/errlist.c \ ! '(' -name configure \ -execdir test -f configure.ac -o -f configure.in ';' ')' \ ! '(' -name preconfigure \ -execdir test -f preconfigure.ac ';' ')' \ -print) and then by running 'make dist-prepare' to regenerate files built from the altered files, and then executing the following to cleanup: chmod a+x sysdeps/unix/sysv/linux/riscv/configure # Omit irrelevant whitespace and comment-only changes, # perhaps from a slightly-different Autoconf version. git checkout -f \ sysdeps/csky/configure \ sysdeps/hppa/configure \ sysdeps/riscv/configure \ sysdeps/unix/sysv/linux/csky/configure # Omit changes that caused a pre-commit check to fail like this: # remote: *** error: sysdeps/powerpc/powerpc64/ppc-mcount.S: trailing lines git checkout -f \ sysdeps/powerpc/powerpc64/ppc-mcount.S \ sysdeps/unix/sysv/linux/s390/s390-64/syscall.S # Omit change that caused a pre-commit check to fail like this: # remote: *** error: sysdeps/sparc/sparc64/multiarch/memcpy-ultra3.S: last line does not end in newline git checkout -f sysdeps/sparc/sparc64/multiarch/memcpy-ultra3.S
2019-09-07 05:40:42 +00:00
# <https://www.gnu.org/licenses/>. */
Start generating libm tests automatically with MPFR.
2013-11-29 16:27:55 +00:00
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
acos 0
acos -0
acos 1
acos -1
acos 0.5
acos -0.5
acos 0.75
acos 2e-17
acos 0.0625
acos 0x0.ffffffp0
acos -0x0.ffffffp0
acos 0x0.ffffffff8p0
acos -0x0.ffffffff8p0
acos 0x0.ffffffffffffp0
acos -0x0.ffffffffffffp0
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
acos 0x0.fffffffffffff8p0
acos -0x0.fffffffffffff8p0
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
acos 0x0.ffffffffffffffffp0
acos -0x0.ffffffffffffffffp0
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
acos 0x0.ffffffffffffffffffffffffffcp0
acos -0x0.ffffffffffffffffffffffffffcp0
acos 0x0.ffffffffffffffffffffffffffff8p0
acos -0x0.ffffffffffffffffffffffffffff8p0
Fix ldbl-128/ldbl-128ibm acosl inaccuracy (bug 18038, bug 18039). The ldbl-128 and ldbl-128ibm implementations of acosl have similar bugs, using a threshold of 0x1p-57L to determine when they just return pi/2. Since the result pi/2 - asinl (x) is roughly pi/2 - x for small x, the relevant cut-off is actually x being < 0.5ulp of 1. This patch fixes the implementations to use that cut-off and adds tests of small acos arguments. Tested for powerpc and mips64. Also tested for x86_64 and x86; no ulps updates needed. [BZ #18038] [BZ #18039] * sysdeps/ieee754/ldbl-128/e_acosl.c (__ieee754_acosl): Only return pi/2 for arguments below 0x1p-113L. * sysdeps/ieee754/ldbl-128ibm/e_acosl.c (__ieee754_acosl): Only return pi/2 for arguments below 0x1p-106L. * math/auto-libm-test-in: Add more tests of acos. * math/auto-libm-test-out: Regenerated.
2015-02-26 21:06:34 +00:00
acos 0x1p-5
acos 0x1p-10
acos 0x1p-15
acos 0x1p-20
acos 0x1p-25
acos 0x1p-30
acos 0x1p-35
acos 0x1p-40
acos 0x1p-45
acos 0x1p-50
acos 0x1p-55
acos 0x1p-60
acos 0x1p-65
acos 0x1p-70
acos 0x1p-75
acos 0x1p-80
acos 0x1p-85
acos 0x1p-90
acos 0x1p-95
acos 0x1p-100
acos 0x1p-105
acos 0x1p-110
acos 0x1p-115
acos 0x1p-120
acos -0x1p-5
acos -0x1p-25
acos -0x1p-45
acos -0x1p-65
acos -0x1p-85
acos -0x1p-105
acos -0x1p-125
Add more tests of acos. This patch adds some randomly-generated tests of acos that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-25 00:30:10 +00:00
acos -0x2.0089a4p-4
acos 0xf.04aeep-4
acos 0x5.dd258006121b8p-4
acos -0x2.35f051e70dbc4p-4
acos 0xe.9a5c0d7fabb9aa1p-4
acos 0xe.17513589de79b75p-4
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
acos 0x3.e57821d368ebap-4
acos 0x2.0bee8p-4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
acos -0x1.da00d8p-4
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
acos 0x1.fffff3634acd6p-1
# the next value generates larger error bounds on x86_64 (ldbl-96)
acos 0xf.fe00271d507ee5dp-4
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
acos min
acos -min
acos min_subnorm
acos -min_subnorm
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
acosh 1
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
acosh 0x1.000002p0
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
acosh 0x1.0000000000001p0
acosh 0x1.0000000000000002p0
acosh 0x1.000000000000000000000000008p0
acosh 0x1.0000000000000000000000000001p0
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
acosh 1.625
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
acosh 7
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
acosh 100
acosh 1e5
Fix ldbl-128ibm acoshl inaccuracy (bug 18019). The ldbl-128ibm implementation of acoshl uses a cut-off of 0x1p28 to determine when to use log(x) + log(2) as a formula. That cut-off is too small for this format, resulting in large errors. This patch changes it to a more appropriate cut-off of 0x1p56, adding tests around the cut-offs for various floating-point formats. Tested for powerpc. Also tested for x86_64 and x86 and updated ulps. [BZ #18019] * sysdeps/ieee754/ldbl-128ibm/e_acoshl.c (__ieee754_acoshl): Use 2**56 not 2**28 as threshold for log (2x) formula. * math/auto-libm-test-in: Add more tests of acosh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-02-25 00:01:15 +00:00
acosh 0x1p8
acosh 0x1p9
acosh 0x1p10
acosh 0x1p11
acosh 0x1p12
acosh 0x1p13
acosh 0x1p24
acosh 0x1p25
acosh 0x1p26
acosh 0x1p27
acosh 0x1p28
acosh 0x1p29
acosh 0x1p30
acosh 0x1p31
acosh 0x1p32
acosh 0x1p33
acosh 0x1p48
acosh 0x1p49
acosh 0x1p50
acosh 0x1p51
acosh 0x1p52
acosh 0x1p53
acosh 0x1p54
acosh 0x1p55
acosh 0x1p56
acosh 0x1p57
acosh 0x1p58
acosh 0x1p59
acosh 0x1p100
acosh 0x1p500
acosh 0x1p5000
Add more tests of acosh, asinh and atanh. This patch adds some randomly-generated tests of acosh, asinh and atanh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, asinh and atanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-25 22:21:20 +00:00
acosh 0x1.80a368p+0
acosh 0x1.0b9d3e9fc19fbp+0
acosh 0x1.11eab6p+0
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
acosh 0x1.0fffaap+0
Add more tests of acosh, atanh, cos, csqrt, erfc, sin, sincos. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, cos, csqrt, erfc, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-06 17:30:18 +00:00
acosh 0x1.068e0eca105a6p+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
acosh 0x2.8d4288p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
acosh 0x1.5d71a6p+36
acosh 0x1.476a3c43d7edep+288
acosh 0xc.84cb1dbbd1b1p+0
acosh 0x1.0722362c26ba5p+0
Add more randomly-generated libm tests. This patch adds more libm test inputs found through random test generation to increase observed ulps on x86_64. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, atanh, cbrt, cosh, csqrt, erfc, expm1 and lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-11 15:03:10 +00:00
acosh 0x1.73515p+0
acosh 0x1.1b836p+0
acosh 0x1.07c956p+0
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
acosh 0x1.1808eep+0
acosh 0x1.1052c4p+0
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (ldbl-96)
acosh 0x1.1ecbdf374bce01cap+0
# the next value generates larger error bounds on x86_64 (binary128)
acosh 0x1.0f97586eba090200118df0902f99p+0
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
acosh max
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
Add narrowing add functions. This patch adds the narrowing add functions from TS 18661-1 to glibc's libm: fadd, faddl, daddl, f32addf64, f32addf32x, f32xaddf64 for all configurations; f32addf64x, f32addf128, f64addf64x, f64addf128, f32xaddf64x, f32xaddf128, f64xaddf128 for configurations with _Float64x and _Float128; __nldbl_daddl for ldbl-opt. As discussed for the build infrastructure patch, tgmath.h support is deliberately deferred, and FP_FAST_* macros are not applicable without optimized function implementations. Function implementations are added for all relevant pairs of formats (including certain cases of a format and itself where more than one type has that format). The main implementations use round-to-odd, or a trivial computation in the case where both formats are the same or where the wider format is IBM long double (in which case we don't attempt to be correctly rounding). The sysdeps/ieee754/soft-fp implementations use soft-fp, and are used automatically for configurations without exceptions and rounding modes by virtue of existing Implies files. As previously discussed, optimized versions for particular architectures are possible, but not included. i386 gets a special version of f32xaddf64 to avoid problems with double rounding (similar to the existing fdim version), since this function must round just once without an intermediate rounding to long double. (No such special version is needed for any other function, because the nontrivial functions use round-to-odd, which does the intermediate computation with the rounding mode set to round-to-zero, and double rounding is OK except in round-to-nearest mode, so is OK for that intermediate round-to-zero computation.) mul and div will need slightly different special versions for i386 (using round-to-odd on long double instead of precision control) because of the possibility of inexact intermediate results in the subnormal range for double. To reduce duplication among the different function implementations, math-narrow.h gets macros CHECK_NARROW_ADD, NARROW_ADD_ROUND_TO_ODD and NARROW_ADD_TRIVIAL. In the trivial cases and for any architecture-specific optimized implementations, the overhead of the errno setting might be significant, but I think that's best handled through compiler built-in functions rather than providing separate no-errno versions in glibc (and likewise there are no __*_finite entry points for these function provided, __*_finite effectively being no-errno versions at present in most cases). Tested for x86_64 and x86, with both GCC 6 and GCC 7. Tested for mips64 (all three ABIs, both hard and soft float) and powerpc with GCC 7. Tested with build-many-glibcs.py with both GCC 6 and GCC 7. * math/Makefile (libm-narrow-fns): Add add. (libm-test-funcs-narrow): Likewise. * math/Versions (GLIBC_2.28): Add narrowing add functions. * math/bits/mathcalls-narrow.h (add): Use __MATHCALL_NARROW . * math/gen-auto-libm-tests.c (test_functions): Add add. * math/math-narrow.h (CHECK_NARROW_ADD): New macro. (NARROW_ADD_ROUND_TO_ODD): Likewise. (NARROW_ADD_TRIVIAL): Likewise. * sysdeps/ieee754/float128/float128_private.h (__faddl): New macro. (__daddl): Likewise. * sysdeps/ieee754/ldbl-opt/Makefile (libnldbl-calls): Add fadd and dadd. (CFLAGS-nldbl-dadd.c): New variable. (CFLAGS-nldbl-fadd.c): Likewise. * sysdeps/ieee754/ldbl-opt/Versions (GLIBC_2.28): Add __nldbl_daddl. * sysdeps/ieee754/ldbl-opt/nldbl-compat.h (__nldbl_daddl): New prototype. * manual/arith.texi (Misc FP Arithmetic): Document fadd, faddl, daddl, fMaddfN, fMaddfNx, fMxaddfN and fMxaddfNx. * math/auto-libm-test-in: Add tests of add. * math/auto-libm-test-out-narrow-add: New generated file. * math/libm-test-narrow-add.inc: New file. * sysdeps/i386/fpu/s_f32xaddf64.c: Likewise. * sysdeps/ieee754/dbl-64/s_f32xaddf64.c: Likewise. * sysdeps/ieee754/dbl-64/s_fadd.c: Likewise. * sysdeps/ieee754/float128/s_f32addf128.c: Likewise. * sysdeps/ieee754/float128/s_f64addf128.c: Likewise. * sysdeps/ieee754/float128/s_f64xaddf128.c: Likewise. * sysdeps/ieee754/ldbl-128/s_daddl.c: Likewise. * sysdeps/ieee754/ldbl-128/s_f64xaddf128.c: Likewise. * sysdeps/ieee754/ldbl-128/s_faddl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_daddl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_faddl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_daddl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_faddl.c: Likewise. * sysdeps/ieee754/ldbl-opt/nldbl-dadd.c: Likewise. * sysdeps/ieee754/ldbl-opt/nldbl-fadd.c: Likewise. * sysdeps/ieee754/soft-fp/s_daddl.c: Likewise. * sysdeps/ieee754/soft-fp/s_fadd.c: Likewise. * sysdeps/ieee754/soft-fp/s_faddl.c: Likewise. * sysdeps/powerpc/fpu/libm-test-ulps: Update. * sysdeps/mach/hurd/i386/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/aarch64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/alpha/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/arm/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/hppa/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/i386/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/m68k/coldfire/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/m68k/m680x0/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/microblaze/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/mips/mips32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/mips/mips64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/nios2/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/fpu/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/nofpu/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/libm-le.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/riscv/rv64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/s390/s390-32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/s390/s390-64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sh/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/tile/tilegx32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/tile/tilegx64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/x86_64/64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/x86_64/x32/libm.abilist: Likewise.
2018-02-10 02:08:43 +00:00
add 0 0
add 0 -0
add -0 0
add -0 -0
add max max
add max -max
add -max max
add -max -max
add min min missing-underflow:arg-ibm128
add min -min missing-underflow:arg-ibm128
add -min min missing-underflow:arg-ibm128
add -min -min missing-underflow:arg-ibm128
add min_subnorm min_subnorm missing-underflow:arg-ibm128
add min_subnorm -min_subnorm missing-underflow:arg-ibm128
add -min_subnorm min_subnorm missing-underflow:arg-ibm128
add -min_subnorm -min_subnorm missing-underflow:arg-ibm128
add 1 2
add 1 -2
add -1 2
add -1 -2
add 100.5 0.75
add 100.5 -0.75
add -100.5 0.75
add -100.5 -0.75
add 1 0x1p-23
add 1 0x1.7fp-23
add 1 0x1.8p-23
add 1 0x1.81p-23
add 1 0x1p-24
add 1 0x1.1p-24
add 1 0x0.fp-24
add 1 min
add 1 -min
add 1 min_subnorm
add 1 -min_subnorm
add -1 min
add -1 -min
add -1 min_subnorm
add -1 -min_subnorm
# Cases where larger argument is half way between two values of a
# floating-point format, so that double rounding would sometimes yield
# the wrong result.
add 0x1.000001p0 min
add 0x1.000001p0 -min
add 0x1.000001p0 min_subnorm
add 0x1.000001p0 -min_subnorm
add -0x1.000001p0 min
add -0x1.000001p0 -min
add -0x1.000001p0 min_subnorm
add -0x1.000001p0 -min_subnorm
add 0x1.00000000000008p0 min
add 0x1.00000000000008p0 -min
add 0x1.00000000000008p0 min_subnorm
add 0x1.00000000000008p0 -min_subnorm
add -0x1.00000000000008p0 min
add -0x1.00000000000008p0 -min
add -0x1.00000000000008p0 min_subnorm
add -0x1.00000000000008p0 -min_subnorm
add 0x1.0000000000000001p0 min
add 0x1.0000000000000001p0 -min
add 0x1.0000000000000001p0 min_subnorm
add 0x1.0000000000000001p0 -min_subnorm
add -0x1.0000000000000001p0 min
add -0x1.0000000000000001p0 -min
add -0x1.0000000000000001p0 min_subnorm
add -0x1.0000000000000001p0 -min_subnorm
# Two values representable in the same format with sum very close to
# half way between two representable values. In particular, verify
# that there is no intermediate rounding to 64 bits when adding two
# double values.
add 1 0x1.000002p-24
add 1 -0x1.000002p-24
add 1 0x0.ffffffp-24
add 1 -0x0.ffffffp-24
add 0x1.000002p0 0x1.000002p-24
add 0x1.000002p0 -0x1.000002p-24
add 0x1.000002p0 0x0.ffffffp-24
add 0x1.000002p0 -0x0.ffffffp-24
add 1 0x1.0000000000001p-53
add 1 -0x1.0000000000001p-53
add 1 0x0.fffffffffffff8p-53
add 1 -0x0.fffffffffffff8p-53
add 0x1.0000000000001p0 0x1.0000000000001p-53
add 0x1.0000000000001p0 -0x1.0000000000001p-53
add 0x1.0000000000001p0 0x0.fffffffffffff8p-53
add 0x1.0000000000001p0 -0x0.fffffffffffff8p-53
add 1 0x1.0000000000000002p-64
add 1 -0x1.0000000000000002p-64
add 1 0x0.ffffffffffffffffp-64
add 1 -0x0.ffffffffffffffffp-64
add 0x1.0000000000000002p0 0x1.0000000000000002p-64
add 0x1.0000000000000002p0 -0x1.0000000000000002p-64
add 0x1.0000000000000002p0 0x0.ffffffffffffffffp-64
add 0x1.0000000000000002p0 -0x0.ffffffffffffffffp-64
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
asin 0
asin -0
asin 0.5
asin -0.5
asin 1.0
asin -1.0
asin 0.75
asin 0x0.ffffffp0
asin -0x0.ffffffp0
asin 0x0.ffffffff8p0
asin -0x0.ffffffff8p0
asin 0x0.ffffffffffffp0
asin -0x0.ffffffffffffp0
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
asin 0x0.fffffffffffff8p0
asin -0x0.fffffffffffff8p0
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
asin 0x0.ffffffffffffffffp0
asin -0x0.ffffffffffffffffp0
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
asin 0x0.ffffffffffffffffffffffffffcp0
asin -0x0.ffffffffffffffffffffffffffcp0
asin 0x0.ffffffffffffffffffffffffffff8p0
asin -0x0.ffffffffffffffffffffffffffff8p0
asin 0x1p-5
asin 0x1p-10
asin 0x1p-15
asin 0x1p-20
asin 0x1p-25
asin 0x1p-30
asin 0x1p-35
asin 0x1p-40
asin 0x1p-45
asin 0x1p-50
asin 0x1p-55
asin 0x1p-60
asin 0x1p-100
asin 0x1p-600
asin 0x1p-10000
Add more tests of asin. This patch adds some randomly-generated tests of asin that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of asin. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-25 17:53:58 +00:00
asin -0x2.18915cp-4
asin -0x3.746774p-4
asin -0x3.1c54d10e5c844p-4
asin 0xf.c9675fa6fe69f12p-4
Add another test of asin. This patch adds a randomly-generated test of asin that is observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add another test of asin. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-25 21:57:04 +00:00
asin -0xa.fc5afp-4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
asin 0xd.805e5p-4
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
asin -0x2.1a02dcp-4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
asin -0x8.1c120ac3e495p-4
asin -0xb.36dacp-4
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
asin 0x1.f6041ffe89d7dp-4
Fix asin missing underflows (bug 16351). Similar to various other bugs in this area, some asin implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86, powerpc and mips64. [BZ #16351] * sysdeps/i386/fpu/e_asin.S (dbl_min): New object. (MO): New macro. (__ieee754_asin): Force underflow exception for results with small absolute value. * sysdeps/i386/fpu/e_asinf.S (flt_min): New object. (MO): New macro. (__ieee754_asinf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/e_asin.c: Include <float.h> and <math.h>. (__ieee754_asin): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/e_asinf.c: Include <float.h>. (__ieee754_asinf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/e_asinl.c: Include <float.h>. (__ieee754_asinl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128ibm/e_asinl.c: Include <float.h>. (__ieee754_asinl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-96/e_asinl.c: Include <float.h>. (__ieee754_asinl): Force underflow exception for results with small absolute value. * sysdeps/x86_64/fpu/multiarch/e_asin.c [HAVE_FMA4_SUPPORT]: Include <math.h>. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 16351. * math/auto-libm-test-out: Regenerated.
2015-02-26 17:18:54 +00:00
asin min
asin -min
asin min_subnorm
asin -min_subnorm
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
asinh 0
asinh -0
asinh 0.75
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
asinh 1
asinh 10
asinh 100
asinh 1e6
Fix ldbl-128ibm asinhl inaccuracy (bug 18020). The ldbl-128ibm implementation of asinhl uses cut-offs of 0x1p28 and 0x1p-29 to determine when to use simpler formulas that avoid possible overflow / underflow. Both those cut-offs are inappropriate for this format, resulting in large errors. This patch changes the code to use more appropriate cut-offs of 0x1p56 and 0x1p-56, adding tests around the cut-offs for various floating-point formats. Tested for powerpc. Also tested for x86_64 and x86 and updated ulps. [BZ #18020] * sysdeps/ieee754/ldbl-128ibm/s_asinhl.c (__asinhl): Use 2**56 and 2**-56 not 2**28 and 2**-29 as thresholds for simpler formulas. * math/auto-libm-test-in: Add more tests of asinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-02-25 11:13:41 +00:00
asinh 0x1p8
asinh 0x1p9
asinh 0x1p10
asinh 0x1p11
asinh 0x1p12
asinh 0x1p13
asinh 0x1p24
asinh 0x1p25
asinh 0x1p26
asinh 0x1p27
asinh 0x1p28
asinh 0x1p29
asinh 0x1p30
asinh 0x1p31
asinh 0x1p32
asinh 0x1p33
asinh 0x1p48
asinh 0x1p49
asinh 0x1p50
asinh 0x1p51
asinh 0x1p52
asinh 0x1p53
asinh 0x1p54
asinh 0x1p55
asinh 0x1p56
asinh 0x1p57
asinh 0x1p58
asinh 0x1p59
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
asinh 0x1p100
Fix ldbl-128ibm asinhl inaccuracy (bug 18020). The ldbl-128ibm implementation of asinhl uses cut-offs of 0x1p28 and 0x1p-29 to determine when to use simpler formulas that avoid possible overflow / underflow. Both those cut-offs are inappropriate for this format, resulting in large errors. This patch changes the code to use more appropriate cut-offs of 0x1p56 and 0x1p-56, adding tests around the cut-offs for various floating-point formats. Tested for powerpc. Also tested for x86_64 and x86 and updated ulps. [BZ #18020] * sysdeps/ieee754/ldbl-128ibm/s_asinhl.c (__asinhl): Use 2**56 and 2**-56 not 2**28 and 2**-29 as thresholds for simpler formulas. * math/auto-libm-test-in: Add more tests of asinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-02-25 11:13:41 +00:00
asinh 0x1p500
asinh 0x1p5000
asinh 0x1p-8
asinh 0x1p-9
asinh 0x1p-10
asinh 0x1p-11
asinh 0x1p-12
asinh 0x1p-13
asinh 0x1p-24
asinh 0x1p-25
asinh 0x1p-26
asinh 0x1p-27
asinh 0x1p-28
asinh 0x1p-29
asinh 0x1p-30
asinh 0x1p-31
asinh 0x1p-32
asinh 0x1p-33
asinh 0x1p-48
asinh 0x1p-49
asinh 0x1p-50
asinh 0x1p-51
asinh 0x1p-52
asinh 0x1p-53
asinh 0x1p-54
asinh 0x1p-55
asinh 0x1p-56
asinh 0x1p-57
asinh 0x1p-58
asinh 0x1p-59
asinh 0x1p-100
Add more tests of acosh, asinh and atanh. This patch adds some randomly-generated tests of acosh, asinh and atanh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, asinh and atanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-25 22:21:20 +00:00
asinh -0x3.d26bb4p-4
asinh -0x3.bdeef4p-4
asinh -0x7.fc7fc8p-8
asinh -0x3.b94a52e6913c2p-4
asinh 0x7.d8e5a8p-4
asinh -0x7.63a06320c42e4p-4
asinh 0x6.f4a93p-4
asinh -0x7.88bcc8p-4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
asinh -0x3.11c35p-4
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
asinh -0x4.39534p-4
asinh -0xd.d62e8p+92
asinh -0x4.cfb9805a53a2065p-4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
asinh -0x5.cabae7a011e33d9p-4
asinh -0x6.e26358p-4
asinh 0x6.c92c08p-4
Fix asinh missing underflows (bug 16350). Similar to various other bugs in this area, some asinh implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86 and mips64. [BZ #16350] * sysdeps/i386/fpu/s_asinh.S (__asinh): Force underflow exception for arguments with small absolute value. * sysdeps/i386/fpu/s_asinhf.S (__asinhf): Likewise. * sysdeps/i386/fpu/s_asinhl.S (__asinhl): Likewise. * sysdeps/ieee754/dbl-64/s_asinh.c: Include <float.h>. (__asinh): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/s_asinhf.c: Include <float.h>. (__asinhf): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/s_asinhl.c: Include <float.h>. (__asinhl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/s_asinhl.c: Include <float.h>. (__asinhl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-96/s_asinhl.c: Include <float.h>. (__asinhl): Force underflow exception for arguments with small absolute value. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 16350. * math/auto-libm-test-out: Regenerated.
2015-06-18 23:27:41 +00:00
asinh 0x1p-500
asinh 0x1p-5000
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
asinh 0xf.96c69p-8
add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)
2020-11-25 18:03:33 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
asinh 0x1.023cfeab10a13p-2
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary128)
asinh -0x5.a95683e302a71be88a35649b24ep-4
Fix asinh missing underflows (bug 16350). Similar to various other bugs in this area, some asinh implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86 and mips64. [BZ #16350] * sysdeps/i386/fpu/s_asinh.S (__asinh): Force underflow exception for arguments with small absolute value. * sysdeps/i386/fpu/s_asinhf.S (__asinhf): Likewise. * sysdeps/i386/fpu/s_asinhl.S (__asinhl): Likewise. * sysdeps/ieee754/dbl-64/s_asinh.c: Include <float.h>. (__asinh): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/s_asinhf.c: Include <float.h>. (__asinhf): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/s_asinhl.c: Include <float.h>. (__asinhl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/s_asinhl.c: Include <float.h>. (__asinhl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-96/s_asinhl.c: Include <float.h>. (__asinhl): Force underflow exception for arguments with small absolute value. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 16350. * math/auto-libm-test-out: Regenerated.
2015-06-18 23:27:41 +00:00
asinh min
asinh -min
asinh min_subnorm
asinh -min_subnorm
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
asinh max
asinh -max
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
atan 0
atan -0
atan max
atan -max
atan 1
atan -1
atan 0.75
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
atan 0x1p-5
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
atan 0x1p-10
atan 0x1p-15
atan 0x1p-20
atan 0x1p-25
atan 0x1p-30
atan 0x1p-35
atan 0x1p-40
atan 0x1p-45
atan 0x1p-50
atan 0x1p-55
atan 0x1p-60
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
atan 2.5
atan 10
atan 1e6
Fix atanf spurious underflows (bug 18196). The sysdeps/ieee754/flt-32 version of atanf produces spurious underflow exceptions for some large arguments, because of computations that compute x^-4. This patch fixes this by adjusting the threshold for large arguments (for which +/- pi/2 can just be returned, the correct result being roughly +/- pi/2 - 1/x) from 2^34 to 2^25. Tested for x86_64 and x86. [BZ #18196] * sysdeps/ieee754/flt-32/s_atanf.c (__atanf): Use 2^25 not 2^34 as threshold for large arguments. * math/auto-libm-test-in: Add another test of atan. * math/auto-libm-test-out: Regenerated.
2015-05-14 23:51:09 +00:00
atan 0x1p31
Fix atan / atan2 missing underflows (bug 15319). This patch fixes bug 15319, missing underflows from atan / atan2 when the result of atan is very close to its small argument (or that of atan2 is very close to the ratio of its arguments, which may be an exact division). The usual approach of doing an underflowing computation if the computed result is subnormal is followed. For 32-bit x86, there are extra complications: the inline __ieee754_atan2 in bits/mathinline.h needs to be disabled for float and double because other libm functions using it generally rely on getting proper underflow exceptions from it, while the out-of-line functions have to remove excess range and precision from the underflowing result so as to return an exact 0 in the case where errno should be set for underflow to 0. (The failures I saw without that are similar to those Carlos reported for other functions, where I haven't seen a response to <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html> confirming if my diagnosis is correct. Arguably all libm functions with float and double returns should remove excess range and precision, but that's a separate matter.) The x86_64 long double case reported in a comment in bug 15319 is not a bug (it's an argument of LDBL_MIN, and x86_64 is an after-rounding architecture so the correct IEEE result is not to raise underflow in the given rounding mode, in addition to treating the result as an exact LDBL_MIN being within the newly clarified documentation of accuracy goals). I'm presuming that the fpatan instruction can be trusted to raise appropriate exceptions when the (long double) result underflows (after rounding) and so no changes are needed for x86 / x86_64 long double functions here; empirically this is the case for the cases covered in the testsuite, on my system. Tested for x86_64, x86, powerpc and mips64. Only 32-bit x86 needs ulps updates (for the changes to inlines meaning some functions no longer get excess precision from their __ieee754_atan2* calls). [BZ #15319] * sysdeps/i386/fpu/e_atan2.S (dbl_min): New object. (MO): New macro. (__ieee754_atan2): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/e_atan2f.S (flt_min): New object. (MO): New macro. (__ieee754_atan2f): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/s_atan.S (dbl_min): New object. (MO): New macro. (__atan): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/s_atanf.S (flt_min): New object. (MO): New macro. (__atanf): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/ieee754/dbl-64/e_atan2.c: Include <float.h> and <math.h>. (__ieee754_atan2): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/s_atan.c: Include <float.h> and <math_private.h>. (atan): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/s_atanf.c: Include <float.h>. (__atanf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/s_atanl.c: Include <float.h> and <math.h>. (__atanl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128ibm/s_atanl.c: Include <float.h>. (__atanl): Force underflow exception for results with small absolute value. * sysdeps/x86/fpu/bits/mathinline.h [!__SSE2_MATH__ && !__x86_64__ && __LIBC_INTERNAL_MATH_INLINES] (__ieee754_atan2): Only define inline for long double. * sysdeps/x86_64/fpu/multiarch/e_atan2.c [HAVE_FMA4_SUPPORT || HAVE_AVX_SUPPORT]: Include <math.h>. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 15319. Add more tests of atan2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (casin_test_data): Do not mark underflow exceptions as possibly missing for bug 15319. (casinh_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-02-18 21:10:49 +00:00
atan 0x1p-100
atan 0x1p-600
atan 0x1p-10000
Fix dbl-64 atan in non-default rounding modes (bug 18197). The dbl-64 implementation of atan does computations that expect to run in round-to-nearest mode, and in other modes the errors can accumulate to more than the maximum accepted 9ulp. This patch makes it use FE_TONEAREST internally, similar to other functions with such issues. Tested for x86_64 and x86; no ulps updates needed. [BZ #18197] * sysdeps/ieee754/dbl-64/s_atan.c: Include <fenv.h>. (atan): Set FE_TONEAREST mode for internal computations. * math/auto-libm-test-in: Add more tests of atan. * math/auto-libm-test-out: Regenerated.
2015-04-08 17:14:12 +00:00
atan -0x3.b02d84p-4
atan -0x3.3fb708p-4
Add more tests of atan. This patch adds some randomly-generated tests of atan that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-08 21:00:03 +00:00
atan -0x2.3249ap+0
atan -0x1.363f46p+0
atan -0x1.ad4c0ap+0
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
atan -0x3.eb8e18p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
atan 0x3.53c188p+0
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
atan -0x1.58c83p+0
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
atan 0x1.626772p-1
Fix atan / atan2 missing underflows (bug 15319). This patch fixes bug 15319, missing underflows from atan / atan2 when the result of atan is very close to its small argument (or that of atan2 is very close to the ratio of its arguments, which may be an exact division). The usual approach of doing an underflowing computation if the computed result is subnormal is followed. For 32-bit x86, there are extra complications: the inline __ieee754_atan2 in bits/mathinline.h needs to be disabled for float and double because other libm functions using it generally rely on getting proper underflow exceptions from it, while the out-of-line functions have to remove excess range and precision from the underflowing result so as to return an exact 0 in the case where errno should be set for underflow to 0. (The failures I saw without that are similar to those Carlos reported for other functions, where I haven't seen a response to <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html> confirming if my diagnosis is correct. Arguably all libm functions with float and double returns should remove excess range and precision, but that's a separate matter.) The x86_64 long double case reported in a comment in bug 15319 is not a bug (it's an argument of LDBL_MIN, and x86_64 is an after-rounding architecture so the correct IEEE result is not to raise underflow in the given rounding mode, in addition to treating the result as an exact LDBL_MIN being within the newly clarified documentation of accuracy goals). I'm presuming that the fpatan instruction can be trusted to raise appropriate exceptions when the (long double) result underflows (after rounding) and so no changes are needed for x86 / x86_64 long double functions here; empirically this is the case for the cases covered in the testsuite, on my system. Tested for x86_64, x86, powerpc and mips64. Only 32-bit x86 needs ulps updates (for the changes to inlines meaning some functions no longer get excess precision from their __ieee754_atan2* calls). [BZ #15319] * sysdeps/i386/fpu/e_atan2.S (dbl_min): New object. (MO): New macro. (__ieee754_atan2): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/e_atan2f.S (flt_min): New object. (MO): New macro. (__ieee754_atan2f): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/s_atan.S (dbl_min): New object. (MO): New macro. (__atan): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/s_atanf.S (flt_min): New object. (MO): New macro. (__atanf): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/ieee754/dbl-64/e_atan2.c: Include <float.h> and <math.h>. (__ieee754_atan2): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/s_atan.c: Include <float.h> and <math_private.h>. (atan): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/s_atanf.c: Include <float.h>. (__atanf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/s_atanl.c: Include <float.h> and <math.h>. (__atanl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128ibm/s_atanl.c: Include <float.h>. (__atanl): Force underflow exception for results with small absolute value. * sysdeps/x86/fpu/bits/mathinline.h [!__SSE2_MATH__ && !__x86_64__ && __LIBC_INTERNAL_MATH_INLINES] (__ieee754_atan2): Only define inline for long double. * sysdeps/x86_64/fpu/multiarch/e_atan2.c [HAVE_FMA4_SUPPORT || HAVE_AVX_SUPPORT]: Include <math.h>. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 15319. Add more tests of atan2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (casin_test_data): Do not mark underflow exceptions as possibly missing for bug 15319. (casinh_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-02-18 21:10:49 +00:00
atan min
atan -min
atan min_subnorm
atan -min_subnorm
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
# atan2 (0,x) == 0 for x > 0.
atan2 0 1
# atan2 (-0,x) == -0 for x > 0.
atan2 -0 1
atan2 0 0
atan2 -0 0
# atan2 (+0,x) == +pi for x < 0.
atan2 0 -1
# atan2 (-0,x) == -pi for x < 0.
atan2 -0 -1
atan2 0 -0
atan2 -0 -0
# atan2 (y,+0) == pi/2 for y > 0.
atan2 1 0
# atan2 (y,-0) == pi/2 for y > 0.
atan2 1 -0
# atan2 (y,+0) == -pi/2 for y < 0.
atan2 -1 0
# atan2 (y,-0) == -pi/2 for y < 0.
atan2 -1 -0
atan2 max max
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
atan2 max -max
atan2 -max max
atan2 -max -max
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
atan2 max min
atan2 -max -min
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
atan2 -max min
atan2 max -min
atan2 max min_subnorm
atan2 -max -min_subnorm
atan2 -max min_subnorm
atan2 max -min_subnorm
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
atan2 0.75 1
atan2 -0.75 1.0
atan2 0.75 -1.0
atan2 -0.75 -1.0
atan2 0.390625 .00029
atan2 1.390625 0.9296875
atan2 -0.00756827042671106339 -.001792735857538728036
atan2 0x1.00000000000001p0 0x1.00000000000001p0
Fix dbl-64 atan2 in non-default rounding modes (bug 18210, bug 18211). The dbl-64 implementation of atan2 does computations that expect to run in round-to-nearest mode, and in other modes the errors can accumulate to more than the maximum accepted 9ulp. This patch makes it use FE_TONEAREST internally, similar to other functions with such issues. Tests that previously produced large errors are added for atan2 and the closely related carg, clog and clog10 functions. Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18210] [BZ #18211] * sysdeps/ieee754/dbl-64/e_atan2.c: Include <fenv.h>. (__ieee754_atan2): Set FE_TONEAREST mode for internal computations. * math/auto-libm-test-in: Add more tests of atan2, carg, clog and clog10. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-08 17:32:17 +00:00
atan2 0x4.c3841p-4 0x2.f2f308p+0
atan2 -0xe.cf143p-40 0xd.3de7ap-36
atan2 0x5.576cf8p-4 0x2.21e65p+0
atan2 -0x4.29411p-4 0x1.f4755cp+0
atan2 -0xa.b4101p+20 -0xf.9c4c8p-4
atan2 0x4.251bb8p-4 0x7.40ac68p+0
atan2 0x1.47239ep+68 0xa.3ac3cp+68
atan2 -0x6.b0794p-4 0x3.8ff10cp+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
atan2 -0x7.15e7af0a1780cp-724 0xf.fffffp+124
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
atan2 0x3.f16f1p+0 -0x1.546056p+0
atan2 -0x1.9e657cp-24 0x7.40bb4p-52
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
atan2 -0x1.f9cf48p+49 0x1.f60598p+51
math: Add more inputs to atan2 accuracy tests [BZ #28765] This patch adds following inputs: 0x1.bcab29da0e947p-54 0x1.bc41f4d2294b8p-54 0x1.a11891ec004d4p-348 0x1.814830510be26p-348 0x1.b836ed678be29p-588 0x1.b7be6f5a03a8cp-588 0x1.a83f842ef3f73p-633 0x1.a799d8a6677ep-633 to atan2 tests and updates x86_64 double atan2 ulps. This fixes BZ #28765. Reviewed-By: Paul Zimmermann <Paul.Zimmermann@inria.fr>
2022-01-12 20:26:35 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
atan2 0x1.bcab29da0e947p-54 0x1.bc41f4d2294b8p-54
# the next value generates larger error bounds on x86_64 (binary64)
atan2 0x1.a11891ec004d4p-348 0x1.814830510be26p-348
# the next value generates larger error bounds on x86_64 (binary64)
atan2 0x1.b836ed678be29p-588 0x1.b7be6f5a03a8cp-588
# the next value generates larger error bounds on x86_64 (binary64)
atan2 0x1.a83f842ef3f73p-633 0x1.a799d8a6677ep-633
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary128)
atan2 0x1.41df5aa214612c7e019fa6ade88p-13316 0x5.e53b26a270a29eb9f77ef8ef7af8p-13316
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
atan2 min min
atan2 min -min
atan2 -min min
atan2 -min -min
atan2 min_subnorm min_subnorm
atan2 min_subnorm -min_subnorm
atan2 -min_subnorm min_subnorm
atan2 -min_subnorm -min_subnorm
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
atan2 min min_subnorm
atan2 min -min_subnorm
atan2 -min min_subnorm
atan2 -min -min_subnorm
atan2 min_subnorm min
atan2 min_subnorm -min
atan2 -min_subnorm min
atan2 -min_subnorm -min
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
atan2 1 -max
atan2 -1 -max
atan2 min -max
atan2 -min -max
atan2 min_subnorm -max
atan2 -min_subnorm -max
Fix atan / atan2 missing underflows (bug 15319). This patch fixes bug 15319, missing underflows from atan / atan2 when the result of atan is very close to its small argument (or that of atan2 is very close to the ratio of its arguments, which may be an exact division). The usual approach of doing an underflowing computation if the computed result is subnormal is followed. For 32-bit x86, there are extra complications: the inline __ieee754_atan2 in bits/mathinline.h needs to be disabled for float and double because other libm functions using it generally rely on getting proper underflow exceptions from it, while the out-of-line functions have to remove excess range and precision from the underflowing result so as to return an exact 0 in the case where errno should be set for underflow to 0. (The failures I saw without that are similar to those Carlos reported for other functions, where I haven't seen a response to <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html> confirming if my diagnosis is correct. Arguably all libm functions with float and double returns should remove excess range and precision, but that's a separate matter.) The x86_64 long double case reported in a comment in bug 15319 is not a bug (it's an argument of LDBL_MIN, and x86_64 is an after-rounding architecture so the correct IEEE result is not to raise underflow in the given rounding mode, in addition to treating the result as an exact LDBL_MIN being within the newly clarified documentation of accuracy goals). I'm presuming that the fpatan instruction can be trusted to raise appropriate exceptions when the (long double) result underflows (after rounding) and so no changes are needed for x86 / x86_64 long double functions here; empirically this is the case for the cases covered in the testsuite, on my system. Tested for x86_64, x86, powerpc and mips64. Only 32-bit x86 needs ulps updates (for the changes to inlines meaning some functions no longer get excess precision from their __ieee754_atan2* calls). [BZ #15319] * sysdeps/i386/fpu/e_atan2.S (dbl_min): New object. (MO): New macro. (__ieee754_atan2): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/e_atan2f.S (flt_min): New object. (MO): New macro. (__ieee754_atan2f): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/s_atan.S (dbl_min): New object. (MO): New macro. (__atan): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/s_atanf.S (flt_min): New object. (MO): New macro. (__atanf): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/ieee754/dbl-64/e_atan2.c: Include <float.h> and <math.h>. (__ieee754_atan2): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/s_atan.c: Include <float.h> and <math_private.h>. (atan): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/s_atanf.c: Include <float.h>. (__atanf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/s_atanl.c: Include <float.h> and <math.h>. (__atanl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128ibm/s_atanl.c: Include <float.h>. (__atanl): Force underflow exception for results with small absolute value. * sysdeps/x86/fpu/bits/mathinline.h [!__SSE2_MATH__ && !__x86_64__ && __LIBC_INTERNAL_MATH_INLINES] (__ieee754_atan2): Only define inline for long double. * sysdeps/x86_64/fpu/multiarch/e_atan2.c [HAVE_FMA4_SUPPORT || HAVE_AVX_SUPPORT]: Include <math.h>. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 15319. Add more tests of atan2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (casin_test_data): Do not mark underflow exceptions as possibly missing for bug 15319. (casinh_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-02-18 21:10:49 +00:00
atan2 1 max
atan2 -1 max
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
atan2 min max
atan2 -min max
atan2 min_subnorm max
atan2 -min_subnorm max
Fix atan / atan2 missing underflows (bug 15319). This patch fixes bug 15319, missing underflows from atan / atan2 when the result of atan is very close to its small argument (or that of atan2 is very close to the ratio of its arguments, which may be an exact division). The usual approach of doing an underflowing computation if the computed result is subnormal is followed. For 32-bit x86, there are extra complications: the inline __ieee754_atan2 in bits/mathinline.h needs to be disabled for float and double because other libm functions using it generally rely on getting proper underflow exceptions from it, while the out-of-line functions have to remove excess range and precision from the underflowing result so as to return an exact 0 in the case where errno should be set for underflow to 0. (The failures I saw without that are similar to those Carlos reported for other functions, where I haven't seen a response to <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html> confirming if my diagnosis is correct. Arguably all libm functions with float and double returns should remove excess range and precision, but that's a separate matter.) The x86_64 long double case reported in a comment in bug 15319 is not a bug (it's an argument of LDBL_MIN, and x86_64 is an after-rounding architecture so the correct IEEE result is not to raise underflow in the given rounding mode, in addition to treating the result as an exact LDBL_MIN being within the newly clarified documentation of accuracy goals). I'm presuming that the fpatan instruction can be trusted to raise appropriate exceptions when the (long double) result underflows (after rounding) and so no changes are needed for x86 / x86_64 long double functions here; empirically this is the case for the cases covered in the testsuite, on my system. Tested for x86_64, x86, powerpc and mips64. Only 32-bit x86 needs ulps updates (for the changes to inlines meaning some functions no longer get excess precision from their __ieee754_atan2* calls). [BZ #15319] * sysdeps/i386/fpu/e_atan2.S (dbl_min): New object. (MO): New macro. (__ieee754_atan2): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/e_atan2f.S (flt_min): New object. (MO): New macro. (__ieee754_atan2f): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/s_atan.S (dbl_min): New object. (MO): New macro. (__atan): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/s_atanf.S (flt_min): New object. (MO): New macro. (__atanf): For results with small absolute value, force underflow exception and remove excess range and precision from return value. * sysdeps/ieee754/dbl-64/e_atan2.c: Include <float.h> and <math.h>. (__ieee754_atan2): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/s_atan.c: Include <float.h> and <math_private.h>. (atan): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/s_atanf.c: Include <float.h>. (__atanf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/s_atanl.c: Include <float.h> and <math.h>. (__atanl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128ibm/s_atanl.c: Include <float.h>. (__atanl): Force underflow exception for results with small absolute value. * sysdeps/x86/fpu/bits/mathinline.h [!__SSE2_MATH__ && !__x86_64__ && __LIBC_INTERNAL_MATH_INLINES] (__ieee754_atan2): Only define inline for long double. * sysdeps/x86_64/fpu/multiarch/e_atan2.c [HAVE_FMA4_SUPPORT || HAVE_AVX_SUPPORT]: Include <math.h>. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 15319. Add more tests of atan2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (casin_test_data): Do not mark underflow exceptions as possibly missing for bug 15319. (casinh_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-02-18 21:10:49 +00:00
atan2 min 1
atan2 -min 1
atan2 min_subnorm 1
atan2 -min_subnorm 1
atan2 min -1
atan2 -min -1
atan2 min_subnorm -1
atan2 -min_subnorm -1
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
atanh 0
atanh -0
atanh 0.75
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
atanh -0.75
atanh 0.25
atanh 0x1p-5
atanh 0x1p-10
atanh 0x1.2345p-20
Fix ldbl-96, ldbl-128ibm atanhl inaccuracy (bug 18046, bug 18047). The threshold in ldbl-96 atanhl for when to return the argument, 0x1p-28, is a bit too big, and that in ldbl-128ibm atanhl is much too big (the relevant condition being x^3/3 being < 0.5ulp of x), resulting in errors a bit above the limits of those considered acceptable in glibc in the ldbl-96 case, and in large errors in the ldbl-128ibm case. This patch changes those implementations to use more appropriate thresholds and adds tests around the thresholds for various formats. Tested for x86_64, x86 and powerpc. x86_64 and x86 ulps updated accordingly. [BZ #18046] [BZ #18047] * sysdeps/ieee754/ldbl-128ibm/e_atanhl.c (__ieee754_atanhl): Use 0x1p-56L as threshold for just returning the argument. * sysdeps/ieee754/ldbl-96/e_atanhl.c (__ieee754_atanhl): Use 0x1p-32L as threshold for just returning the argument. * math/auto-libm-test-in: Add more tests of atanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulp: Likewise.
2015-02-27 17:48:37 +00:00
atanh 0x1p-8
atanh 0x1p-9
atanh 0x1p-10
atanh 0x1p-11
atanh 0x1p-12
atanh 0x1p-13
atanh 0x1p-24
atanh 0x1p-25
atanh 0x1p-26
atanh 0x1p-27
atanh 0x1p-28
atanh 0x1p-29
atanh 0x1p-30
atanh 0x1p-31
atanh 0x1p-32
atanh 0x1p-33
atanh 0x1p-48
atanh 0x1p-49
atanh 0x1p-50
atanh 0x1p-51
atanh 0x1p-52
atanh 0x1p-53
atanh 0x1p-54
atanh 0x1p-55
atanh 0x1p-56
atanh 0x1p-57
atanh 0x1p-58
atanh 0x1p-59
atanh 0x1p-100
Fix i386 atanhl spurious underflows (bug 18049). The i386 implementation of atanhl, for small arguments, does a calculation that involves computing twice the square of the argument, resulting in spurious underflows for some arguments. This patch fixes this by just returning the argument when its exponent is below -32, with underflow being forced as needed for subnormal arguments. Tested for x86 and x86_64. [BZ #18049] * sysdeps/i386/fpu/e_atanhl.S (__ieee754_atanhl): For exponents below -32, return the argument, with underflow if subnormal. * math/auto-libm-test-in: Add more tests of atanh. * math/auto-libm-test-out: Regenerated.
2015-05-19 23:05:22 +00:00
atanh -0x1p-100
atanh 0x1p-600
atanh -0x1p-600
atanh 0x1p-10000
atanh -0x1p-10000
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
atanh 0x0.ffffffp0
atanh -0x0.ffffffp0
atanh 0x0.ffffffff8p0
atanh -0x0.ffffffff8p0
atanh 0x0.ffffffffffffp0
atanh -0x0.ffffffffffffp0
atanh 0x0.fffffffffffff8p0
atanh -0x0.fffffffffffff8p0
atanh 0x0.ffffffffffffffffp0
atanh -0x0.ffffffffffffffffp0
atanh 0x0.ffffffffffffffffffffffffffcp0
atanh -0x0.ffffffffffffffffffffffffffcp0
atanh 0x0.ffffffffffffffffffffffffffff8p0
atanh -0x0.ffffffffffffffffffffffffffff8p0
Add more tests of acosh, asinh and atanh. This patch adds some randomly-generated tests of acosh, asinh and atanh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, asinh and atanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-25 22:21:20 +00:00
atanh -0x6.e6c77p-20
atanh 0x3.2ca824p-4
atanh -0x1.cc1d66p-4
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
atanh -0xf.cd3809ca8fd28p-4
Add more tests of acosh, asinh and atanh. This patch adds some randomly-generated tests of acosh, asinh and atanh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, asinh and atanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-25 22:21:20 +00:00
atanh -0x1.04f386p-4
atanh -0x2.084568p-4
atanh -0x3.e0a5d8p-4
atanh 0x3.dfb1f5db0ceccp-4
atanh 0x2.251b2a64c85dep-4
Add more tests of atanh. This patch adds some randomly-generated tests of atanh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-08 21:13:35 +00:00
atanh -0x2.e3458cp-4
atanh 0x3.91d9f3c80c72d7acp-4
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
atanh -0x2.6c52c26567198p-4
Add more tests of acosh, atanh, cos, csqrt, erfc, sin, sincos. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, cos, csqrt, erfc, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-06 17:30:18 +00:00
atanh 0x3.a274ecp-4
atanh -0x3.f0f519a687b64p-8
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
atanh 0x6.fd4ec8p-4
atanh -0x2.6cb2a8p-4
atanh -0xc.21df7c7f51508p-4
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
atanh 0x5.8be99p-40
atanh 0x3.cbed35fe733d8p-4
Add more randomly-generated libm tests. This patch adds more libm test inputs found through random test generation to increase observed ulps on x86_64. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, atanh, cbrt, cosh, csqrt, erfc, expm1 and lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-11 15:03:10 +00:00
atanh -0x5.c18b6p-4
atanh -0x7.c88a5p-8
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
atanh -0x2.c72b7cp-4
atanh -0x3.98eaf4p-4
atanh 0x2.c1085p-4
Fix atanhl missing underflows (bug 16352). Similar to various other bugs in this area, some atanh implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. (No change in this regard is needed for the i386 implementation; special handling to force underflows in these cases will only be needed there when the spurious underflows, bug 18049, get fixed.) Tested for x86_64, x86, powerpc and mips64. [BZ #16352] * sysdeps/i386/fpu/e_atanh.S (dbl_min): New object. (__ieee754_atanh): Force underflow exception for results with small absolute value. * sysdeps/i386/fpu/e_atanhf.S (flt_min): New object. (__ieee754_atanhf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/e_atanh.c: Include <float.h>. (__ieee754_atanh): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/e_atanhf.c: Include <float.h>. (__ieee754_atanhf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/e_atanhl.c: Include <float.h>. (__ieee754_atanhl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128ibm/e_atanhl.c: Include <float.h>. (__ieee754_atanhl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-96/e_atanhl.c: Include <float.h>. (__ieee754_atanhl): Force underflow exception for results with small absolute value. * math/auto-libm-test-in: Do not allow missing underflow exceptions from atanh. * math/auto-libm-test-out: Regenerated.
2015-05-15 22:07:57 +00:00
atanh 0x1p-500
atanh 0x1p-5000
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary128)
atanh 0x2.c02a24f3472c7840afbd8cfb68bap-4
Fix atanhl missing underflows (bug 16352). Similar to various other bugs in this area, some atanh implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. (No change in this regard is needed for the i386 implementation; special handling to force underflows in these cases will only be needed there when the spurious underflows, bug 18049, get fixed.) Tested for x86_64, x86, powerpc and mips64. [BZ #16352] * sysdeps/i386/fpu/e_atanh.S (dbl_min): New object. (__ieee754_atanh): Force underflow exception for results with small absolute value. * sysdeps/i386/fpu/e_atanhf.S (flt_min): New object. (__ieee754_atanhf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/e_atanh.c: Include <float.h>. (__ieee754_atanh): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/e_atanhf.c: Include <float.h>. (__ieee754_atanhf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/e_atanhl.c: Include <float.h>. (__ieee754_atanhl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128ibm/e_atanhl.c: Include <float.h>. (__ieee754_atanhl): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-96/e_atanhl.c: Include <float.h>. (__ieee754_atanhl): Force underflow exception for results with small absolute value. * math/auto-libm-test-in: Do not allow missing underflow exceptions from atanh. * math/auto-libm-test-out: Regenerated.
2015-05-15 22:07:57 +00:00
atanh min
atanh -min
atanh min_subnorm
atanh -min_subnorm
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
Move tests of cabs and carg from libm-test.inc to auto-libm-test-in. This patch moves tests of cabs and carg to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86; no ulps updates needed. * math/auto-libm-test-in: Add tests of cabs and carg. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cabs_test_data): Use AUTO_TESTS_c_f. (carg_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_f. (func_calc_desc): Add mpc_c_f union field. (test_functions): Add cabs and carg. (calc_generic_results): Handle mpc_c_f.
2013-12-19 21:28:30 +00:00
# cabs (x,y) == cabs (y,x).
cabs 0.75 12.390625
# cabs (x,y) == cabs (-x,y).
cabs -12.390625 0.75
# cabs (x,y) == cabs (-y,x).
cabs -0.75 12.390625
# cabs (x,y) == cabs (-x,-y).
cabs -12.390625 -0.75
# cabs (x,y) == cabs (-y,-x).
cabs -0.75 -12.390625
# cabs (x,0) == fabs (x).
cabs -0.75 0
cabs 0.75 0
cabs -1.0 0
cabs 1.0 0
cabs -5.7e7 0
cabs 5.7e7 0
cabs 0.75 1.25
Add more tests of cabs. This patch adds some randomly-generated tests of cabs that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cabs. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-08 17:46:07 +00:00
cabs -0x1.34be3p-4 -0xc.56623p+0
cabs -0x1.2b0ff8p+28 -0x2.549fc4p+16
cabs -0x1.0932cp-80 -0x2.51109p-24
cabs -0x1.055fb2p+48 0x9.1ce86p+24
cabs -0x1.26a566p+120 0x4.017b28p+92
cabs -0x1.0eda54p+28 0xb.09476p+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
cabs -0x1.133b84p+84 -0xa.7d925f57f60cp+768
cabs -0 -0x3.4e5d7877324cp+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
cabs -0xa.f59b8p+4 0xa.21a95p+20
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
cabs -0x1.30ed4cp+0 0x1.e9d956p+56
cabs -0x1.250366p-36 -0x5.a5046p-4
cabs -0x1.88858cp+84 0x5.bd9198p+36
Move tests of cabs and carg from libm-test.inc to auto-libm-test-in. This patch moves tests of cabs and carg to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86; no ulps updates needed. * math/auto-libm-test-in: Add tests of cabs and carg. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cabs_test_data): Use AUTO_TESTS_c_f. (carg_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_f. (func_calc_desc): Add mpc_c_f union field. (test_functions): Add cabs and carg. (calc_generic_results): Handle mpc_c_f.
2013-12-19 21:28:30 +00:00
Move tests of cacos, cacosh to auto-libm-test-*. This patch moves tests of cacos and cacosh with finite inputs to using the auto-libm-test machinery. Each of auto-libm-test-out-cacos and auto-libm-test-out-cacosh takes about 80 minutes to generate on my system because of MPC slowness on special cases that appear in the tests (with MPC 1.0.3; I don't know to what extent current MPC master might speed it up). Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of cacos and cacosh. * math/auto-libm-test-out-cacos: New generated file. * math/auto-libm-test-out-cacosh: Likewise. * math/libm-test-cacos.inc (cacos_test_data): Use AUTO_TESTS_c_c. Move tests with finite inputs to auto-libm-test-in. * math/libm-test-cacosh.inc (cacosh_test_data): Likewise. * math/Makefile (libm-test-funcs-auto): Add cacos and cacosh. (libm-test-funcs-noauto): Remove cacos and cacosh. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2017-02-17 17:44:23 +00:00
cacos 0 0
cacos -0 0
cacos -0 -0
cacos 0 -0
cacos 0 -1.5
cacos -0 -1.5
cacos 0 -1.0
cacos -0 -1.0
cacos 0 -0.5
cacos -0 -0.5
cacos 0 0.5
cacos -0 0.5
cacos 0 1.0
cacos -0 1.0
cacos 0 1.5
cacos -0 1.5
cacos -1.5 0
cacos -1.5 -0
cacos -1.0 0
cacos -1.0 -0
cacos -0.5 0
cacos -0.5 -0
cacos 0.5 0
cacos 0.5 -0
cacos 1.0 0
cacos 1.0 -0
cacos 1.5 0
cacos 1.5 -0
cacos 0x1p50 1.0
cacos 0x1p50 -1.0
cacos -0x1p50 1.0
cacos -0x1p50 -1.0
cacos 1.0 0x1p50
cacos -1.0 0x1p50
cacos 1.0 -0x1p50
cacos -1.0 -0x1p50
cacos 0x1p500 1.0
cacos 0x1p500 -1.0
cacos -0x1p500 1.0
cacos -0x1p500 -1.0
cacos 1.0 0x1p500
cacos -1.0 0x1p500
cacos 1.0 -0x1p500
cacos -1.0 -0x1p500
cacos 0x1p5000 1.0
cacos 0x1p5000 -1.0
cacos -0x1p5000 1.0
cacos -0x1p5000 -1.0
cacos 1.0 0x1p5000
cacos -1.0 0x1p5000
cacos 1.0 -0x1p5000
cacos -1.0 -0x1p5000
cacos 0x1.fp127 0x1.fp127
cacos 0x1.fp1023 0x1.fp1023
cacos 0x1.fp16383 0x1.fp16383
cacos 0x1.fp-129 1.5
cacos 0x1.fp-129 -1.5
cacos -0x1.fp-129 1.5
cacos -0x1.fp-129 -1.5
cacos 1.5 0x1.fp-129
cacos -1.5 0x1.fp-129
cacos 1.5 -0x1.fp-129
cacos -1.5 -0x1.fp-129
cacos 0x1.fp-1025 1.5
cacos 0x1.fp-1025 -1.5
cacos -0x1.fp-1025 1.5
cacos -0x1.fp-1025 -1.5
cacos 1.5 0x1.fp-1025
cacos -1.5 0x1.fp-1025
cacos 1.5 -0x1.fp-1025
cacos -1.5 -0x1.fp-1025
cacos 0x1.fp-16385 1.5
cacos 0x1.fp-16385 -1.5
cacos -0x1.fp-16385 1.5
cacos -0x1.fp-16385 -1.5
cacos 1.5 0x1.fp-16385
cacos -1.5 0x1.fp-16385
cacos 1.5 -0x1.fp-16385
cacos -1.5 -0x1.fp-16385
cacos 0.5 1.0
cacos 0.5 -1.0
cacos -0.5 1.0
cacos -0.5 -1.0
cacos 1.0 0.5
cacos -1.0 0.5
cacos 1.0 -0.5
cacos -1.0 -0.5
cacos 0.25 1.0
cacos 0.25 -1.0
cacos -0.25 1.0
cacos -0.25 -1.0
cacos 1.0 0.25
cacos -1.0 0.25
cacos 1.0 -0.25
cacos -1.0 -0.25
cacos 0x1.fp-10 1.0
cacos 0x1.fp-10 -1.0
cacos -0x1.fp-10 1.0
cacos -0x1.fp-10 -1.0
cacos 1.0 0x1.fp-10
cacos -1.0 0x1.fp-10
cacos 1.0 -0x1.fp-10
cacos -1.0 -0x1.fp-10
cacos 0x1.fp-30 1.0
cacos 0x1.fp-30 -1.0
cacos -0x1.fp-30 1.0
cacos -0x1.fp-30 -1.0
cacos 1.0 0x1.fp-30
cacos -1.0 0x1.fp-30
cacos 1.0 -0x1.fp-30
cacos -1.0 -0x1.fp-30
cacos 0x1.fp-100 1.0
cacos 0x1.fp-100 -1.0
cacos -0x1.fp-100 1.0
cacos -0x1.fp-100 -1.0
cacos 1.0 0x1.fp-100
cacos -1.0 0x1.fp-100
cacos 1.0 -0x1.fp-100
cacos -1.0 -0x1.fp-100
cacos 0x1.fp-129 1.0
cacos 0x1.fp-129 -1.0
cacos -0x1.fp-129 1.0
cacos -0x1.fp-129 -1.0
cacos 1.0 0x1.fp-129
cacos -1.0 0x1.fp-129
cacos 1.0 -0x1.fp-129
cacos -1.0 -0x1.fp-129
cacos 0x1.fp-1000 1.0
cacos 0x1.fp-1000 -1.0
cacos -0x1.fp-1000 1.0
cacos -0x1.fp-1000 -1.0
cacos 1.0 0x1.fp-1000
cacos -1.0 0x1.fp-1000
cacos 1.0 -0x1.fp-1000
cacos -1.0 -0x1.fp-1000
cacos 0x1.fp-1025 1.0
cacos 0x1.fp-1025 -1.0
cacos -0x1.fp-1025 1.0
cacos -0x1.fp-1025 -1.0
cacos 1.0 0x1.fp-1025
cacos -1.0 0x1.fp-1025
cacos 1.0 -0x1.fp-1025
cacos -1.0 -0x1.fp-1025
cacos 0x1.fp-10000 1.0
cacos 0x1.fp-10000 -1.0
cacos -0x1.fp-10000 1.0
cacos -0x1.fp-10000 -1.0
cacos 1.0 0x1.fp-10000
cacos -1.0 0x1.fp-10000
cacos 1.0 -0x1.fp-10000
cacos -1.0 -0x1.fp-10000
cacos 0x1.fp-16385 1.0
cacos 0x1.fp-16385 -1.0
cacos -0x1.fp-16385 1.0
cacos -0x1.fp-16385 -1.0
cacos 1.0 0x1.fp-16385
cacos -1.0 0x1.fp-16385
cacos 1.0 -0x1.fp-16385
cacos -1.0 -0x1.fp-16385
cacos 0x1p-23 0x1.000002p0
cacos 0x1p-23 -0x1.000002p0
cacos -0x1p-23 0x1.000002p0
cacos -0x1p-23 -0x1.000002p0
cacos 0x1.000002p0 0x1p-23
cacos -0x1.000002p0 0x1p-23
cacos 0x1.000002p0 -0x1p-23
cacos -0x1.000002p0 -0x1p-23
cacos 0x1.fp-129 0x1.000002p0
cacos 0x1.fp-129 -0x1.000002p0
cacos -0x1.fp-129 0x1.000002p0
cacos -0x1.fp-129 -0x1.000002p0
cacos 0x1.000002p0 0x1.fp-129
cacos -0x1.000002p0 0x1.fp-129
cacos 0x1.000002p0 -0x1.fp-129
cacos -0x1.000002p0 -0x1.fp-129
cacos 0.0 0x1.000002p0
cacos 0.0 -0x1.000002p0
cacos -0.0 0x1.000002p0
cacos -0.0 -0x1.000002p0
cacos 0x1.000002p0 0.0
cacos -0x1.000002p0 0.0
cacos 0x1.000002p0 -0.0
cacos -0x1.000002p0 -0.0
cacos 0x1p-52 0x1.0000000000001p0
cacos 0x1p-52 -0x1.0000000000001p0
cacos -0x1p-52 0x1.0000000000001p0
cacos -0x1p-52 -0x1.0000000000001p0
cacos 0x1.0000000000001p0 0x1p-52
cacos -0x1.0000000000001p0 0x1p-52
cacos 0x1.0000000000001p0 -0x1p-52
cacos -0x1.0000000000001p0 -0x1p-52
cacos 0x1.fp-1025 0x1.0000000000001p0
cacos 0x1.fp-1025 -0x1.0000000000001p0
cacos -0x1.fp-1025 0x1.0000000000001p0
cacos -0x1.fp-1025 -0x1.0000000000001p0
cacos 0x1.0000000000001p0 0x1.fp-1025
cacos -0x1.0000000000001p0 0x1.fp-1025
cacos 0x1.0000000000001p0 -0x1.fp-1025
cacos -0x1.0000000000001p0 -0x1.fp-1025
cacos 0.0 0x1.0000000000001p0
cacos 0.0 -0x1.0000000000001p0
cacos -0.0 0x1.0000000000001p0
cacos -0.0 -0x1.0000000000001p0
cacos 0x1.0000000000001p0 0.0
cacos -0x1.0000000000001p0 0.0
cacos 0x1.0000000000001p0 -0.0
cacos -0x1.0000000000001p0 -0.0
cacos 0x1p-63 0x1.0000000000000002p0
cacos 0x1p-63 -0x1.0000000000000002p0
cacos -0x1p-63 0x1.0000000000000002p0
cacos -0x1p-63 -0x1.0000000000000002p0
cacos 0x1.0000000000000002p0 0x1p-63
cacos -0x1.0000000000000002p0 0x1p-63
cacos 0x1.0000000000000002p0 -0x1p-63
cacos -0x1.0000000000000002p0 -0x1p-63
cacos 0x1.fp-16385 0x1.0000000000000002p0
cacos 0x1.fp-16385 -0x1.0000000000000002p0
cacos -0x1.fp-16385 0x1.0000000000000002p0
cacos -0x1.fp-16385 -0x1.0000000000000002p0
cacos 0x1.0000000000000002p0 0x1.fp-16385
cacos -0x1.0000000000000002p0 0x1.fp-16385
cacos 0x1.0000000000000002p0 -0x1.fp-16385
cacos -0x1.0000000000000002p0 -0x1.fp-16385
cacos 0.0 0x1.0000000000000002p0
cacos 0.0 -0x1.0000000000000002p0
cacos -0.0 0x1.0000000000000002p0
cacos -0.0 -0x1.0000000000000002p0
cacos 0x1.0000000000000002p0 0.0
cacos -0x1.0000000000000002p0 0.0
cacos 0x1.0000000000000002p0 -0.0
cacos -0x1.0000000000000002p0 -0.0
cacos 0x1p-106 0x1.000000000000000000000000008p0
cacos 0x1p-106 -0x1.000000000000000000000000008p0
cacos -0x1p-106 0x1.000000000000000000000000008p0
cacos -0x1p-106 -0x1.000000000000000000000000008p0
cacos 0x1.000000000000000000000000008p0 0x1p-106
cacos -0x1.000000000000000000000000008p0 0x1p-106
cacos 0x1.000000000000000000000000008p0 -0x1p-106
cacos -0x1.000000000000000000000000008p0 -0x1p-106
cacos 0x1.fp-1025 0x1.000000000000000000000000008p0
cacos 0x1.fp-1025 -0x1.000000000000000000000000008p0
cacos -0x1.fp-1025 0x1.000000000000000000000000008p0
cacos -0x1.fp-1025 -0x1.000000000000000000000000008p0
cacos 0x1.000000000000000000000000008p0 0x1.fp-1025
cacos -0x1.000000000000000000000000008p0 0x1.fp-1025
cacos 0x1.000000000000000000000000008p0 -0x1.fp-1025
cacos -0x1.000000000000000000000000008p0 -0x1.fp-1025
cacos 0.0 0x1.000000000000000000000000008p0
cacos 0.0 -0x1.000000000000000000000000008p0
cacos -0.0 0x1.000000000000000000000000008p0
cacos -0.0 -0x1.000000000000000000000000008p0
cacos 0x1.000000000000000000000000008p0 0.0
cacos -0x1.000000000000000000000000008p0 0.0
cacos 0x1.000000000000000000000000008p0 -0.0
cacos -0x1.000000000000000000000000008p0 -0.0
cacos 0x1p-113 0x1.0000000000000000000000000001p0
cacos 0x1p-113 -0x1.0000000000000000000000000001p0
cacos -0x1p-113 0x1.0000000000000000000000000001p0
cacos -0x1p-113 -0x1.0000000000000000000000000001p0
cacos 0x1.0000000000000000000000000001p0 0x1p-113
cacos -0x1.0000000000000000000000000001p0 0x1p-113
cacos 0x1.0000000000000000000000000001p0 -0x1p-113
cacos -0x1.0000000000000000000000000001p0 -0x1p-113
cacos 0x1.fp-16385 0x1.0000000000000000000000000001p0
cacos 0x1.fp-16385 -0x1.0000000000000000000000000001p0
cacos -0x1.fp-16385 0x1.0000000000000000000000000001p0
cacos -0x1.fp-16385 -0x1.0000000000000000000000000001p0
cacos 0x1.0000000000000000000000000001p0 0x1.fp-16385
cacos -0x1.0000000000000000000000000001p0 0x1.fp-16385
cacos 0x1.0000000000000000000000000001p0 -0x1.fp-16385
cacos -0x1.0000000000000000000000000001p0 -0x1.fp-16385
cacos 0.0 0x1.0000000000000000000000000001p0
cacos 0.0 -0x1.0000000000000000000000000001p0
cacos -0.0 0x1.0000000000000000000000000001p0
cacos -0.0 -0x1.0000000000000000000000000001p0
cacos 0x1.0000000000000000000000000001p0 0.0
cacos -0x1.0000000000000000000000000001p0 0.0
cacos 0x1.0000000000000000000000000001p0 -0.0
cacos -0x1.0000000000000000000000000001p0 -0.0
cacos 0x1p-23 0x0.ffffffp0
cacos 0x1p-23 -0x0.ffffffp0
cacos -0x1p-23 0x0.ffffffp0
cacos -0x1p-23 -0x0.ffffffp0
cacos 0x0.ffffffp0 0x1p-23
cacos -0x0.ffffffp0 0x1p-23
cacos 0x0.ffffffp0 -0x1p-23
cacos -0x0.ffffffp0 -0x1p-23
cacos 0x1.fp-129 0x0.ffffffp0
cacos 0x1.fp-129 -0x0.ffffffp0
cacos -0x1.fp-129 0x0.ffffffp0
cacos -0x1.fp-129 -0x0.ffffffp0
cacos 0x0.ffffffp0 0x1.fp-129
cacos -0x0.ffffffp0 0x1.fp-129
cacos 0x0.ffffffp0 -0x1.fp-129
cacos -0x0.ffffffp0 -0x1.fp-129
cacos 0.0 0x0.ffffffp0
cacos 0.0 -0x0.ffffffp0
cacos -0.0 0x0.ffffffp0
cacos -0.0 -0x0.ffffffp0
cacos 0x0.ffffffp0 0.0
cacos -0x0.ffffffp0 0.0
cacos 0x0.ffffffp0 -0.0
cacos -0x0.ffffffp0 -0.0
cacos 0x1p-23 0.5
cacos 0x1p-23 -0.5
cacos -0x1p-23 0.5
cacos -0x1p-23 -0.5
cacos 0.5 0x1p-23
cacos -0.5 0x1p-23
cacos 0.5 -0x1p-23
cacos -0.5 -0x1p-23
cacos 0x1.fp-129 0.5
cacos 0x1.fp-129 -0.5
cacos -0x1.fp-129 0.5
cacos -0x1.fp-129 -0.5
cacos 0.5 0x1.fp-129
cacos -0.5 0x1.fp-129
cacos 0.5 -0x1.fp-129
cacos -0.5 -0x1.fp-129
cacos 0x1p-23 0x1p-23
cacos 0x1p-23 -0x1p-23
cacos -0x1p-23 0x1p-23
cacos -0x1p-23 -0x1p-23
cacos 0x1.fp-129 0x1p-23
cacos 0x1.fp-129 -0x1p-23
cacos -0x1.fp-129 0x1p-23
cacos -0x1.fp-129 -0x1p-23
cacos 0x1p-23 0x1.fp-129
cacos -0x1p-23 0x1.fp-129
cacos 0x1p-23 -0x1.fp-129
cacos -0x1p-23 -0x1.fp-129
cacos 0.0 0x1p-23
cacos 0.0 -0x1p-23
cacos -0.0 0x1p-23
cacos -0.0 -0x1p-23
cacos 0x1p-23 0.0
cacos -0x1p-23 0.0
cacos 0x1p-23 -0.0
cacos -0x1p-23 -0.0
cacos 0x1.fp-129 0x1.fp-129
cacos 0x1.fp-129 -0x1.fp-129
cacos -0x1.fp-129 0x1.fp-129
cacos -0x1.fp-129 -0x1.fp-129
cacos 0x1p-52 0x0.fffffffffffff8p0
cacos 0x1p-52 -0x0.fffffffffffff8p0
cacos -0x1p-52 0x0.fffffffffffff8p0
cacos -0x1p-52 -0x0.fffffffffffff8p0
cacos 0x0.fffffffffffff8p0 0x1p-52
cacos -0x0.fffffffffffff8p0 0x1p-52
cacos 0x0.fffffffffffff8p0 -0x1p-52
cacos -0x0.fffffffffffff8p0 -0x1p-52
cacos 0x1.fp-1025 0x0.fffffffffffff8p0
cacos 0x1.fp-1025 -0x0.fffffffffffff8p0
cacos -0x1.fp-1025 0x0.fffffffffffff8p0
cacos -0x1.fp-1025 -0x0.fffffffffffff8p0
cacos 0x0.fffffffffffff8p0 0x1.fp-1025
cacos -0x0.fffffffffffff8p0 0x1.fp-1025
cacos 0x0.fffffffffffff8p0 -0x1.fp-1025
cacos -0x0.fffffffffffff8p0 -0x1.fp-1025
cacos 0.0 0x0.fffffffffffff8p0
cacos 0.0 -0x0.fffffffffffff8p0
cacos -0.0 0x0.fffffffffffff8p0
cacos -0.0 -0x0.fffffffffffff8p0
cacos 0x0.fffffffffffff8p0 0.0
cacos -0x0.fffffffffffff8p0 0.0
cacos 0x0.fffffffffffff8p0 -0.0
cacos -0x0.fffffffffffff8p0 -0.0
cacos 0x1p-52 0.5
cacos 0x1p-52 -0.5
cacos -0x1p-52 0.5
cacos -0x1p-52 -0.5
cacos 0.5 0x1p-52
cacos -0.5 0x1p-52
cacos 0.5 -0x1p-52
cacos -0.5 -0x1p-52
cacos 0x1.fp-1025 0.5
cacos 0x1.fp-1025 -0.5
cacos -0x1.fp-1025 0.5
cacos -0x1.fp-1025 -0.5
cacos 0.5 0x1.fp-1025
cacos -0.5 0x1.fp-1025
cacos 0.5 -0x1.fp-1025
cacos -0.5 -0x1.fp-1025
cacos 0x1p-52 0x1p-52
cacos 0x1p-52 -0x1p-52
cacos -0x1p-52 0x1p-52
cacos -0x1p-52 -0x1p-52
cacos 0x1.fp-1025 0x1p-52
cacos 0x1.fp-1025 -0x1p-52
cacos -0x1.fp-1025 0x1p-52
cacos -0x1.fp-1025 -0x1p-52
cacos 0x1p-52 0x1.fp-1025
cacos -0x1p-52 0x1.fp-1025
cacos 0x1p-52 -0x1.fp-1025
cacos -0x1p-52 -0x1.fp-1025
cacos 0.0 0x1p-52
cacos 0.0 -0x1p-52
cacos -0.0 0x1p-52
cacos -0.0 -0x1p-52
cacos 0x1p-52 0.0
cacos -0x1p-52 0.0
cacos 0x1p-52 -0.0
cacos -0x1p-52 -0.0
cacos 0x1.fp-1025 0x1.fp-1025
cacos 0x1.fp-1025 -0x1.fp-1025
cacos -0x1.fp-1025 0x1.fp-1025
cacos -0x1.fp-1025 -0x1.fp-1025
cacos 0x1p-63 0x0.ffffffffffffffffp0
cacos 0x1p-63 -0x0.ffffffffffffffffp0
cacos -0x1p-63 0x0.ffffffffffffffffp0
cacos -0x1p-63 -0x0.ffffffffffffffffp0
cacos 0x0.ffffffffffffffffp0 0x1p-63
cacos -0x0.ffffffffffffffffp0 0x1p-63
cacos 0x0.ffffffffffffffffp0 -0x1p-63
cacos -0x0.ffffffffffffffffp0 -0x1p-63
cacos 0x1.fp-16385 0x0.ffffffffffffffffp0
cacos 0x1.fp-16385 -0x0.ffffffffffffffffp0
cacos -0x1.fp-16385 0x0.ffffffffffffffffp0
cacos -0x1.fp-16385 -0x0.ffffffffffffffffp0
cacos 0x0.ffffffffffffffffp0 0x1.fp-16385
cacos -0x0.ffffffffffffffffp0 0x1.fp-16385
cacos 0x0.ffffffffffffffffp0 -0x1.fp-16385
cacos -0x0.ffffffffffffffffp0 -0x1.fp-16385
cacos 0.0 0x0.ffffffffffffffffp0
cacos 0.0 -0x0.ffffffffffffffffp0
cacos -0.0 0x0.ffffffffffffffffp0
cacos -0.0 -0x0.ffffffffffffffffp0
cacos 0x0.ffffffffffffffffp0 0.0
cacos -0x0.ffffffffffffffffp0 0.0
cacos 0x0.ffffffffffffffffp0 -0.0
cacos -0x0.ffffffffffffffffp0 -0.0
cacos 0x1p-63 0.5
cacos 0x1p-63 -0.5
cacos -0x1p-63 0.5
cacos -0x1p-63 -0.5
cacos 0.5 0x1p-63
cacos -0.5 0x1p-63
cacos 0.5 -0x1p-63
cacos -0.5 -0x1p-63
cacos 0x1.fp-16385 0.5
cacos 0x1.fp-16385 -0.5
cacos -0x1.fp-16385 0.5
cacos -0x1.fp-16385 -0.5
cacos 0.5 0x1.fp-16385
cacos -0.5 0x1.fp-16385
cacos 0.5 -0x1.fp-16385
cacos -0.5 -0x1.fp-16385
cacos 0x1p-63 0x1p-63
cacos 0x1p-63 -0x1p-63
cacos -0x1p-63 0x1p-63
cacos -0x1p-63 -0x1p-63
cacos 0x1.fp-16385 0x1p-63
cacos 0x1.fp-16385 -0x1p-63
cacos -0x1.fp-16385 0x1p-63
cacos -0x1.fp-16385 -0x1p-63
cacos 0x1p-63 0x1.fp-16385
cacos -0x1p-63 0x1.fp-16385
cacos 0x1p-63 -0x1.fp-16385
cacos -0x1p-63 -0x1.fp-16385
cacos 0.0 0x1p-63
cacos 0.0 -0x1p-63
cacos -0.0 0x1p-63
cacos -0.0 -0x1p-63
cacos 0x1p-63 0.0
cacos -0x1p-63 0.0
cacos 0x1p-63 -0.0
cacos -0x1p-63 -0.0
cacos 0x1.fp-16385 0x1.fp-16385
cacos 0x1.fp-16385 -0x1.fp-16385
cacos -0x1.fp-16385 0x1.fp-16385
cacos -0x1.fp-16385 -0x1.fp-16385
cacos 0x1p-105 0x0.ffffffffffffffffffffffffffcp0
cacos 0x1p-105 -0x0.ffffffffffffffffffffffffffcp0
cacos -0x1p-105 0x0.ffffffffffffffffffffffffffcp0
cacos -0x1p-105 -0x0.ffffffffffffffffffffffffffcp0
cacos 0x0.ffffffffffffffffffffffffffcp0 0x1p-105
cacos -0x0.ffffffffffffffffffffffffffcp0 0x1p-105
cacos 0x0.ffffffffffffffffffffffffffcp0 -0x1p-105
cacos -0x0.ffffffffffffffffffffffffffcp0 -0x1p-105
cacos 0x1.fp-1025 0x0.ffffffffffffffffffffffffffcp0
cacos 0x1.fp-1025 -0x0.ffffffffffffffffffffffffffcp0
cacos -0x1.fp-1025 0x0.ffffffffffffffffffffffffffcp0
cacos -0x1.fp-1025 -0x0.ffffffffffffffffffffffffffcp0
cacos 0x0.ffffffffffffffffffffffffffcp0 0x1.fp-1025
cacos -0x0.ffffffffffffffffffffffffffcp0 0x1.fp-1025
cacos 0x0.ffffffffffffffffffffffffffcp0 -0x1.fp-1025
cacos -0x0.ffffffffffffffffffffffffffcp0 -0x1.fp-1025
cacos 0.0 0x0.ffffffffffffffffffffffffffcp0
cacos 0.0 -0x0.ffffffffffffffffffffffffffcp0
cacos -0.0 0x0.ffffffffffffffffffffffffffcp0
cacos -0.0 -0x0.ffffffffffffffffffffffffffcp0
cacos 0x0.ffffffffffffffffffffffffffcp0 0.0
cacos -0x0.ffffffffffffffffffffffffffcp0 0.0
cacos 0x0.ffffffffffffffffffffffffffcp0 -0.0
cacos -0x0.ffffffffffffffffffffffffffcp0 -0.0
cacos 0x1p-105 0.5
cacos 0x1p-105 -0.5
cacos -0x1p-105 0.5
cacos -0x1p-105 -0.5
cacos 0.5 0x1p-105
cacos -0.5 0x1p-105
cacos 0.5 -0x1p-105
cacos -0.5 -0x1p-105
cacos 0x1p-105 0x1p-105
cacos 0x1p-105 -0x1p-105
cacos -0x1p-105 0x1p-105
cacos -0x1p-105 -0x1p-105
cacos 0x1.fp-1025 0x1p-105
cacos 0x1.fp-1025 -0x1p-105
cacos -0x1.fp-1025 0x1p-105
cacos -0x1.fp-1025 -0x1p-105
cacos 0x1p-105 0x1.fp-1025
cacos -0x1p-105 0x1.fp-1025
cacos 0x1p-105 -0x1.fp-1025
cacos -0x1p-105 -0x1.fp-1025
cacos 0.0 0x1p-105
cacos 0.0 -0x1p-105
cacos -0.0 0x1p-105
cacos -0.0 -0x1p-105
cacos 0x1p-105 0.0
cacos -0x1p-105 0.0
cacos 0x1p-105 -0.0
cacos -0x1p-105 -0.0
cacos 0x1p-112 0x0.ffffffffffffffffffffffffffff8p0
cacos 0x1p-112 -0x0.ffffffffffffffffffffffffffff8p0
cacos -0x1p-112 0x0.ffffffffffffffffffffffffffff8p0
cacos -0x1p-112 -0x0.ffffffffffffffffffffffffffff8p0
cacos 0x0.ffffffffffffffffffffffffffff8p0 0x1p-112
cacos -0x0.ffffffffffffffffffffffffffff8p0 0x1p-112
cacos 0x0.ffffffffffffffffffffffffffff8p0 -0x1p-112
cacos -0x0.ffffffffffffffffffffffffffff8p0 -0x1p-112
cacos 0x1.fp-16385 0x0.ffffffffffffffffffffffffffff8p0
cacos 0x1.fp-16385 -0x0.ffffffffffffffffffffffffffff8p0
cacos -0x1.fp-16385 0x0.ffffffffffffffffffffffffffff8p0
cacos -0x1.fp-16385 -0x0.ffffffffffffffffffffffffffff8p0
cacos 0x0.ffffffffffffffffffffffffffff8p0 0x1.fp-16385
cacos -0x0.ffffffffffffffffffffffffffff8p0 0x1.fp-16385
cacos 0x0.ffffffffffffffffffffffffffff8p0 -0x1.fp-16385
cacos -0x0.ffffffffffffffffffffffffffff8p0 -0x1.fp-16385
cacos 0.0 0x0.ffffffffffffffffffffffffffff8p0
cacos 0.0 -0x0.ffffffffffffffffffffffffffff8p0
cacos -0.0 0x0.ffffffffffffffffffffffffffff8p0
cacos -0.0 -0x0.ffffffffffffffffffffffffffff8p0
cacos 0x0.ffffffffffffffffffffffffffff8p0 0.0
cacos -0x0.ffffffffffffffffffffffffffff8p0 0.0
cacos 0x0.ffffffffffffffffffffffffffff8p0 -0.0
cacos -0x0.ffffffffffffffffffffffffffff8p0 -0.0
cacos 0x1p-112 0.5
cacos 0x1p-112 -0.5
cacos -0x1p-112 0.5
cacos -0x1p-112 -0.5
cacos 0.5 0x1p-112
cacos -0.5 0x1p-112
cacos 0.5 -0x1p-112
cacos -0.5 -0x1p-112
cacos 0x1p-112 0x1p-112
cacos 0x1p-112 -0x1p-112
cacos -0x1p-112 0x1p-112
cacos -0x1p-112 -0x1p-112
cacos 0x1.fp-16385 0x1p-112
cacos 0x1.fp-16385 -0x1p-112
cacos -0x1.fp-16385 0x1p-112
cacos -0x1.fp-16385 -0x1p-112
cacos 0x1p-112 0x1.fp-16385
cacos -0x1p-112 0x1.fp-16385
cacos 0x1p-112 -0x1.fp-16385
cacos -0x1p-112 -0x1.fp-16385
cacos 0.0 0x1p-112
cacos 0.0 -0x1p-112
cacos -0.0 0x1p-112
cacos -0.0 -0x1p-112
cacos 0x1p-112 0.0
cacos -0x1p-112 0.0
cacos 0x1p-112 -0.0
cacos -0x1p-112 -0.0
cacos 0.75 1.25
cacos -2 -3
cacosh 0 0
cacosh -0 0
cacosh 0 -0
cacosh -0 -0
cacosh 0 -1.5
cacosh -0 -1.5
cacosh 0 -1.0
cacosh -0 -1.0
cacosh 0 -0.5
cacosh -0 -0.5
cacosh 0 0.5
cacosh -0 0.5
cacosh 0 1.0
cacosh -0 1.0
cacosh 0 1.5
cacosh -0 1.5
cacosh -1.5 0
cacosh -1.5 -0
cacosh -1.0 0
cacosh -1.0 -0
cacosh -0.5 0
cacosh -0.5 -0
cacosh 0.5 0
cacosh 0.5 -0
cacosh 1.0 0
cacosh 1.0 -0
cacosh 1.5 0
cacosh 1.5 -0
cacosh 0x1p50 1.0
cacosh 0x1p50 -1.0
cacosh -0x1p50 1.0
cacosh -0x1p50 -1.0
cacosh 1.0 0x1p50
cacosh -1.0 0x1p50
cacosh 1.0 -0x1p50
cacosh -1.0 -0x1p50
cacosh 0x1p500 1.0
cacosh 0x1p500 -1.0
cacosh -0x1p500 1.0
cacosh -0x1p500 -1.0
cacosh 1.0 0x1p500
cacosh -1.0 0x1p500
cacosh 1.0 -0x1p500
cacosh -1.0 -0x1p500
cacosh 0x1p5000 1.0
cacosh 0x1p5000 -1.0
cacosh -0x1p5000 1.0
cacosh -0x1p5000 -1.0
cacosh 1.0 0x1p5000
cacosh -1.0 0x1p5000
cacosh 1.0 -0x1p5000
cacosh -1.0 -0x1p5000
cacosh 0x1.fp127 0x1.fp127
cacosh 0x1.fp1023 0x1.fp1023
cacosh 0x1.fp16383 0x1.fp16383
cacosh 0x1.fp-129 1.5
cacosh 0x1.fp-129 -1.5
cacosh -0x1.fp-129 1.5
cacosh -0x1.fp-129 -1.5
cacosh 1.5 0x1.fp-129
cacosh -1.5 0x1.fp-129
cacosh 1.5 -0x1.fp-129
cacosh -1.5 -0x1.fp-129
cacosh 0x1.fp-1025 1.5
cacosh 0x1.fp-1025 -1.5
cacosh -0x1.fp-1025 1.5
cacosh -0x1.fp-1025 -1.5
cacosh 1.5 0x1.fp-1025
cacosh -1.5 0x1.fp-1025
cacosh 1.5 -0x1.fp-1025
cacosh -1.5 -0x1.fp-1025
cacosh 0x1.fp-16385 1.5
cacosh 0x1.fp-16385 -1.5
cacosh -0x1.fp-16385 1.5
cacosh -0x1.fp-16385 -1.5
cacosh 1.5 0x1.fp-16385
cacosh -1.5 0x1.fp-16385
cacosh 1.5 -0x1.fp-16385
cacosh -1.5 -0x1.fp-16385
cacosh 0.5 1.0
cacosh 0.5 -1.0
cacosh -0.5 1.0
cacosh -0.5 -1.0
cacosh 1.0 0.5
cacosh -1.0 0.5
cacosh 1.0 -0.5
cacosh -1.0 -0.5
cacosh 0.25 1.0
cacosh 0.25 -1.0
cacosh -0.25 1.0
cacosh -0.25 -1.0
cacosh 1.0 0.25
cacosh -1.0 0.25
cacosh 1.0 -0.25
cacosh -1.0 -0.25
cacosh 0x1.fp-10 1.0
cacosh 0x1.fp-10 -1.0
cacosh -0x1.fp-10 1.0
cacosh -0x1.fp-10 -1.0
cacosh 1.0 0x1.fp-10
cacosh -1.0 0x1.fp-10
cacosh 1.0 -0x1.fp-10
cacosh -1.0 -0x1.fp-10
cacosh 0x1.fp-30 1.0
cacosh 0x1.fp-30 -1.0
cacosh -0x1.fp-30 1.0
cacosh -0x1.fp-30 -1.0
cacosh 1.0 0x1.fp-30
cacosh -1.0 0x1.fp-30
cacosh 1.0 -0x1.fp-30
cacosh -1.0 -0x1.fp-30
cacosh 0x1.fp-100 1.0
cacosh 0x1.fp-100 -1.0
cacosh -0x1.fp-100 1.0
cacosh -0x1.fp-100 -1.0
cacosh 1.0 0x1.fp-100
cacosh -1.0 0x1.fp-100
cacosh 1.0 -0x1.fp-100
cacosh -1.0 -0x1.fp-100
cacosh 0x1.fp-129 1.0
cacosh 0x1.fp-129 -1.0
cacosh -0x1.fp-129 1.0
cacosh -0x1.fp-129 -1.0
cacosh 1.0 0x1.fp-129
cacosh -1.0 0x1.fp-129
cacosh 1.0 -0x1.fp-129
cacosh -1.0 -0x1.fp-129
cacosh 0x1.fp-1000 1.0
cacosh 0x1.fp-1000 -1.0
cacosh -0x1.fp-1000 1.0
cacosh -0x1.fp-1000 -1.0
cacosh 1.0 0x1.fp-1000
cacosh -1.0 0x1.fp-1000
cacosh 1.0 -0x1.fp-1000
cacosh -1.0 -0x1.fp-1000
cacosh 0x1.fp-1025 1.0
cacosh 0x1.fp-1025 -1.0
cacosh -0x1.fp-1025 1.0
cacosh -0x1.fp-1025 -1.0
cacosh 1.0 0x1.fp-1025
cacosh -1.0 0x1.fp-1025
cacosh 1.0 -0x1.fp-1025
cacosh -1.0 -0x1.fp-1025
cacosh 0x1.fp-10000 1.0
cacosh 0x1.fp-10000 -1.0
cacosh -0x1.fp-10000 1.0
cacosh -0x1.fp-10000 -1.0
cacosh 1.0 0x1.fp-10000
cacosh -1.0 0x1.fp-10000
cacosh 1.0 -0x1.fp-10000
cacosh -1.0 -0x1.fp-10000
cacosh 0x1.fp-16385 1.0
cacosh 0x1.fp-16385 -1.0
cacosh -0x1.fp-16385 1.0
cacosh -0x1.fp-16385 -1.0
cacosh 1.0 0x1.fp-16385
cacosh -1.0 0x1.fp-16385
cacosh 1.0 -0x1.fp-16385
cacosh -1.0 -0x1.fp-16385
cacosh 0x1p-23 0x1.000002p0
cacosh 0x1p-23 -0x1.000002p0
cacosh -0x1p-23 0x1.000002p0
cacosh -0x1p-23 -0x1.000002p0
cacosh 0x1.000002p0 0x1p-23
cacosh -0x1.000002p0 0x1p-23
cacosh 0x1.000002p0 -0x1p-23
cacosh -0x1.000002p0 -0x1p-23
cacosh 0x1.fp-129 0x1.000002p0
cacosh 0x1.fp-129 -0x1.000002p0
cacosh -0x1.fp-129 0x1.000002p0
cacosh -0x1.fp-129 -0x1.000002p0
cacosh 0x1.000002p0 0x1.fp-129
cacosh -0x1.000002p0 0x1.fp-129
cacosh 0x1.000002p0 -0x1.fp-129
cacosh -0x1.000002p0 -0x1.fp-129
cacosh 0.0 0x1.000002p0
cacosh 0.0 -0x1.000002p0
cacosh -0.0 0x1.000002p0
cacosh -0.0 -0x1.000002p0
cacosh 0x1.000002p0 0.0
cacosh -0x1.000002p0 0.0
cacosh 0x1.000002p0 -0.0
cacosh -0x1.000002p0 -0.0
cacosh 0x1p-52 0x1.0000000000001p0
cacosh 0x1p-52 -0x1.0000000000001p0
cacosh -0x1p-52 0x1.0000000000001p0
cacosh -0x1p-52 -0x1.0000000000001p0
cacosh 0x1.0000000000001p0 0x1p-52
cacosh -0x1.0000000000001p0 0x1p-52
cacosh 0x1.0000000000001p0 -0x1p-52
cacosh -0x1.0000000000001p0 -0x1p-52
cacosh 0x1.fp-1025 0x1.0000000000001p0
cacosh 0x1.fp-1025 -0x1.0000000000001p0
cacosh -0x1.fp-1025 0x1.0000000000001p0
cacosh -0x1.fp-1025 -0x1.0000000000001p0
cacosh 0x1.0000000000001p0 0x1.fp-1025
cacosh -0x1.0000000000001p0 0x1.fp-1025
cacosh 0x1.0000000000001p0 -0x1.fp-1025
cacosh -0x1.0000000000001p0 -0x1.fp-1025
cacosh 0.0 0x1.0000000000001p0
cacosh 0.0 -0x1.0000000000001p0
cacosh -0.0 0x1.0000000000001p0
cacosh -0.0 -0x1.0000000000001p0
cacosh 0x1.0000000000001p0 0.0
cacosh -0x1.0000000000001p0 0.0
cacosh 0x1.0000000000001p0 -0.0
cacosh -0x1.0000000000001p0 -0.0
cacosh 0x1p-63 0x1.0000000000000002p0
cacosh 0x1p-63 -0x1.0000000000000002p0
cacosh -0x1p-63 0x1.0000000000000002p0
cacosh -0x1p-63 -0x1.0000000000000002p0
cacosh 0x1.0000000000000002p0 0x1p-63
cacosh -0x1.0000000000000002p0 0x1p-63
cacosh 0x1.0000000000000002p0 -0x1p-63
cacosh -0x1.0000000000000002p0 -0x1p-63
cacosh 0x1.fp-16385 0x1.0000000000000002p0
cacosh 0x1.fp-16385 -0x1.0000000000000002p0
cacosh -0x1.fp-16385 0x1.0000000000000002p0
cacosh -0x1.fp-16385 -0x1.0000000000000002p0
cacosh 0x1.0000000000000002p0 0x1.fp-16385
cacosh -0x1.0000000000000002p0 0x1.fp-16385
cacosh 0x1.0000000000000002p0 -0x1.fp-16385
cacosh -0x1.0000000000000002p0 -0x1.fp-16385
cacosh 0.0 0x1.0000000000000002p0
cacosh 0.0 -0x1.0000000000000002p0
cacosh -0.0 0x1.0000000000000002p0
cacosh -0.0 -0x1.0000000000000002p0
cacosh 0x1.0000000000000002p0 0.0
cacosh -0x1.0000000000000002p0 0.0
cacosh 0x1.0000000000000002p0 -0.0
cacosh -0x1.0000000000000002p0 -0.0
cacosh 0x1p-106 0x1.000000000000000000000000008p0
cacosh 0x1p-106 -0x1.000000000000000000000000008p0
cacosh -0x1p-106 0x1.000000000000000000000000008p0
cacosh -0x1p-106 -0x1.000000000000000000000000008p0
cacosh 0x1.000000000000000000000000008p0 0x1p-106
cacosh -0x1.000000000000000000000000008p0 0x1p-106
cacosh 0x1.000000000000000000000000008p0 -0x1p-106
cacosh -0x1.000000000000000000000000008p0 -0x1p-106
cacosh 0x1.fp-1025 0x1.000000000000000000000000008p0
cacosh 0x1.fp-1025 -0x1.000000000000000000000000008p0
cacosh -0x1.fp-1025 0x1.000000000000000000000000008p0
cacosh -0x1.fp-1025 -0x1.000000000000000000000000008p0
cacosh 0x1.000000000000000000000000008p0 0x1.fp-1025
cacosh -0x1.000000000000000000000000008p0 0x1.fp-1025
cacosh 0x1.000000000000000000000000008p0 -0x1.fp-1025
cacosh -0x1.000000000000000000000000008p0 -0x1.fp-1025
cacosh 0.0 0x1.000000000000000000000000008p0
cacosh 0.0 -0x1.000000000000000000000000008p0
cacosh -0.0 0x1.000000000000000000000000008p0
cacosh -0.0 -0x1.000000000000000000000000008p0
cacosh 0x1.000000000000000000000000008p0 0.0
cacosh -0x1.000000000000000000000000008p0 0.0
cacosh 0x1.000000000000000000000000008p0 -0.0
cacosh -0x1.000000000000000000000000008p0 -0.0
cacosh 0x1p-113 0x1.0000000000000000000000000001p0
cacosh 0x1p-113 -0x1.0000000000000000000000000001p0
cacosh -0x1p-113 0x1.0000000000000000000000000001p0
cacosh -0x1p-113 -0x1.0000000000000000000000000001p0
cacosh 0x1.0000000000000000000000000001p0 0x1p-113
cacosh -0x1.0000000000000000000000000001p0 0x1p-113
cacosh 0x1.0000000000000000000000000001p0 -0x1p-113
cacosh -0x1.0000000000000000000000000001p0 -0x1p-113
cacosh 0x1.fp-16385 0x1.0000000000000000000000000001p0
cacosh 0x1.fp-16385 -0x1.0000000000000000000000000001p0
cacosh -0x1.fp-16385 0x1.0000000000000000000000000001p0
cacosh -0x1.fp-16385 -0x1.0000000000000000000000000001p0
cacosh 0x1.0000000000000000000000000001p0 0x1.fp-16385
cacosh -0x1.0000000000000000000000000001p0 0x1.fp-16385
cacosh 0x1.0000000000000000000000000001p0 -0x1.fp-16385
cacosh -0x1.0000000000000000000000000001p0 -0x1.fp-16385
cacosh 0.0 0x1.0000000000000000000000000001p0
cacosh 0.0 -0x1.0000000000000000000000000001p0
cacosh -0.0 0x1.0000000000000000000000000001p0
cacosh -0.0 -0x1.0000000000000000000000000001p0
cacosh 0x1.0000000000000000000000000001p0 0.0
cacosh -0x1.0000000000000000000000000001p0 0.0
cacosh 0x1.0000000000000000000000000001p0 -0.0
cacosh -0x1.0000000000000000000000000001p0 -0.0
cacosh 0x1p-23 0x0.ffffffp0
cacosh 0x1p-23 -0x0.ffffffp0
cacosh -0x1p-23 0x0.ffffffp0
cacosh -0x1p-23 -0x0.ffffffp0
cacosh 0x0.ffffffp0 0x1p-23
cacosh -0x0.ffffffp0 0x1p-23
cacosh 0x0.ffffffp0 -0x1p-23
cacosh -0x0.ffffffp0 -0x1p-23
cacosh 0x1.fp-129 0x0.ffffffp0
cacosh 0x1.fp-129 -0x0.ffffffp0
cacosh -0x1.fp-129 0x0.ffffffp0
cacosh -0x1.fp-129 -0x0.ffffffp0
cacosh 0x0.ffffffp0 0x1.fp-129
cacosh -0x0.ffffffp0 0x1.fp-129
cacosh 0x0.ffffffp0 -0x1.fp-129
cacosh -0x0.ffffffp0 -0x1.fp-129
cacosh 0.0 0x0.ffffffp0
cacosh 0.0 -0x0.ffffffp0
cacosh -0.0 0x0.ffffffp0
cacosh -0.0 -0x0.ffffffp0
cacosh 0x0.ffffffp0 0.0
cacosh -0x0.ffffffp0 0.0
cacosh 0x0.ffffffp0 -0.0
cacosh -0x0.ffffffp0 -0.0
cacosh 0x1p-23 0.5
cacosh 0x1p-23 -0.5
cacosh -0x1p-23 0.5
cacosh -0x1p-23 -0.5
cacosh 0.5 0x1p-23
cacosh -0.5 0x1p-23
cacosh 0.5 -0x1p-23
cacosh -0.5 -0x1p-23
cacosh 0x1.fp-129 0.5
cacosh 0x1.fp-129 -0.5
cacosh -0x1.fp-129 0.5
cacosh -0x1.fp-129 -0.5
cacosh 0.5 0x1.fp-129
cacosh -0.5 0x1.fp-129
cacosh 0.5 -0x1.fp-129
cacosh -0.5 -0x1.fp-129
cacosh 0x1p-23 0x1p-23
cacosh 0x1p-23 -0x1p-23
cacosh -0x1p-23 0x1p-23
cacosh -0x1p-23 -0x1p-23
cacosh 0x1.fp-129 0x1p-23
cacosh 0x1.fp-129 -0x1p-23
cacosh -0x1.fp-129 0x1p-23
cacosh -0x1.fp-129 -0x1p-23
cacosh 0x1p-23 0x1.fp-129
cacosh -0x1p-23 0x1.fp-129
cacosh 0x1p-23 -0x1.fp-129
cacosh -0x1p-23 -0x1.fp-129
cacosh 0.0 0x1p-23
cacosh 0.0 -0x1p-23
cacosh -0.0 0x1p-23
cacosh -0.0 -0x1p-23
cacosh 0x1p-23 0.0
cacosh -0x1p-23 0.0
cacosh 0x1p-23 -0.0
cacosh -0x1p-23 -0.0
cacosh 0x1.fp-129 0x1.fp-129
cacosh 0x1.fp-129 -0x1.fp-129
cacosh -0x1.fp-129 0x1.fp-129
cacosh -0x1.fp-129 -0x1.fp-129
cacosh 0x1p-52 0x0.fffffffffffff8p0
cacosh 0x1p-52 -0x0.fffffffffffff8p0
cacosh -0x1p-52 0x0.fffffffffffff8p0
cacosh -0x1p-52 -0x0.fffffffffffff8p0
cacosh 0x0.fffffffffffff8p0 0x1p-52
cacosh -0x0.fffffffffffff8p0 0x1p-52
cacosh 0x0.fffffffffffff8p0 -0x1p-52
cacosh -0x0.fffffffffffff8p0 -0x1p-52
cacosh 0x1.fp-1025 0x0.fffffffffffff8p0
cacosh 0x1.fp-1025 -0x0.fffffffffffff8p0
cacosh -0x1.fp-1025 0x0.fffffffffffff8p0
cacosh -0x1.fp-1025 -0x0.fffffffffffff8p0
cacosh 0x0.fffffffffffff8p0 0x1.fp-1025
cacosh -0x0.fffffffffffff8p0 0x1.fp-1025
cacosh 0x0.fffffffffffff8p0 -0x1.fp-1025
cacosh -0x0.fffffffffffff8p0 -0x1.fp-1025
cacosh 0.0 0x0.fffffffffffff8p0
cacosh 0.0 -0x0.fffffffffffff8p0
cacosh -0.0 0x0.fffffffffffff8p0
cacosh -0.0 -0x0.fffffffffffff8p0
cacosh 0x0.fffffffffffff8p0 0.0
cacosh -0x0.fffffffffffff8p0 0.0
cacosh 0x0.fffffffffffff8p0 -0.0
cacosh -0x0.fffffffffffff8p0 -0.0
cacosh 0x1p-52 0.5
cacosh 0x1p-52 -0.5
cacosh -0x1p-52 0.5
cacosh -0x1p-52 -0.5
cacosh 0.5 0x1p-52
cacosh -0.5 0x1p-52
cacosh 0.5 -0x1p-52
cacosh -0.5 -0x1p-52
cacosh 0x1.fp-1025 0.5
cacosh 0x1.fp-1025 -0.5
cacosh -0x1.fp-1025 0.5
cacosh -0x1.fp-1025 -0.5
cacosh 0.5 0x1.fp-1025
cacosh -0.5 0x1.fp-1025
cacosh 0.5 -0x1.fp-1025
cacosh -0.5 -0x1.fp-1025
cacosh 0x1p-52 0x1p-52
cacosh 0x1p-52 -0x1p-52
cacosh -0x1p-52 0x1p-52
cacosh -0x1p-52 -0x1p-52
cacosh 0x1.fp-1025 0x1p-52
cacosh 0x1.fp-1025 -0x1p-52
cacosh -0x1.fp-1025 0x1p-52
cacosh -0x1.fp-1025 -0x1p-52
cacosh 0x1p-52 0x1.fp-1025
cacosh -0x1p-52 0x1.fp-1025
cacosh 0x1p-52 -0x1.fp-1025
cacosh -0x1p-52 -0x1.fp-1025
cacosh 0.0 0x1p-52
cacosh 0.0 -0x1p-52
cacosh -0.0 0x1p-52
cacosh -0.0 -0x1p-52
cacosh 0x1p-52 0.0
cacosh -0x1p-52 0.0
cacosh 0x1p-52 -0.0
cacosh -0x1p-52 -0.0
cacosh 0x1.fp-1025 0x1.fp-1025
cacosh 0x1.fp-1025 -0x1.fp-1025
cacosh -0x1.fp-1025 0x1.fp-1025
cacosh -0x1.fp-1025 -0x1.fp-1025
cacosh 0x1p-63 0x0.ffffffffffffffffp0
cacosh 0x1p-63 -0x0.ffffffffffffffffp0
cacosh -0x1p-63 0x0.ffffffffffffffffp0
cacosh -0x1p-63 -0x0.ffffffffffffffffp0
cacosh 0x0.ffffffffffffffffp0 0x1p-63
cacosh -0x0.ffffffffffffffffp0 0x1p-63
cacosh 0x0.ffffffffffffffffp0 -0x1p-63
cacosh -0x0.ffffffffffffffffp0 -0x1p-63
cacosh 0x1.fp-16385 0x0.ffffffffffffffffp0
cacosh 0x1.fp-16385 -0x0.ffffffffffffffffp0
cacosh -0x1.fp-16385 0x0.ffffffffffffffffp0
cacosh -0x1.fp-16385 -0x0.ffffffffffffffffp0
cacosh 0x0.ffffffffffffffffp0 0x1.fp-16385
cacosh -0x0.ffffffffffffffffp0 0x1.fp-16385
cacosh 0x0.ffffffffffffffffp0 -0x1.fp-16385
cacosh -0x0.ffffffffffffffffp0 -0x1.fp-16385
cacosh 0.0 0x0.ffffffffffffffffp0
cacosh 0.0 -0x0.ffffffffffffffffp0
cacosh -0.0 0x0.ffffffffffffffffp0
cacosh -0.0 -0x0.ffffffffffffffffp0
cacosh 0x0.ffffffffffffffffp0 0.0
cacosh -0x0.ffffffffffffffffp0 0.0
cacosh 0x0.ffffffffffffffffp0 -0.0
cacosh -0x0.ffffffffffffffffp0 -0.0
cacosh 0x1p-63 0.5
cacosh 0x1p-63 -0.5
cacosh -0x1p-63 0.5
cacosh -0x1p-63 -0.5
cacosh 0.5 0x1p-63
cacosh -0.5 0x1p-63
cacosh 0.5 -0x1p-63
cacosh -0.5 -0x1p-63
cacosh 0x1.fp-16385 0.5
cacosh 0x1.fp-16385 -0.5
cacosh -0x1.fp-16385 0.5
cacosh -0x1.fp-16385 -0.5
cacosh 0.5 0x1.fp-16385
cacosh -0.5 0x1.fp-16385
cacosh 0.5 -0x1.fp-16385
cacosh -0.5 -0x1.fp-16385
cacosh 0x1p-63 0x1p-63
cacosh 0x1p-63 -0x1p-63
cacosh -0x1p-63 0x1p-63
cacosh -0x1p-63 -0x1p-63
cacosh 0x1.fp-16385 0x1p-63
cacosh 0x1.fp-16385 -0x1p-63
cacosh -0x1.fp-16385 0x1p-63
cacosh -0x1.fp-16385 -0x1p-63
cacosh 0x1p-63 0x1.fp-16385
cacosh -0x1p-63 0x1.fp-16385
cacosh 0x1p-63 -0x1.fp-16385
cacosh -0x1p-63 -0x1.fp-16385
cacosh 0.0 0x1p-63
cacosh 0.0 -0x1p-63
cacosh -0.0 0x1p-63
cacosh -0.0 -0x1p-63
cacosh 0x1p-63 0.0
cacosh -0x1p-63 0.0
cacosh 0x1p-63 -0.0
cacosh -0x1p-63 -0.0
cacosh 0x1.fp-16385 0x1.fp-16385
cacosh 0x1.fp-16385 -0x1.fp-16385
cacosh -0x1.fp-16385 0x1.fp-16385
cacosh -0x1.fp-16385 -0x1.fp-16385
cacosh 0x1p-105 0x0.ffffffffffffffffffffffffffcp0
cacosh 0x1p-105 -0x0.ffffffffffffffffffffffffffcp0
cacosh -0x1p-105 0x0.ffffffffffffffffffffffffffcp0
cacosh -0x1p-105 -0x0.ffffffffffffffffffffffffffcp0
cacosh 0x0.ffffffffffffffffffffffffffcp0 0x1p-105
cacosh -0x0.ffffffffffffffffffffffffffcp0 0x1p-105
cacosh 0x0.ffffffffffffffffffffffffffcp0 -0x1p-105
cacosh -0x0.ffffffffffffffffffffffffffcp0 -0x1p-105
cacosh 0x1.fp-1025 0x0.ffffffffffffffffffffffffffcp0
cacosh 0x1.fp-1025 -0x0.ffffffffffffffffffffffffffcp0
cacosh -0x1.fp-1025 0x0.ffffffffffffffffffffffffffcp0
cacosh -0x1.fp-1025 -0x0.ffffffffffffffffffffffffffcp0
cacosh 0x0.ffffffffffffffffffffffffffcp0 0x1.fp-1025
cacosh -0x0.ffffffffffffffffffffffffffcp0 0x1.fp-1025
cacosh 0x0.ffffffffffffffffffffffffffcp0 -0x1.fp-1025
cacosh -0x0.ffffffffffffffffffffffffffcp0 -0x1.fp-1025
cacosh 0.0 0x0.ffffffffffffffffffffffffffcp0
cacosh 0.0 -0x0.ffffffffffffffffffffffffffcp0
cacosh -0.0 0x0.ffffffffffffffffffffffffffcp0
cacosh -0.0 -0x0.ffffffffffffffffffffffffffcp0
cacosh 0x0.ffffffffffffffffffffffffffcp0 0.0
cacosh -0x0.ffffffffffffffffffffffffffcp0 0.0
cacosh 0x0.ffffffffffffffffffffffffffcp0 -0.0
cacosh -0x0.ffffffffffffffffffffffffffcp0 -0.0
cacosh 0x1p-105 0.5
cacosh 0x1p-105 -0.5
cacosh -0x1p-105 0.5
cacosh -0x1p-105 -0.5
cacosh 0.5 0x1p-105
cacosh -0.5 0x1p-105
cacosh 0.5 -0x1p-105
cacosh -0.5 -0x1p-105
cacosh 0x1p-105 0x1p-105
cacosh 0x1p-105 -0x1p-105
cacosh -0x1p-105 0x1p-105
cacosh -0x1p-105 -0x1p-105
cacosh 0x1.fp-1025 0x1p-105
cacosh 0x1.fp-1025 -0x1p-105
cacosh -0x1.fp-1025 0x1p-105
cacosh -0x1.fp-1025 -0x1p-105
cacosh 0x1p-105 0x1.fp-1025
cacosh -0x1p-105 0x1.fp-1025
cacosh 0x1p-105 -0x1.fp-1025
cacosh -0x1p-105 -0x1.fp-1025
cacosh 0.0 0x1p-105
cacosh 0.0 -0x1p-105
cacosh -0.0 0x1p-105
cacosh -0.0 -0x1p-105
cacosh 0x1p-105 0.0
cacosh -0x1p-105 0.0
cacosh 0x1p-105 -0.0
cacosh -0x1p-105 -0.0
cacosh 0x1p-112 0x0.ffffffffffffffffffffffffffff8p0
cacosh 0x1p-112 -0x0.ffffffffffffffffffffffffffff8p0
cacosh -0x1p-112 0x0.ffffffffffffffffffffffffffff8p0
cacosh -0x1p-112 -0x0.ffffffffffffffffffffffffffff8p0
cacosh 0x0.ffffffffffffffffffffffffffff8p0 0x1p-112
cacosh -0x0.ffffffffffffffffffffffffffff8p0 0x1p-112
cacosh 0x0.ffffffffffffffffffffffffffff8p0 -0x1p-112
cacosh -0x0.ffffffffffffffffffffffffffff8p0 -0x1p-112
cacosh 0x1.fp-16385 0x0.ffffffffffffffffffffffffffff8p0
cacosh 0x1.fp-16385 -0x0.ffffffffffffffffffffffffffff8p0
cacosh -0x1.fp-16385 0x0.ffffffffffffffffffffffffffff8p0
cacosh -0x1.fp-16385 -0x0.ffffffffffffffffffffffffffff8p0
cacosh 0x0.ffffffffffffffffffffffffffff8p0 0x1.fp-16385
cacosh -0x0.ffffffffffffffffffffffffffff8p0 0x1.fp-16385
cacosh 0x0.ffffffffffffffffffffffffffff8p0 -0x1.fp-16385
cacosh -0x0.ffffffffffffffffffffffffffff8p0 -0x1.fp-16385
cacosh 0.0 0x0.ffffffffffffffffffffffffffff8p0
cacosh 0.0 -0x0.ffffffffffffffffffffffffffff8p0
cacosh -0.0 0x0.ffffffffffffffffffffffffffff8p0
cacosh -0.0 -0x0.ffffffffffffffffffffffffffff8p0
cacosh 0x0.ffffffffffffffffffffffffffff8p0 0.0
cacosh -0x0.ffffffffffffffffffffffffffff8p0 0.0
cacosh 0x0.ffffffffffffffffffffffffffff8p0 -0.0
cacosh -0x0.ffffffffffffffffffffffffffff8p0 -0.0
cacosh 0x1p-112 0.5
cacosh 0x1p-112 -0.5
cacosh -0x1p-112 0.5
cacosh -0x1p-112 -0.5
cacosh 0.5 0x1p-112
cacosh -0.5 0x1p-112
cacosh 0.5 -0x1p-112
cacosh -0.5 -0x1p-112
cacosh 0x1p-112 0x1p-112
cacosh 0x1p-112 -0x1p-112
cacosh -0x1p-112 0x1p-112
cacosh -0x1p-112 -0x1p-112
cacosh 0x1.fp-16385 0x1p-112
cacosh 0x1.fp-16385 -0x1p-112
cacosh -0x1.fp-16385 0x1p-112
cacosh -0x1.fp-16385 -0x1p-112
cacosh 0x1p-112 0x1.fp-16385
cacosh -0x1p-112 0x1.fp-16385
cacosh 0x1p-112 -0x1.fp-16385
cacosh -0x1p-112 -0x1.fp-16385
cacosh 0.0 0x1p-112
cacosh 0.0 -0x1p-112
cacosh -0.0 0x1p-112
cacosh -0.0 -0x1p-112
cacosh 0x1p-112 0.0
cacosh -0x1p-112 0.0
cacosh 0x1p-112 -0.0
cacosh -0x1p-112 -0.0
cacosh 0.75 1.25
cacosh -2 -3
Move tests of cabs and carg from libm-test.inc to auto-libm-test-in. This patch moves tests of cabs and carg to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86; no ulps updates needed. * math/auto-libm-test-in: Add tests of cabs and carg. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cabs_test_data): Use AUTO_TESTS_c_f. (carg_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_f. (func_calc_desc): Add mpc_c_f union field. (test_functions): Add cabs and carg. (calc_generic_results): Handle mpc_c_f.
2013-12-19 21:28:30 +00:00
# carg (x + i 0) == 0 for x > 0.
carg 2.0 0
# carg (x - i 0) == -0 for x > 0.
carg 2.0 -0
carg 0 0
carg 0 -0
# carg (x + i 0) == +pi for x < 0.
carg -2.0 0
# carg (x - i 0) == -pi for x < 0.
carg -2.0 -0
carg -0 0
carg -0 -0
# carg (+0 + i y) == pi/2 for y > 0.
carg 0 2.0
# carg (-0 + i y) == pi/2 for y > 0.
carg -0 2.0
# carg (+0 + i y) == -pi/2 for y < 0.
carg 0 -2.0
# carg (-0 + i y) == -pi/2 for y < 0.
carg -0 -2.0
Fix dbl-64 atan2 in non-default rounding modes (bug 18210, bug 18211). The dbl-64 implementation of atan2 does computations that expect to run in round-to-nearest mode, and in other modes the errors can accumulate to more than the maximum accepted 9ulp. This patch makes it use FE_TONEAREST internally, similar to other functions with such issues. Tests that previously produced large errors are added for atan2 and the closely related carg, clog and clog10 functions. Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18210] [BZ #18211] * sysdeps/ieee754/dbl-64/e_atan2.c: Include <fenv.h>. (__ieee754_atan2): Set FE_TONEAREST mode for internal computations. * math/auto-libm-test-in: Add more tests of atan2, carg, clog and clog10. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-08 17:32:17 +00:00
carg 0x2.f2f308p+0 0x4.c3841p-4
carg 0xd.3de7ap-36 -0xe.cf143p-40
carg 0x2.21e65p+0 0x5.576cf8p-4
carg 0x1.f4755cp+0 -0x4.29411p-4
carg -0xf.9c4c8p-4 -0xa.b4101p+20
carg 0x7.40ac68p+0 0x4.251bb8p-4
carg 0xa.3ac3cp+68 0x1.47239ep+68
carg 0x3.8ff10cp+0 -0x6.b0794p-4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
carg -0x3.973cc4p+72 -0xf.fffffp+124
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
carg -0x1.0a512ap-120 0xf.54681p-108
carg -0x3.be0054531569p-4 -0xb.0c5a9p-4
carg -0x1.0236b6p-20 0x2.a6e504p+108
carg 0x9.27b5fd9157b6c92b151371ca23d8p+0 0x1.d8759b9024992p+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
carg 0x8p-152 0x4p-1076
Move tests of cabs and carg from libm-test.inc to auto-libm-test-in. This patch moves tests of cabs and carg to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86; no ulps updates needed. * math/auto-libm-test-in: Add tests of cabs and carg. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cabs_test_data): Use AUTO_TESTS_c_f. (carg_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_f. (func_calc_desc): Add mpc_c_f union field. (test_functions): Add cabs and carg. (calc_generic_results): Handle mpc_c_f.
2013-12-19 21:28:30 +00:00
Move tests of casin, casinh to auto-libm-test-*. This patch moves tests of casin and casinh with finite inputs to using the auto-libm-test machinery. Each of auto-libm-test-out-casin and auto-libm-test-out-casinh takes about 38 minutes to generate on my system because of MPC slowness on special cases that appear in the tests (with MPC 1.0.3; I don't know to what extent current MPC master might speed it up). Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of casin and casinh. * math/auto-libm-test-out-casin: New generated file. * math/auto-libm-test-out-casinh: Likewise. * math/libm-test-casin.inc (casin_test_data): Use AUTO_TESTS_c_c. Move tests with finite inputs to auto-libm-test-in. * math/libm-test-casinh.inc (casinh_test_data): Likewise. * math/Makefile (libm-test-funcs-auto): Add casin and casinh. (libm-test-funcs-noauto): Remove casin and casinh. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2017-02-17 18:14:02 +00:00
casin 0 0
casin -0 0
casin 0 -0
casin -0 -0
casin 0 -1.5
casin -0 -1.5
casin 0 -1.0
casin -0 -1.0
casin 0 -0.5
casin -0 -0.5
casin 0 0.5
casin -0 0.5
casin 0 1.0
casin -0 1.0
casin 0 1.5
casin -0 1.5
casin -1.5 0
casin -1.5 -0
casin -1.0 0
casin -1.0 -0
casin -0.5 0
casin -0.5 -0
casin 0.5 0
casin 0.5 -0
casin 1.0 0
casin 1.0 -0
casin 1.5 0
casin 1.5 -0
casin 0x1p50 1.0
casin 0x1p50 -1.0
casin -0x1p50 1.0
casin -0x1p50 -1.0
casin 1.0 0x1p50
casin -1.0 0x1p50
casin 1.0 -0x1p50
casin -1.0 -0x1p50
casin 0x1p500 1.0
casin 0x1p500 -1.0
casin -0x1p500 1.0
casin -0x1p500 -1.0
casin 1.0 0x1p500
casin -1.0 0x1p500
casin 1.0 -0x1p500
casin -1.0 -0x1p500
casin 0x1p5000 1.0
casin 0x1p5000 -1.0
casin -0x1p5000 1.0
casin -0x1p5000 -1.0
casin 1.0 0x1p5000
casin -1.0 0x1p5000
casin 1.0 -0x1p5000
casin -1.0 -0x1p5000
casin 0x1.fp127 0x1.fp127
casin 0x1.fp1023 0x1.fp1023
casin 0x1.fp16383 0x1.fp16383
casin 0x1.fp-129 1.5
casin 0x1.fp-129 -1.5
casin -0x1.fp-129 1.5
casin -0x1.fp-129 -1.5
casin 1.5 0x1.fp-129
casin -1.5 0x1.fp-129
casin 1.5 -0x1.fp-129
casin -1.5 -0x1.fp-129
casin 0x1.fp-1025 1.5
casin 0x1.fp-1025 -1.5
casin -0x1.fp-1025 1.5
casin -0x1.fp-1025 -1.5
casin 1.5 0x1.fp-1025
casin -1.5 0x1.fp-1025
casin 1.5 -0x1.fp-1025
casin -1.5 -0x1.fp-1025
casin 0x1.fp-16385 1.5
casin 0x1.fp-16385 -1.5
casin -0x1.fp-16385 1.5
casin -0x1.fp-16385 -1.5
casin 1.5 0x1.fp-16385
casin -1.5 0x1.fp-16385
casin 1.5 -0x1.fp-16385
casin -1.5 -0x1.fp-16385
casin 0.5 1.0
casin 0.5 -1.0
casin -0.5 1.0
casin -0.5 -1.0
casin 1.0 0.5
casin -1.0 0.5
casin 1.0 -0.5
casin -1.0 -0.5
casin 0.25 1.0
casin 0.25 -1.0
casin -0.25 1.0
casin -0.25 -1.0
casin 1.0 0.25
casin -1.0 0.25
casin 1.0 -0.25
casin -1.0 -0.25
casin 0x1.fp-10 1.0
casin 0x1.fp-10 -1.0
casin -0x1.fp-10 1.0
casin -0x1.fp-10 -1.0
casin 1.0 0x1.fp-10
casin -1.0 0x1.fp-10
casin 1.0 -0x1.fp-10
casin -1.0 -0x1.fp-10
casin 0x1.fp-30 1.0
casin 0x1.fp-30 -1.0
casin -0x1.fp-30 1.0
casin -0x1.fp-30 -1.0
casin 1.0 0x1.fp-30
casin -1.0 0x1.fp-30
casin 1.0 -0x1.fp-30
casin -1.0 -0x1.fp-30
casin 0x1.fp-100 1.0
casin 0x1.fp-100 -1.0
casin -0x1.fp-100 1.0
casin -0x1.fp-100 -1.0
casin 1.0 0x1.fp-100
casin -1.0 0x1.fp-100
casin 1.0 -0x1.fp-100
casin -1.0 -0x1.fp-100
casin 0x1.fp-129 1.0
casin 0x1.fp-129 -1.0
casin -0x1.fp-129 1.0
casin -0x1.fp-129 -1.0
casin 1.0 0x1.fp-129
casin -1.0 0x1.fp-129
casin 1.0 -0x1.fp-129
casin -1.0 -0x1.fp-129
casin 0x1.fp-1000 1.0
casin 0x1.fp-1000 -1.0
casin -0x1.fp-1000 1.0
casin -0x1.fp-1000 -1.0
casin 1.0 0x1.fp-1000
casin -1.0 0x1.fp-1000
casin 1.0 -0x1.fp-1000
casin -1.0 -0x1.fp-1000
casin 0x1.fp-1025 1.0
casin 0x1.fp-1025 -1.0
casin -0x1.fp-1025 1.0
casin -0x1.fp-1025 -1.0
casin 1.0 0x1.fp-1025
casin -1.0 0x1.fp-1025
casin 1.0 -0x1.fp-1025
casin -1.0 -0x1.fp-1025
casin 0x1.fp-10000 1.0
casin 0x1.fp-10000 -1.0
casin -0x1.fp-10000 1.0
casin -0x1.fp-10000 -1.0
casin 1.0 0x1.fp-10000
casin -1.0 0x1.fp-10000
casin 1.0 -0x1.fp-10000
casin -1.0 -0x1.fp-10000
casin 0x1.fp-16385 1.0
casin 0x1.fp-16385 -1.0
casin -0x1.fp-16385 1.0
casin -0x1.fp-16385 -1.0
casin 1.0 0x1.fp-16385
casin -1.0 0x1.fp-16385
casin 1.0 -0x1.fp-16385
casin -1.0 -0x1.fp-16385
casin 0x1p-23 0x1.000002p0
casin 0x1p-23 -0x1.000002p0
casin -0x1p-23 0x1.000002p0
casin -0x1p-23 -0x1.000002p0
casin 0x1.000002p0 0x1p-23
casin -0x1.000002p0 0x1p-23
casin 0x1.000002p0 -0x1p-23
casin -0x1.000002p0 -0x1p-23
casin 0x1.fp-129 0x1.000002p0
casin 0x1.fp-129 -0x1.000002p0
casin -0x1.fp-129 0x1.000002p0
casin -0x1.fp-129 -0x1.000002p0
casin 0x1.000002p0 0x1.fp-129
casin -0x1.000002p0 0x1.fp-129
casin 0x1.000002p0 -0x1.fp-129
casin -0x1.000002p0 -0x1.fp-129
casin 0.0 0x1.000002p0
casin 0.0 -0x1.000002p0
casin -0.0 0x1.000002p0
casin -0.0 -0x1.000002p0
casin 0x1.000002p0 0.0
casin -0x1.000002p0 0.0
casin 0x1.000002p0 -0.0
casin -0x1.000002p0 -0.0
casin 0x1p-52 0x1.0000000000001p0
casin 0x1p-52 -0x1.0000000000001p0
casin -0x1p-52 0x1.0000000000001p0
casin -0x1p-52 -0x1.0000000000001p0
casin 0x1.0000000000001p0 0x1p-52
casin -0x1.0000000000001p0 0x1p-52
casin 0x1.0000000000001p0 -0x1p-52
casin -0x1.0000000000001p0 -0x1p-52
casin 0x1.fp-1025 0x1.0000000000001p0
casin 0x1.fp-1025 -0x1.0000000000001p0
casin -0x1.fp-1025 0x1.0000000000001p0
casin -0x1.fp-1025 -0x1.0000000000001p0
casin 0x1.0000000000001p0 0x1.fp-1025
casin -0x1.0000000000001p0 0x1.fp-1025
casin 0x1.0000000000001p0 -0x1.fp-1025
casin -0x1.0000000000001p0 -0x1.fp-1025
casin 0.0 0x1.0000000000001p0
casin 0.0 -0x1.0000000000001p0
casin -0.0 0x1.0000000000001p0
casin -0.0 -0x1.0000000000001p0
casin 0x1.0000000000001p0 0.0
casin -0x1.0000000000001p0 0.0
casin 0x1.0000000000001p0 -0.0
casin -0x1.0000000000001p0 -0.0
casin 0x1p-63 0x1.0000000000000002p0
casin 0x1p-63 -0x1.0000000000000002p0
casin -0x1p-63 0x1.0000000000000002p0
casin -0x1p-63 -0x1.0000000000000002p0
casin 0x1.0000000000000002p0 0x1p-63
casin -0x1.0000000000000002p0 0x1p-63
casin 0x1.0000000000000002p0 -0x1p-63
casin -0x1.0000000000000002p0 -0x1p-63
casin 0x1.fp-16385 0x1.0000000000000002p0
casin 0x1.fp-16385 -0x1.0000000000000002p0
casin -0x1.fp-16385 0x1.0000000000000002p0
casin -0x1.fp-16385 -0x1.0000000000000002p0
casin 0x1.0000000000000002p0 0x1.fp-16385
casin -0x1.0000000000000002p0 0x1.fp-16385
casin 0x1.0000000000000002p0 -0x1.fp-16385
casin -0x1.0000000000000002p0 -0x1.fp-16385
casin 0.0 0x1.0000000000000002p0
casin 0.0 -0x1.0000000000000002p0
casin -0.0 0x1.0000000000000002p0
casin -0.0 -0x1.0000000000000002p0
casin 0x1.0000000000000002p0 0.0
casin -0x1.0000000000000002p0 0.0
casin 0x1.0000000000000002p0 -0.0
casin -0x1.0000000000000002p0 -0.0
casin 0x1p-106 0x1.000000000000000000000000008p0
casin 0x1p-106 -0x1.000000000000000000000000008p0
casin -0x1p-106 0x1.000000000000000000000000008p0
casin -0x1p-106 -0x1.000000000000000000000000008p0
casin 0x1.000000000000000000000000008p0 0x1p-106
casin -0x1.000000000000000000000000008p0 0x1p-106
casin 0x1.000000000000000000000000008p0 -0x1p-106
casin -0x1.000000000000000000000000008p0 -0x1p-106
casin 0x1.fp-1025 0x1.000000000000000000000000008p0
casin 0x1.fp-1025 -0x1.000000000000000000000000008p0
casin -0x1.fp-1025 0x1.000000000000000000000000008p0
casin -0x1.fp-1025 -0x1.000000000000000000000000008p0
casin 0x1.000000000000000000000000008p0 0x1.fp-1025
casin -0x1.000000000000000000000000008p0 0x1.fp-1025
casin 0x1.000000000000000000000000008p0 -0x1.fp-1025
casin -0x1.000000000000000000000000008p0 -0x1.fp-1025
casin 0.0 0x1.000000000000000000000000008p0
casin 0.0 -0x1.000000000000000000000000008p0
casin -0.0 0x1.000000000000000000000000008p0
casin -0.0 -0x1.000000000000000000000000008p0
casin 0x1.000000000000000000000000008p0 0.0
casin -0x1.000000000000000000000000008p0 0.0
casin 0x1.000000000000000000000000008p0 -0.0
casin -0x1.000000000000000000000000008p0 -0.0
casin 0x1p-113 0x1.0000000000000000000000000001p0
casin 0x1p-113 -0x1.0000000000000000000000000001p0
casin -0x1p-113 0x1.0000000000000000000000000001p0
casin -0x1p-113 -0x1.0000000000000000000000000001p0
casin 0x1.0000000000000000000000000001p0 0x1p-113
casin -0x1.0000000000000000000000000001p0 0x1p-113
casin 0x1.0000000000000000000000000001p0 -0x1p-113
casin -0x1.0000000000000000000000000001p0 -0x1p-113
casin 0x1.fp-16385 0x1.0000000000000000000000000001p0
casin 0x1.fp-16385 -0x1.0000000000000000000000000001p0
casin -0x1.fp-16385 0x1.0000000000000000000000000001p0
casin -0x1.fp-16385 -0x1.0000000000000000000000000001p0
casin 0x1.0000000000000000000000000001p0 0x1.fp-16385
casin -0x1.0000000000000000000000000001p0 0x1.fp-16385
casin 0x1.0000000000000000000000000001p0 -0x1.fp-16385
casin -0x1.0000000000000000000000000001p0 -0x1.fp-16385
casin 0.0 0x1.0000000000000000000000000001p0
casin 0.0 -0x1.0000000000000000000000000001p0
casin -0.0 0x1.0000000000000000000000000001p0
casin -0.0 -0x1.0000000000000000000000000001p0
casin 0x1.0000000000000000000000000001p0 0.0
casin -0x1.0000000000000000000000000001p0 0.0
casin 0x1.0000000000000000000000000001p0 -0.0
casin -0x1.0000000000000000000000000001p0 -0.0
casin 0x1p-23 0x0.ffffffp0
casin 0x1p-23 -0x0.ffffffp0
casin -0x1p-23 0x0.ffffffp0
casin -0x1p-23 -0x0.ffffffp0
casin 0x0.ffffffp0 0x1p-23
casin -0x0.ffffffp0 0x1p-23
casin 0x0.ffffffp0 -0x1p-23
casin -0x0.ffffffp0 -0x1p-23
casin 0x1.fp-129 0x0.ffffffp0
casin 0x1.fp-129 -0x0.ffffffp0
casin -0x1.fp-129 0x0.ffffffp0
casin -0x1.fp-129 -0x0.ffffffp0
casin 0x0.ffffffp0 0x1.fp-129
casin -0x0.ffffffp0 0x1.fp-129
casin 0x0.ffffffp0 -0x1.fp-129
casin -0x0.ffffffp0 -0x1.fp-129
casin 0.0 0x0.ffffffp0
casin 0.0 -0x0.ffffffp0
casin -0.0 0x0.ffffffp0
casin -0.0 -0x0.ffffffp0
casin 0x0.ffffffp0 0.0
casin -0x0.ffffffp0 0.0
casin 0x0.ffffffp0 -0.0
casin -0x0.ffffffp0 -0.0
casin 0x1p-23 0.5
casin 0x1p-23 -0.5
casin -0x1p-23 0.5
casin -0x1p-23 -0.5
casin 0.5 0x1p-23
casin -0.5 0x1p-23
casin 0.5 -0x1p-23
casin -0.5 -0x1p-23
casin 0x1.fp-129 0.5
casin 0x1.fp-129 -0.5
casin -0x1.fp-129 0.5
casin -0x1.fp-129 -0.5
casin 0.5 0x1.fp-129
casin -0.5 0x1.fp-129
casin 0.5 -0x1.fp-129
casin -0.5 -0x1.fp-129
casin 0x1p-23 0x1p-23
casin 0x1p-23 -0x1p-23
casin -0x1p-23 0x1p-23
casin -0x1p-23 -0x1p-23
casin 0x1.fp-129 0x1p-23
casin 0x1.fp-129 -0x1p-23
casin -0x1.fp-129 0x1p-23
casin -0x1.fp-129 -0x1p-23
casin 0x1p-23 0x1.fp-129
casin -0x1p-23 0x1.fp-129
casin 0x1p-23 -0x1.fp-129
casin -0x1p-23 -0x1.fp-129
casin 0.0 0x1p-23
casin 0.0 -0x1p-23
casin -0.0 0x1p-23
casin -0.0 -0x1p-23
casin 0x1p-23 0.0
casin -0x1p-23 0.0
casin 0x1p-23 -0.0
casin -0x1p-23 -0.0
casin 0x1.fp-129 0x1.fp-129
casin 0x1.fp-129 -0x1.fp-129
casin -0x1.fp-129 0x1.fp-129
casin -0x1.fp-129 -0x1.fp-129
casin 0x1p-52 0x0.fffffffffffff8p0
casin 0x1p-52 -0x0.fffffffffffff8p0
casin -0x1p-52 0x0.fffffffffffff8p0
casin -0x1p-52 -0x0.fffffffffffff8p0
casin 0x0.fffffffffffff8p0 0x1p-52
casin -0x0.fffffffffffff8p0 0x1p-52
casin 0x0.fffffffffffff8p0 -0x1p-52
casin -0x0.fffffffffffff8p0 -0x1p-52
casin 0x1.fp-1025 0x0.fffffffffffff8p0
casin 0x1.fp-1025 -0x0.fffffffffffff8p0
casin -0x1.fp-1025 0x0.fffffffffffff8p0
casin -0x1.fp-1025 -0x0.fffffffffffff8p0
casin 0x0.fffffffffffff8p0 0x1.fp-1025
casin -0x0.fffffffffffff8p0 0x1.fp-1025
casin 0x0.fffffffffffff8p0 -0x1.fp-1025
casin -0x0.fffffffffffff8p0 -0x1.fp-1025
casin 0.0 0x0.fffffffffffff8p0
casin 0.0 -0x0.fffffffffffff8p0
casin -0.0 0x0.fffffffffffff8p0
casin -0.0 -0x0.fffffffffffff8p0
casin 0x0.fffffffffffff8p0 0.0
casin -0x0.fffffffffffff8p0 0.0
casin 0x0.fffffffffffff8p0 -0.0
casin -0x0.fffffffffffff8p0 -0.0
casin 0x1p-52 0.5
casin 0x1p-52 -0.5
casin -0x1p-52 0.5
casin -0x1p-52 -0.5
casin 0.5 0x1p-52
casin -0.5 0x1p-52
casin 0.5 -0x1p-52
casin -0.5 -0x1p-52
casin 0x1.fp-1025 0.5
casin 0x1.fp-1025 -0.5
casin -0x1.fp-1025 0.5
casin -0x1.fp-1025 -0.5
casin 0.5 0x1.fp-1025
casin -0.5 0x1.fp-1025
casin 0.5 -0x1.fp-1025
casin -0.5 -0x1.fp-1025
casin 0x1p-52 0x1p-52
casin 0x1p-52 -0x1p-52
casin -0x1p-52 0x1p-52
casin -0x1p-52 -0x1p-52
casin 0x1.fp-1025 0x1p-52
casin 0x1.fp-1025 -0x1p-52
casin -0x1.fp-1025 0x1p-52
casin -0x1.fp-1025 -0x1p-52
casin 0x1p-52 0x1.fp-1025
casin -0x1p-52 0x1.fp-1025
casin 0x1p-52 -0x1.fp-1025
casin -0x1p-52 -0x1.fp-1025
casin 0.0 0x1p-52
casin 0.0 -0x1p-52
casin -0.0 0x1p-52
casin -0.0 -0x1p-52
casin 0x1p-52 0.0
casin -0x1p-52 0.0
casin 0x1p-52 -0.0
casin -0x1p-52 -0.0
casin 0x1.fp-1025 0x1.fp-1025
casin 0x1.fp-1025 -0x1.fp-1025
casin -0x1.fp-1025 0x1.fp-1025
casin -0x1.fp-1025 -0x1.fp-1025
casin 0x1p-63 0x0.ffffffffffffffffp0
casin 0x1p-63 -0x0.ffffffffffffffffp0
casin -0x1p-63 0x0.ffffffffffffffffp0
casin -0x1p-63 -0x0.ffffffffffffffffp0
casin 0x0.ffffffffffffffffp0 0x1p-63
casin -0x0.ffffffffffffffffp0 0x1p-63
casin 0x0.ffffffffffffffffp0 -0x1p-63
casin -0x0.ffffffffffffffffp0 -0x1p-63
casin 0x1.fp-16385 0x0.ffffffffffffffffp0
casin 0x1.fp-16385 -0x0.ffffffffffffffffp0
casin -0x1.fp-16385 0x0.ffffffffffffffffp0
casin -0x1.fp-16385 -0x0.ffffffffffffffffp0
casin 0x0.ffffffffffffffffp0 0x1.fp-16385
casin -0x0.ffffffffffffffffp0 0x1.fp-16385
casin 0x0.ffffffffffffffffp0 -0x1.fp-16385
casin -0x0.ffffffffffffffffp0 -0x1.fp-16385
casin 0.0 0x0.ffffffffffffffffp0
casin 0.0 -0x0.ffffffffffffffffp0
casin -0.0 0x0.ffffffffffffffffp0
casin -0.0 -0x0.ffffffffffffffffp0
casin 0x0.ffffffffffffffffp0 0.0
casin -0x0.ffffffffffffffffp0 0.0
casin 0x0.ffffffffffffffffp0 -0.0
casin -0x0.ffffffffffffffffp0 -0.0
casin 0x1p-63 0.5
casin 0x1p-63 -0.5
casin -0x1p-63 0.5
casin -0x1p-63 -0.5
casin 0.5 0x1p-63
casin -0.5 0x1p-63
casin 0.5 -0x1p-63
casin -0.5 -0x1p-63
casin 0x1.fp-16385 0.5
casin 0x1.fp-16385 -0.5
casin -0x1.fp-16385 0.5
casin -0x1.fp-16385 -0.5
casin 0.5 0x1.fp-16385
casin -0.5 0x1.fp-16385
casin 0.5 -0x1.fp-16385
casin -0.5 -0x1.fp-16385
casin 0x1p-63 0x1p-63
casin 0x1p-63 -0x1p-63
casin -0x1p-63 0x1p-63
casin -0x1p-63 -0x1p-63
casin 0x1.fp-16385 0x1p-63
casin 0x1.fp-16385 -0x1p-63
casin -0x1.fp-16385 0x1p-63
casin -0x1.fp-16385 -0x1p-63
casin 0x1p-63 0x1.fp-16385
casin -0x1p-63 0x1.fp-16385
casin 0x1p-63 -0x1.fp-16385
casin -0x1p-63 -0x1.fp-16385
casin 0.0 0x1p-63
casin 0.0 -0x1p-63
casin -0.0 0x1p-63
casin -0.0 -0x1p-63
casin 0x1p-63 0.0
casin -0x1p-63 0.0
casin 0x1p-63 -0.0
casin -0x1p-63 -0.0
casin 0x1.fp-16385 0x1.fp-16385
casin 0x1.fp-16385 -0x1.fp-16385
casin -0x1.fp-16385 0x1.fp-16385
casin -0x1.fp-16385 -0x1.fp-16385
casin 0x1p-105 0x0.ffffffffffffffffffffffffffcp0
casin 0x1p-105 -0x0.ffffffffffffffffffffffffffcp0
casin -0x1p-105 0x0.ffffffffffffffffffffffffffcp0
casin -0x1p-105 -0x0.ffffffffffffffffffffffffffcp0
casin 0x0.ffffffffffffffffffffffffffcp0 0x1p-105
casin -0x0.ffffffffffffffffffffffffffcp0 0x1p-105
casin 0x0.ffffffffffffffffffffffffffcp0 -0x1p-105
casin -0x0.ffffffffffffffffffffffffffcp0 -0x1p-105
casin 0x1.fp-1025 0x0.ffffffffffffffffffffffffffcp0
casin 0x1.fp-1025 -0x0.ffffffffffffffffffffffffffcp0
casin -0x1.fp-1025 0x0.ffffffffffffffffffffffffffcp0
casin -0x1.fp-1025 -0x0.ffffffffffffffffffffffffffcp0
casin 0x0.ffffffffffffffffffffffffffcp0 0x1.fp-1025
casin -0x0.ffffffffffffffffffffffffffcp0 0x1.fp-1025
casin 0x0.ffffffffffffffffffffffffffcp0 -0x1.fp-1025
casin -0x0.ffffffffffffffffffffffffffcp0 -0x1.fp-1025
casin 0.0 0x0.ffffffffffffffffffffffffffcp0
casin 0.0 -0x0.ffffffffffffffffffffffffffcp0
casin -0.0 0x0.ffffffffffffffffffffffffffcp0
casin -0.0 -0x0.ffffffffffffffffffffffffffcp0
casin 0x0.ffffffffffffffffffffffffffcp0 0.0
casin -0x0.ffffffffffffffffffffffffffcp0 0.0
casin 0x0.ffffffffffffffffffffffffffcp0 -0.0
casin -0x0.ffffffffffffffffffffffffffcp0 -0.0
casin 0x1p-105 0.5
casin 0x1p-105 -0.5
casin -0x1p-105 0.5
casin -0x1p-105 -0.5
casin 0.5 0x1p-105
casin -0.5 0x1p-105
casin 0.5 -0x1p-105
casin -0.5 -0x1p-105
casin 0x1p-105 0x1p-105
casin 0x1p-105 -0x1p-105
casin -0x1p-105 0x1p-105
casin -0x1p-105 -0x1p-105
casin 0x1.fp-1025 0x1p-105
casin 0x1.fp-1025 -0x1p-105
casin -0x1.fp-1025 0x1p-105
casin -0x1.fp-1025 -0x1p-105
casin 0x1p-105 0x1.fp-1025
casin -0x1p-105 0x1.fp-1025
casin 0x1p-105 -0x1.fp-1025
casin -0x1p-105 -0x1.fp-1025
casin 0.0 0x1p-105
casin 0.0 -0x1p-105
casin -0.0 0x1p-105
casin -0.0 -0x1p-105
casin 0x1p-105 0.0
casin -0x1p-105 0.0
casin 0x1p-105 -0.0
casin -0x1p-105 -0.0
casin 0x1p-112 0x0.ffffffffffffffffffffffffffff8p0
casin 0x1p-112 -0x0.ffffffffffffffffffffffffffff8p0
casin -0x1p-112 0x0.ffffffffffffffffffffffffffff8p0
casin -0x1p-112 -0x0.ffffffffffffffffffffffffffff8p0
casin 0x0.ffffffffffffffffffffffffffff8p0 0x1p-112
casin -0x0.ffffffffffffffffffffffffffff8p0 0x1p-112
casin 0x0.ffffffffffffffffffffffffffff8p0 -0x1p-112
casin -0x0.ffffffffffffffffffffffffffff8p0 -0x1p-112
casin 0x1.fp-16385 0x0.ffffffffffffffffffffffffffff8p0
casin 0x1.fp-16385 -0x0.ffffffffffffffffffffffffffff8p0
casin -0x1.fp-16385 0x0.ffffffffffffffffffffffffffff8p0
casin -0x1.fp-16385 -0x0.ffffffffffffffffffffffffffff8p0
casin 0x0.ffffffffffffffffffffffffffff8p0 0x1.fp-16385
casin -0x0.ffffffffffffffffffffffffffff8p0 0x1.fp-16385
casin 0x0.ffffffffffffffffffffffffffff8p0 -0x1.fp-16385
casin -0x0.ffffffffffffffffffffffffffff8p0 -0x1.fp-16385
casin 0.0 0x0.ffffffffffffffffffffffffffff8p0
casin 0.0 -0x0.ffffffffffffffffffffffffffff8p0
casin -0.0 0x0.ffffffffffffffffffffffffffff8p0
casin -0.0 -0x0.ffffffffffffffffffffffffffff8p0
casin 0x0.ffffffffffffffffffffffffffff8p0 0.0
casin -0x0.ffffffffffffffffffffffffffff8p0 0.0
casin 0x0.ffffffffffffffffffffffffffff8p0 -0.0
casin -0x0.ffffffffffffffffffffffffffff8p0 -0.0
casin 0x1p-112 0.5
casin 0x1p-112 -0.5
casin -0x1p-112 0.5
casin -0x1p-112 -0.5
casin 0.5 0x1p-112
casin -0.5 0x1p-112
casin 0.5 -0x1p-112
casin -0.5 -0x1p-112
casin 0x1p-112 0x1p-112
casin 0x1p-112 -0x1p-112
casin -0x1p-112 0x1p-112
casin -0x1p-112 -0x1p-112
casin 0x1.fp-16385 0x1p-112
casin 0x1.fp-16385 -0x1p-112
casin -0x1.fp-16385 0x1p-112
casin -0x1.fp-16385 -0x1p-112
casin 0x1p-112 0x1.fp-16385
casin -0x1p-112 0x1.fp-16385
casin 0x1p-112 -0x1.fp-16385
casin -0x1p-112 -0x1.fp-16385
casin 0.0 0x1p-112
casin 0.0 -0x1p-112
casin -0.0 0x1p-112
casin -0.0 -0x1p-112
casin 0x1p-112 0.0
casin -0x1p-112 0.0
casin 0x1p-112 -0.0
casin -0x1p-112 -0.0
casin 0.75 1.25
casin -2 -3
casinh 0 0
casinh -0 0
casinh 0 -0
casinh -0 -0
casinh 0 -1.5
casinh -0 -1.5
casinh 0 -1.0
casinh -0 -1.0
casinh 0 -0.5
casinh -0 -0.5
casinh 0 0.5
casinh -0 0.5
casinh 0 1.0
casinh -0 1.0
casinh 0 1.5
casinh -0 1.5
casinh -1.5 0
casinh -1.5 -0
casinh -1.0 0
casinh -1.0 -0
casinh -0.5 0
casinh -0.5 -0
casinh 0.5 0
casinh 0.5 -0
casinh 1.0 0
casinh 1.0 -0
casinh 1.5 0
casinh 1.5 -0
casinh 0x1p50 1.0
casinh 0x1p50 -1.0
casinh -0x1p50 1.0
casinh -0x1p50 -1.0
casinh 1.0 0x1p50
casinh -1.0 0x1p50
casinh 1.0 -0x1p50
casinh -1.0 -0x1p50
casinh 0x1p500 1.0
casinh 0x1p500 -1.0
casinh -0x1p500 1.0
casinh -0x1p500 -1.0
casinh 1.0 0x1p500
casinh -1.0 0x1p500
casinh 1.0 -0x1p500
casinh -1.0 -0x1p500
casinh 0x1p5000 1.0
casinh 0x1p5000 -1.0
casinh -0x1p5000 1.0
casinh -0x1p5000 -1.0
casinh 1.0 0x1p5000
casinh -1.0 0x1p5000
casinh 1.0 -0x1p5000
casinh -1.0 -0x1p5000
casinh 0x1.fp127 0x1.fp127
casinh 0x1.fp1023 0x1.fp1023
casinh 0x1.fp16383 0x1.fp16383
casinh 0x1.fp-129 1.5
casinh 0x1.fp-129 -1.5
casinh -0x1.fp-129 1.5
casinh -0x1.fp-129 -1.5
casinh 1.5 0x1.fp-129
casinh -1.5 0x1.fp-129
casinh 1.5 -0x1.fp-129
casinh -1.5 -0x1.fp-129
casinh 0x1.fp-1025 1.5
casinh 0x1.fp-1025 -1.5
casinh -0x1.fp-1025 1.5
casinh -0x1.fp-1025 -1.5
casinh 1.5 0x1.fp-1025
casinh -1.5 0x1.fp-1025
casinh 1.5 -0x1.fp-1025
casinh -1.5 -0x1.fp-1025
casinh 0x1.fp-16385 1.5
casinh 0x1.fp-16385 -1.5
casinh -0x1.fp-16385 1.5
casinh -0x1.fp-16385 -1.5
casinh 1.5 0x1.fp-16385
casinh -1.5 0x1.fp-16385
casinh 1.5 -0x1.fp-16385
casinh -1.5 -0x1.fp-16385
casinh 0.5 1.0
casinh 0.5 -1.0
casinh -0.5 1.0
casinh -0.5 -1.0
casinh 1.0 0.5
casinh -1.0 0.5
casinh 1.0 -0.5
casinh -1.0 -0.5
casinh 0.25 1.0
casinh 0.25 -1.0
casinh -0.25 1.0
casinh -0.25 -1.0
casinh 1.0 0.25
casinh -1.0 0.25
casinh 1.0 -0.25
casinh -1.0 -0.25
casinh 0x1.fp-10 1.0
casinh 0x1.fp-10 -1.0
casinh -0x1.fp-10 1.0
casinh -0x1.fp-10 -1.0
casinh 1.0 0x1.fp-10
casinh -1.0 0x1.fp-10
casinh 1.0 -0x1.fp-10
casinh -1.0 -0x1.fp-10
casinh 0x1.fp-30 1.0
casinh 0x1.fp-30 -1.0
casinh -0x1.fp-30 1.0
casinh -0x1.fp-30 -1.0
casinh 1.0 0x1.fp-30
casinh -1.0 0x1.fp-30
casinh 1.0 -0x1.fp-30
casinh -1.0 -0x1.fp-30
casinh 0x1.fp-100 1.0
casinh 0x1.fp-100 -1.0
casinh -0x1.fp-100 1.0
casinh -0x1.fp-100 -1.0
casinh 1.0 0x1.fp-100
casinh -1.0 0x1.fp-100
casinh 1.0 -0x1.fp-100
casinh -1.0 -0x1.fp-100
casinh 0x1.fp-129 1.0
casinh 0x1.fp-129 -1.0
casinh -0x1.fp-129 1.0
casinh -0x1.fp-129 -1.0
casinh 1.0 0x1.fp-129
casinh -1.0 0x1.fp-129
casinh 1.0 -0x1.fp-129
casinh -1.0 -0x1.fp-129
casinh 0x1.fp-1000 1.0
casinh 0x1.fp-1000 -1.0
casinh -0x1.fp-1000 1.0
casinh -0x1.fp-1000 -1.0
casinh 1.0 0x1.fp-1000
casinh -1.0 0x1.fp-1000
casinh 1.0 -0x1.fp-1000
casinh -1.0 -0x1.fp-1000
casinh 0x1.fp-1025 1.0
casinh 0x1.fp-1025 -1.0
casinh -0x1.fp-1025 1.0
casinh -0x1.fp-1025 -1.0
casinh 1.0 0x1.fp-1025
casinh -1.0 0x1.fp-1025
casinh 1.0 -0x1.fp-1025
casinh -1.0 -0x1.fp-1025
casinh 0x1.fp-10000 1.0
casinh 0x1.fp-10000 -1.0
casinh -0x1.fp-10000 1.0
casinh -0x1.fp-10000 -1.0
casinh 1.0 0x1.fp-10000
casinh -1.0 0x1.fp-10000
casinh 1.0 -0x1.fp-10000
casinh -1.0 -0x1.fp-10000
casinh 0x1.fp-16385 1.0
casinh 0x1.fp-16385 -1.0
casinh -0x1.fp-16385 1.0
casinh -0x1.fp-16385 -1.0
casinh 1.0 0x1.fp-16385
casinh -1.0 0x1.fp-16385
casinh 1.0 -0x1.fp-16385
casinh -1.0 -0x1.fp-16385
casinh 0x1p-23 0x1.000002p0
casinh 0x1p-23 -0x1.000002p0
casinh -0x1p-23 0x1.000002p0
casinh -0x1p-23 -0x1.000002p0
casinh 0x1.000002p0 0x1p-23
casinh -0x1.000002p0 0x1p-23
casinh 0x1.000002p0 -0x1p-23
casinh -0x1.000002p0 -0x1p-23
casinh 0x1.fp-129 0x1.000002p0
casinh 0x1.fp-129 -0x1.000002p0
casinh -0x1.fp-129 0x1.000002p0
casinh -0x1.fp-129 -0x1.000002p0
casinh 0x1.000002p0 0x1.fp-129
casinh -0x1.000002p0 0x1.fp-129
casinh 0x1.000002p0 -0x1.fp-129
casinh -0x1.000002p0 -0x1.fp-129
casinh 0.0 0x1.000002p0
casinh 0.0 -0x1.000002p0
casinh -0.0 0x1.000002p0
casinh -0.0 -0x1.000002p0
casinh 0x1.000002p0 0.0
casinh -0x1.000002p0 0.0
casinh 0x1.000002p0 -0.0
casinh -0x1.000002p0 -0.0
casinh 0x1p-52 0x1.0000000000001p0
casinh 0x1p-52 -0x1.0000000000001p0
casinh -0x1p-52 0x1.0000000000001p0
casinh -0x1p-52 -0x1.0000000000001p0
casinh 0x1.0000000000001p0 0x1p-52
casinh -0x1.0000000000001p0 0x1p-52
casinh 0x1.0000000000001p0 -0x1p-52
casinh -0x1.0000000000001p0 -0x1p-52
casinh 0x1.fp-1025 0x1.0000000000001p0
casinh 0x1.fp-1025 -0x1.0000000000001p0
casinh -0x1.fp-1025 0x1.0000000000001p0
casinh -0x1.fp-1025 -0x1.0000000000001p0
casinh 0x1.0000000000001p0 0x1.fp-1025
casinh -0x1.0000000000001p0 0x1.fp-1025
casinh 0x1.0000000000001p0 -0x1.fp-1025
casinh -0x1.0000000000001p0 -0x1.fp-1025
casinh 0.0 0x1.0000000000001p0
casinh 0.0 -0x1.0000000000001p0
casinh -0.0 0x1.0000000000001p0
casinh -0.0 -0x1.0000000000001p0
casinh 0x1.0000000000001p0 0.0
casinh -0x1.0000000000001p0 0.0
casinh 0x1.0000000000001p0 -0.0
casinh -0x1.0000000000001p0 -0.0
casinh 0x1p-63 0x1.0000000000000002p0
casinh 0x1p-63 -0x1.0000000000000002p0
casinh -0x1p-63 0x1.0000000000000002p0
casinh -0x1p-63 -0x1.0000000000000002p0
casinh 0x1.0000000000000002p0 0x1p-63
casinh -0x1.0000000000000002p0 0x1p-63
casinh 0x1.0000000000000002p0 -0x1p-63
casinh -0x1.0000000000000002p0 -0x1p-63
casinh 0x1.fp-16385 0x1.0000000000000002p0
casinh 0x1.fp-16385 -0x1.0000000000000002p0
casinh -0x1.fp-16385 0x1.0000000000000002p0
casinh -0x1.fp-16385 -0x1.0000000000000002p0
casinh 0x1.0000000000000002p0 0x1.fp-16385
casinh -0x1.0000000000000002p0 0x1.fp-16385
casinh 0x1.0000000000000002p0 -0x1.fp-16385
casinh -0x1.0000000000000002p0 -0x1.fp-16385
casinh 0.0 0x1.0000000000000002p0
casinh 0.0 -0x1.0000000000000002p0
casinh -0.0 0x1.0000000000000002p0
casinh -0.0 -0x1.0000000000000002p0
casinh 0x1.0000000000000002p0 0.0
casinh -0x1.0000000000000002p0 0.0
casinh 0x1.0000000000000002p0 -0.0
casinh -0x1.0000000000000002p0 -0.0
casinh 0x1p-106 0x1.000000000000000000000000008p0
casinh 0x1p-106 -0x1.000000000000000000000000008p0
casinh -0x1p-106 0x1.000000000000000000000000008p0
casinh -0x1p-106 -0x1.000000000000000000000000008p0
casinh 0x1.000000000000000000000000008p0 0x1p-106
casinh -0x1.000000000000000000000000008p0 0x1p-106
casinh 0x1.000000000000000000000000008p0 -0x1p-106
casinh -0x1.000000000000000000000000008p0 -0x1p-106
casinh 0x1.fp-1025 0x1.000000000000000000000000008p0
casinh 0x1.fp-1025 -0x1.000000000000000000000000008p0
casinh -0x1.fp-1025 0x1.000000000000000000000000008p0
casinh -0x1.fp-1025 -0x1.000000000000000000000000008p0
casinh 0x1.000000000000000000000000008p0 0x1.fp-1025
casinh -0x1.000000000000000000000000008p0 0x1.fp-1025
casinh 0x1.000000000000000000000000008p0 -0x1.fp-1025
casinh -0x1.000000000000000000000000008p0 -0x1.fp-1025
casinh 0.0 0x1.000000000000000000000000008p0
casinh 0.0 -0x1.000000000000000000000000008p0
casinh -0.0 0x1.000000000000000000000000008p0
casinh -0.0 -0x1.000000000000000000000000008p0
casinh 0x1.000000000000000000000000008p0 0.0
casinh -0x1.000000000000000000000000008p0 0.0
casinh 0x1.000000000000000000000000008p0 -0.0
casinh -0x1.000000000000000000000000008p0 -0.0
casinh 0x1p-113 0x1.0000000000000000000000000001p0
casinh 0x1p-113 -0x1.0000000000000000000000000001p0
casinh -0x1p-113 0x1.0000000000000000000000000001p0
casinh -0x1p-113 -0x1.0000000000000000000000000001p0
casinh 0x1.0000000000000000000000000001p0 0x1p-113
casinh -0x1.0000000000000000000000000001p0 0x1p-113
casinh 0x1.0000000000000000000000000001p0 -0x1p-113
casinh -0x1.0000000000000000000000000001p0 -0x1p-113
casinh 0x1.fp-16385 0x1.0000000000000000000000000001p0
casinh 0x1.fp-16385 -0x1.0000000000000000000000000001p0
casinh -0x1.fp-16385 0x1.0000000000000000000000000001p0
casinh -0x1.fp-16385 -0x1.0000000000000000000000000001p0
casinh 0x1.0000000000000000000000000001p0 0x1.fp-16385
casinh -0x1.0000000000000000000000000001p0 0x1.fp-16385
casinh 0x1.0000000000000000000000000001p0 -0x1.fp-16385
casinh -0x1.0000000000000000000000000001p0 -0x1.fp-16385
casinh 0.0 0x1.0000000000000000000000000001p0
casinh 0.0 -0x1.0000000000000000000000000001p0
casinh -0.0 0x1.0000000000000000000000000001p0
casinh -0.0 -0x1.0000000000000000000000000001p0
casinh 0x1.0000000000000000000000000001p0 0.0
casinh -0x1.0000000000000000000000000001p0 0.0
casinh 0x1.0000000000000000000000000001p0 -0.0
casinh -0x1.0000000000000000000000000001p0 -0.0
casinh 0x1p-23 0x0.ffffffp0
casinh 0x1p-23 -0x0.ffffffp0
casinh -0x1p-23 0x0.ffffffp0
casinh -0x1p-23 -0x0.ffffffp0
casinh 0x0.ffffffp0 0x1p-23
casinh -0x0.ffffffp0 0x1p-23
casinh 0x0.ffffffp0 -0x1p-23
casinh -0x0.ffffffp0 -0x1p-23
casinh 0x1.fp-129 0x0.ffffffp0
casinh 0x1.fp-129 -0x0.ffffffp0
casinh -0x1.fp-129 0x0.ffffffp0
casinh -0x1.fp-129 -0x0.ffffffp0
casinh 0x0.ffffffp0 0x1.fp-129
casinh -0x0.ffffffp0 0x1.fp-129
casinh 0x0.ffffffp0 -0x1.fp-129
casinh -0x0.ffffffp0 -0x1.fp-129
casinh 0.0 0x0.ffffffp0
casinh 0.0 -0x0.ffffffp0
casinh -0.0 0x0.ffffffp0
casinh -0.0 -0x0.ffffffp0
casinh 0x0.ffffffp0 0.0
casinh -0x0.ffffffp0 0.0
casinh 0x0.ffffffp0 -0.0
casinh -0x0.ffffffp0 -0.0
casinh 0x1p-23 0.5
casinh 0x1p-23 -0.5
casinh -0x1p-23 0.5
casinh -0x1p-23 -0.5
casinh 0.5 0x1p-23
casinh -0.5 0x1p-23
casinh 0.5 -0x1p-23
casinh -0.5 -0x1p-23
casinh 0x1.fp-129 0.5
casinh 0x1.fp-129 -0.5
casinh -0x1.fp-129 0.5
casinh -0x1.fp-129 -0.5
casinh 0.5 0x1.fp-129
casinh -0.5 0x1.fp-129
casinh 0.5 -0x1.fp-129
casinh -0.5 -0x1.fp-129
casinh 0x1p-23 0x1p-23
casinh 0x1p-23 -0x1p-23
casinh -0x1p-23 0x1p-23
casinh -0x1p-23 -0x1p-23
casinh 0x1.fp-129 0x1p-23
casinh 0x1.fp-129 -0x1p-23
casinh -0x1.fp-129 0x1p-23
casinh -0x1.fp-129 -0x1p-23
casinh 0x1p-23 0x1.fp-129
casinh -0x1p-23 0x1.fp-129
casinh 0x1p-23 -0x1.fp-129
casinh -0x1p-23 -0x1.fp-129
casinh 0.0 0x1p-23
casinh 0.0 -0x1p-23
casinh -0.0 0x1p-23
casinh -0.0 -0x1p-23
casinh 0x1p-23 0.0
casinh -0x1p-23 0.0
casinh 0x1p-23 -0.0
casinh -0x1p-23 -0.0
casinh 0x1.fp-129 0x1.fp-129
casinh 0x1.fp-129 -0x1.fp-129
casinh -0x1.fp-129 0x1.fp-129
casinh -0x1.fp-129 -0x1.fp-129
casinh 0x1p-52 0x0.fffffffffffff8p0
casinh 0x1p-52 -0x0.fffffffffffff8p0
casinh -0x1p-52 0x0.fffffffffffff8p0
casinh -0x1p-52 -0x0.fffffffffffff8p0
casinh 0x0.fffffffffffff8p0 0x1p-52
casinh -0x0.fffffffffffff8p0 0x1p-52
casinh 0x0.fffffffffffff8p0 -0x1p-52
casinh -0x0.fffffffffffff8p0 -0x1p-52
casinh 0x1.fp-1025 0x0.fffffffffffff8p0
casinh 0x1.fp-1025 -0x0.fffffffffffff8p0
casinh -0x1.fp-1025 0x0.fffffffffffff8p0
casinh -0x1.fp-1025 -0x0.fffffffffffff8p0
casinh 0x0.fffffffffffff8p0 0x1.fp-1025
casinh -0x0.fffffffffffff8p0 0x1.fp-1025
casinh 0x0.fffffffffffff8p0 -0x1.fp-1025
casinh -0x0.fffffffffffff8p0 -0x1.fp-1025
casinh 0.0 0x0.fffffffffffff8p0
casinh 0.0 -0x0.fffffffffffff8p0
casinh -0.0 0x0.fffffffffffff8p0
casinh -0.0 -0x0.fffffffffffff8p0
casinh 0x0.fffffffffffff8p0 0.0
casinh -0x0.fffffffffffff8p0 0.0
casinh 0x0.fffffffffffff8p0 -0.0
casinh -0x0.fffffffffffff8p0 -0.0
casinh 0x1p-52 0.5
casinh 0x1p-52 -0.5
casinh -0x1p-52 0.5
casinh -0x1p-52 -0.5
casinh 0.5 0x1p-52
casinh -0.5 0x1p-52
casinh 0.5 -0x1p-52
casinh -0.5 -0x1p-52
casinh 0x1.fp-1025 0.5
casinh 0x1.fp-1025 -0.5
casinh -0x1.fp-1025 0.5
casinh -0x1.fp-1025 -0.5
casinh 0.5 0x1.fp-1025
casinh -0.5 0x1.fp-1025
casinh 0.5 -0x1.fp-1025
casinh -0.5 -0x1.fp-1025
casinh 0x1p-52 0x1p-52
casinh 0x1p-52 -0x1p-52
casinh -0x1p-52 0x1p-52
casinh -0x1p-52 -0x1p-52
casinh 0x1.fp-1025 0x1p-52
casinh 0x1.fp-1025 -0x1p-52
casinh -0x1.fp-1025 0x1p-52
casinh -0x1.fp-1025 -0x1p-52
casinh 0x1p-52 0x1.fp-1025
casinh -0x1p-52 0x1.fp-1025
casinh 0x1p-52 -0x1.fp-1025
casinh -0x1p-52 -0x1.fp-1025
casinh 0.0 0x1p-52
casinh 0.0 -0x1p-52
casinh -0.0 0x1p-52
casinh -0.0 -0x1p-52
casinh 0x1p-52 0.0
casinh -0x1p-52 0.0
casinh 0x1p-52 -0.0
casinh -0x1p-52 -0.0
casinh 0x1.fp-1025 0x1.fp-1025
casinh 0x1.fp-1025 -0x1.fp-1025
casinh -0x1.fp-1025 0x1.fp-1025
casinh -0x1.fp-1025 -0x1.fp-1025
casinh 0x1p-63 0x0.ffffffffffffffffp0
casinh 0x1p-63 -0x0.ffffffffffffffffp0
casinh -0x1p-63 0x0.ffffffffffffffffp0
casinh -0x1p-63 -0x0.ffffffffffffffffp0
casinh 0x0.ffffffffffffffffp0 0x1p-63
casinh -0x0.ffffffffffffffffp0 0x1p-63
casinh 0x0.ffffffffffffffffp0 -0x1p-63
casinh -0x0.ffffffffffffffffp0 -0x1p-63
casinh 0x1.fp-16385 0x0.ffffffffffffffffp0
casinh 0x1.fp-16385 -0x0.ffffffffffffffffp0
casinh -0x1.fp-16385 0x0.ffffffffffffffffp0
casinh -0x1.fp-16385 -0x0.ffffffffffffffffp0
casinh 0x0.ffffffffffffffffp0 0x1.fp-16385
casinh -0x0.ffffffffffffffffp0 0x1.fp-16385
casinh 0x0.ffffffffffffffffp0 -0x1.fp-16385
casinh -0x0.ffffffffffffffffp0 -0x1.fp-16385
casinh 0.0 0x0.ffffffffffffffffp0
casinh 0.0 -0x0.ffffffffffffffffp0
casinh -0.0 0x0.ffffffffffffffffp0
casinh -0.0 -0x0.ffffffffffffffffp0
casinh 0x0.ffffffffffffffffp0 0.0
casinh -0x0.ffffffffffffffffp0 0.0
casinh 0x0.ffffffffffffffffp0 -0.0
casinh -0x0.ffffffffffffffffp0 -0.0
casinh 0x1p-63 0.5
casinh 0x1p-63 -0.5
casinh -0x1p-63 0.5
casinh -0x1p-63 -0.5
casinh 0.5 0x1p-63
casinh -0.5 0x1p-63
casinh 0.5 -0x1p-63
casinh -0.5 -0x1p-63
casinh 0x1.fp-16385 0.5
casinh 0x1.fp-16385 -0.5
casinh -0x1.fp-16385 0.5
casinh -0x1.fp-16385 -0.5
casinh 0.5 0x1.fp-16385
casinh -0.5 0x1.fp-16385
casinh 0.5 -0x1.fp-16385
casinh -0.5 -0x1.fp-16385
casinh 0x1p-63 0x1p-63
casinh 0x1p-63 -0x1p-63
casinh -0x1p-63 0x1p-63
casinh -0x1p-63 -0x1p-63
casinh 0x1.fp-16385 0x1p-63
casinh 0x1.fp-16385 -0x1p-63
casinh -0x1.fp-16385 0x1p-63
casinh -0x1.fp-16385 -0x1p-63
casinh 0x1p-63 0x1.fp-16385
casinh -0x1p-63 0x1.fp-16385
casinh 0x1p-63 -0x1.fp-16385
casinh -0x1p-63 -0x1.fp-16385
casinh 0.0 0x1p-63
casinh 0.0 -0x1p-63
casinh -0.0 0x1p-63
casinh -0.0 -0x1p-63
casinh 0x1p-63 0.0
casinh -0x1p-63 0.0
casinh 0x1p-63 -0.0
casinh -0x1p-63 -0.0
casinh 0x1.fp-16385 0x1.fp-16385
casinh 0x1.fp-16385 -0x1.fp-16385
casinh -0x1.fp-16385 0x1.fp-16385
casinh -0x1.fp-16385 -0x1.fp-16385
casinh 0x1p-105 0x0.ffffffffffffffffffffffffffcp0
casinh 0x1p-105 -0x0.ffffffffffffffffffffffffffcp0
casinh -0x1p-105 0x0.ffffffffffffffffffffffffffcp0
casinh -0x1p-105 -0x0.ffffffffffffffffffffffffffcp0
casinh 0x0.ffffffffffffffffffffffffffcp0 0x1p-105
casinh -0x0.ffffffffffffffffffffffffffcp0 0x1p-105
casinh 0x0.ffffffffffffffffffffffffffcp0 -0x1p-105
casinh -0x0.ffffffffffffffffffffffffffcp0 -0x1p-105
casinh 0x1.fp-1025 0x0.ffffffffffffffffffffffffffcp0
casinh 0x1.fp-1025 -0x0.ffffffffffffffffffffffffffcp0
casinh -0x1.fp-1025 0x0.ffffffffffffffffffffffffffcp0
casinh -0x1.fp-1025 -0x0.ffffffffffffffffffffffffffcp0
casinh 0x0.ffffffffffffffffffffffffffcp0 0x1.fp-1025
casinh -0x0.ffffffffffffffffffffffffffcp0 0x1.fp-1025
casinh 0x0.ffffffffffffffffffffffffffcp0 -0x1.fp-1025
casinh -0x0.ffffffffffffffffffffffffffcp0 -0x1.fp-1025
casinh 0.0 0x0.ffffffffffffffffffffffffffcp0
casinh 0.0 -0x0.ffffffffffffffffffffffffffcp0
casinh -0.0 0x0.ffffffffffffffffffffffffffcp0
casinh -0.0 -0x0.ffffffffffffffffffffffffffcp0
casinh 0x0.ffffffffffffffffffffffffffcp0 0.0
casinh -0x0.ffffffffffffffffffffffffffcp0 0.0
casinh 0x0.ffffffffffffffffffffffffffcp0 -0.0
casinh -0x0.ffffffffffffffffffffffffffcp0 -0.0
casinh 0x1p-105 0.5
casinh 0x1p-105 -0.5
casinh -0x1p-105 0.5
casinh -0x1p-105 -0.5
casinh 0.5 0x1p-105
casinh -0.5 0x1p-105
casinh 0.5 -0x1p-105
casinh -0.5 -0x1p-105
casinh 0x1p-105 0x1p-105
casinh 0x1p-105 -0x1p-105
casinh -0x1p-105 0x1p-105
casinh -0x1p-105 -0x1p-105
casinh 0x1.fp-1025 0x1p-105
casinh 0x1.fp-1025 -0x1p-105
casinh -0x1.fp-1025 0x1p-105
casinh -0x1.fp-1025 -0x1p-105
casinh 0x1p-105 0x1.fp-1025
casinh -0x1p-105 0x1.fp-1025
casinh 0x1p-105 -0x1.fp-1025
casinh -0x1p-105 -0x1.fp-1025
casinh 0.0 0x1p-105
casinh 0.0 -0x1p-105
casinh -0.0 0x1p-105
casinh -0.0 -0x1p-105
casinh 0x1p-105 0.0
casinh -0x1p-105 0.0
casinh 0x1p-105 -0.0
casinh -0x1p-105 -0.0
casinh 0x1p-112 0x0.ffffffffffffffffffffffffffff8p0
casinh 0x1p-112 -0x0.ffffffffffffffffffffffffffff8p0
casinh -0x1p-112 0x0.ffffffffffffffffffffffffffff8p0
casinh -0x1p-112 -0x0.ffffffffffffffffffffffffffff8p0
casinh 0x0.ffffffffffffffffffffffffffff8p0 0x1p-112
casinh -0x0.ffffffffffffffffffffffffffff8p0 0x1p-112
casinh 0x0.ffffffffffffffffffffffffffff8p0 -0x1p-112
casinh -0x0.ffffffffffffffffffffffffffff8p0 -0x1p-112
casinh 0x1.fp-16385 0x0.ffffffffffffffffffffffffffff8p0
casinh 0x1.fp-16385 -0x0.ffffffffffffffffffffffffffff8p0
casinh -0x1.fp-16385 0x0.ffffffffffffffffffffffffffff8p0
casinh -0x1.fp-16385 -0x0.ffffffffffffffffffffffffffff8p0
casinh 0x0.ffffffffffffffffffffffffffff8p0 0x1.fp-16385
casinh -0x0.ffffffffffffffffffffffffffff8p0 0x1.fp-16385
casinh 0x0.ffffffffffffffffffffffffffff8p0 -0x1.fp-16385
casinh -0x0.ffffffffffffffffffffffffffff8p0 -0x1.fp-16385
casinh 0.0 0x0.ffffffffffffffffffffffffffff8p0
casinh 0.0 -0x0.ffffffffffffffffffffffffffff8p0
casinh -0.0 0x0.ffffffffffffffffffffffffffff8p0
casinh -0.0 -0x0.ffffffffffffffffffffffffffff8p0
casinh 0x0.ffffffffffffffffffffffffffff8p0 0.0
casinh -0x0.ffffffffffffffffffffffffffff8p0 0.0
casinh 0x0.ffffffffffffffffffffffffffff8p0 -0.0
casinh -0x0.ffffffffffffffffffffffffffff8p0 -0.0
casinh 0x1p-112 0.5
casinh 0x1p-112 -0.5
casinh -0x1p-112 0.5
casinh -0x1p-112 -0.5
casinh 0.5 0x1p-112
casinh -0.5 0x1p-112
casinh 0.5 -0x1p-112
casinh -0.5 -0x1p-112
casinh 0x1p-112 0x1p-112
casinh 0x1p-112 -0x1p-112
casinh -0x1p-112 0x1p-112
casinh -0x1p-112 -0x1p-112
casinh 0x1.fp-16385 0x1p-112
casinh 0x1.fp-16385 -0x1p-112
casinh -0x1.fp-16385 0x1p-112
casinh -0x1.fp-16385 -0x1p-112
casinh 0x1p-112 0x1.fp-16385
casinh -0x1p-112 0x1.fp-16385
casinh 0x1p-112 -0x1.fp-16385
casinh -0x1p-112 -0x1.fp-16385
casinh 0.0 0x1p-112
casinh 0.0 -0x1p-112
casinh -0.0 0x1p-112
casinh -0.0 -0x1p-112
casinh 0x1p-112 0.0
casinh -0x1p-112 0.0
casinh 0x1p-112 -0.0
casinh -0x1p-112 -0.0
casinh 0.75 1.25
casinh -2 -3
Move tests of catan, catanh to auto-libm-test-*. This patch moves tests of catan and catanh with finite inputs (other than the divide-by-zero cases producing an exact infinity) to using the auto-libm-test machinery. Each of auto-libm-test-out-catan and auto-libm-test-out-catanh takes about three seconds to generate on my system (so in fact it wasn't necessary after all to defer the move to auto-libm-test-* until the output files were split up by function). Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of catan and catanh. * math/auto-libm-test-out-catan: New generated file. * math/auto-libm-test-out-catanh: Likewise. * math/libm-test-catan.inc (catan_test_data): Use AUTO_TESTS_c_c. Move tests with finite inputs, except divide-by-zero cases, to auto-libm-test-in. * math/libm-test-catanh.inc (catanh_test_data): Likewise. * math/Makefile (libm-test-funcs-auto): Add catan and catanh. (libm-test-funcs-noauto): Remove catan and catanh. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2017-02-17 18:42:37 +00:00
catan 0 0
catan -0 0
catan 0 -0
catan -0 -0
catan 1.0 0
catan 1.0 -0
catan -1.0 0
catan -1.0 -0
catan 0x1p50 0.0
catan 0x1p50 -0.0
catan -0x1p50 0.0
catan -0x1p50 -0.0
catan 0.0 0x1p50
catan -0.0 0x1p50
catan 0.0 -0x1p50
catan -0.0 -0x1p50
catan 0x1p500 0.0
catan 0x1p500 -0.0
catan -0x1p500 0.0
catan -0x1p500 -0.0
catan 0.0 0x1p500
catan -0.0 0x1p500
catan 0.0 -0x1p500
catan -0.0 -0x1p500
catan 0x1p5000 0.0
catan 0x1p5000 -0.0
catan -0x1p5000 0.0
catan -0x1p5000 -0.0
catan 0.0 0x1p5000
catan -0.0 0x1p5000
catan 0.0 -0x1p5000
catan -0.0 -0x1p5000
catan 0x1p63 0.5
catan 0x1p63 -0.5
catan -0x1p63 0.5
catan -0x1p63 -0.5
catan 0.5 0x1p63
catan -0.5 0x1p63
catan 0.5 -0x1p63
catan -0.5 -0x1p63
catan 0x1p511 0.5
catan 0x1p511 -0.5
catan -0x1p511 0.5
catan -0x1p511 -0.5
catan 0.5 0x1p511
catan -0.5 0x1p511
catan 0.5 -0x1p511
catan -0.5 -0x1p511
catan 0x1p8191 0.5
catan 0x1p8191 -0.5
catan -0x1p8191 0.5
catan -0x1p8191 -0.5
catan 0.5 0x1p8191
catan -0.5 0x1p8191
catan 0.5 -0x1p8191
catan -0.5 -0x1p8191
catan 0x1p100 0.0
catan 0x1p100 -0.0
catan -0x1p100 0.0
catan -0x1p100 -0.0
catan 0.0 0x1p100
catan -0.0 0x1p100
catan 0.0 -0x1p100
catan -0.0 -0x1p100
catan 0x1.fp127 0.0
catan 0x1.fp127 -0.0
catan -0x1.fp127 0.0
catan -0x1.fp127 -0.0
catan 0.0 0x1.fp127
catan -0.0 0x1.fp127
catan 0.0 -0x1.fp127
catan -0.0 -0x1.fp127
catan 0x1.fp127 0x1.fp127
catan 0x1.fp127 -0x1.fp127
catan -0x1.fp127 0x1.fp127
catan -0x1.fp127 -0x1.fp127
catan 0x1p900 0.0
catan 0x1p900 -0.0
catan -0x1p900 0.0
catan -0x1p900 -0.0
catan 0.0 0x1p900
catan -0.0 0x1p900
catan 0.0 -0x1p900
catan -0.0 -0x1p900
catan 0x1.fp1023 0.0
catan 0x1.fp1023 -0.0
catan -0x1.fp1023 0.0
catan -0x1.fp1023 -0.0
catan 0.0 0x1.fp1023
catan -0.0 0x1.fp1023
catan 0.0 -0x1.fp1023
catan -0.0 -0x1.fp1023
catan 0x1.fp1023 0x1.fp1023
catan 0x1.fp1023 -0x1.fp1023
catan -0x1.fp1023 0x1.fp1023
catan -0x1.fp1023 -0x1.fp1023
catan 0x1p10000 0.0
catan 0x1p10000 -0.0
catan -0x1p10000 0.0
catan -0x1p10000 -0.0
catan 0.0 0x1p10000
catan -0.0 0x1p10000
catan 0.0 -0x1p10000
catan -0.0 -0x1p10000
catan 0x1.fp16383 0.0
catan 0x1.fp16383 -0.0
catan -0x1.fp16383 0.0
catan -0x1.fp16383 -0.0
catan 0.0 0x1.fp16383
catan -0.0 0x1.fp16383
catan 0.0 -0x1.fp16383
catan -0.0 -0x1.fp16383
XFAIL catan and catanh tests on ibm128 Mark 4 catan and catanh tests as xfail-rounding:ibm128-libgcc. After this patch all catan and catanh tests pass on ibm128. Regenerated auto-libm-test-out using gmp 6.1.2, mpfr 3.1.5 and upstream mpc (9ef8030e50), Tested on powerpc, powerpc64 and powerpc64le. * math/auto-libm-test-in: Mark some catan and catanh as xfail-rounding:ibm128-libgcc. * math/auto-libm-test-out-catan: Regenerate. * math/auto-libm-test-out-catanh: Likewise.
2017-03-16 13:18:46 +00:00
catan 0x1.fp16383 0x1.fp16383 xfail-rounding:ibm128-libgcc
catan 0x1.fp16383 -0x1.fp16383 xfail-rounding:ibm128-libgcc
catan -0x1.fp16383 0x1.fp16383 xfail-rounding:ibm128-libgcc
catan -0x1.fp16383 -0x1.fp16383 xfail-rounding:ibm128-libgcc
Move tests of catan, catanh to auto-libm-test-*. This patch moves tests of catan and catanh with finite inputs (other than the divide-by-zero cases producing an exact infinity) to using the auto-libm-test machinery. Each of auto-libm-test-out-catan and auto-libm-test-out-catanh takes about three seconds to generate on my system (so in fact it wasn't necessary after all to defer the move to auto-libm-test-* until the output files were split up by function). Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of catan and catanh. * math/auto-libm-test-out-catan: New generated file. * math/auto-libm-test-out-catanh: Likewise. * math/libm-test-catan.inc (catan_test_data): Use AUTO_TESTS_c_c. Move tests with finite inputs, except divide-by-zero cases, to auto-libm-test-in. * math/libm-test-catanh.inc (catanh_test_data): Likewise. * math/Makefile (libm-test-funcs-auto): Add catan and catanh. (libm-test-funcs-noauto): Remove catan and catanh. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2017-02-17 18:42:37 +00:00
catan 0x1p-13 1.0
catan 0x1p-13 -1.0
catan -0x1p-13 1.0
catan -0x1p-13 -1.0
catan 1.0 0x1p-13
catan -1.0 0x1p-13
catan 1.0 -0x1p-13
catan -1.0 -0x1p-13
catan 0x1p-27 1.0
catan 0x1p-27 -1.0
catan -0x1p-27 1.0
catan -0x1p-27 -1.0
catan 1.0 0x1p-27
catan -1.0 0x1p-27
catan 1.0 -0x1p-27
catan -1.0 -0x1p-27
catan 0x1p-33 1.0
catan 0x1p-33 -1.0
catan -0x1p-33 1.0
catan -0x1p-33 -1.0
catan 1.0 0x1p-33
catan -1.0 0x1p-33
catan 1.0 -0x1p-33
catan -1.0 -0x1p-33
catan 0x1p-54 1.0
catan 0x1p-54 -1.0
catan -0x1p-54 1.0
catan -0x1p-54 -1.0
catan 1.0 0x1p-54
catan -1.0 0x1p-54
catan 1.0 -0x1p-54
catan -1.0 -0x1p-54
catan 0x1p-57 1.0
catan 0x1p-57 -1.0
catan -0x1p-57 1.0
catan -0x1p-57 -1.0
catan 1.0 0x1p-57
catan -1.0 0x1p-57
catan 1.0 -0x1p-57
catan -1.0 -0x1p-57
catan 0x1p-13 0x1.000002p0
catan 0x1p-13 -0x1.000002p0
catan -0x1p-13 0x1.000002p0
catan -0x1p-13 -0x1.000002p0
catan 0x1.000002p0 0x1p-13
catan -0x1.000002p0 0x1p-13
catan 0x1.000002p0 -0x1p-13
catan -0x1.000002p0 -0x1p-13
catan 0x1p-13 0x0.ffffffp0
catan 0x1p-13 -0x0.ffffffp0
catan -0x1p-13 0x0.ffffffp0
catan -0x1p-13 -0x0.ffffffp0
catan 0x0.ffffffp0 0x1p-13
catan -0x0.ffffffp0 0x1p-13
catan 0x0.ffffffp0 -0x1p-13
catan -0x0.ffffffp0 -0x1p-13
catan 0x1p-27 0x1.0000000000001p0
catan 0x1p-27 -0x1.0000000000001p0
catan -0x1p-27 0x1.0000000000001p0
catan -0x1p-27 -0x1.0000000000001p0
catan 0x1.0000000000001p0 0x1p-27
catan -0x1.0000000000001p0 0x1p-27
catan 0x1.0000000000001p0 -0x1p-27
catan -0x1.0000000000001p0 -0x1p-27
catan 0x1p-27 0x0.fffffffffffff8p0
catan 0x1p-27 -0x0.fffffffffffff8p0
catan -0x1p-27 0x0.fffffffffffff8p0
catan -0x1p-27 -0x0.fffffffffffff8p0
catan 0x0.fffffffffffff8p0 0x1p-27
catan -0x0.fffffffffffff8p0 0x1p-27
catan 0x0.fffffffffffff8p0 -0x1p-27
catan -0x0.fffffffffffff8p0 -0x1p-27
catan 0x1p-33 0x1.0000000000000002p0
catan 0x1p-33 -0x1.0000000000000002p0
catan -0x1p-33 0x1.0000000000000002p0
catan -0x1p-33 -0x1.0000000000000002p0
catan 0x1.0000000000000002p0 0x1p-33
catan -0x1.0000000000000002p0 0x1p-33
catan 0x1.0000000000000002p0 -0x1p-33
catan -0x1.0000000000000002p0 -0x1p-33
catan 0x1p-33 0x0.ffffffffffffffffp0
catan 0x1p-33 -0x0.ffffffffffffffffp0
catan -0x1p-33 0x0.ffffffffffffffffp0
catan -0x1p-33 -0x0.ffffffffffffffffp0
catan 0x0.ffffffffffffffffp0 0x1p-33
catan -0x0.ffffffffffffffffp0 0x1p-33
catan 0x0.ffffffffffffffffp0 -0x1p-33
catan -0x0.ffffffffffffffffp0 -0x1p-33
catan 0x1p-54 0x1.000000000000000000000000008p0
catan 0x1p-54 -0x1.000000000000000000000000008p0
catan -0x1p-54 0x1.000000000000000000000000008p0
catan -0x1p-54 -0x1.000000000000000000000000008p0
catan 0x1.000000000000000000000000008p0 0x1p-54
catan -0x1.000000000000000000000000008p0 0x1p-54
catan 0x1.000000000000000000000000008p0 -0x1p-54
catan -0x1.000000000000000000000000008p0 -0x1p-54
catan 0x1p-54 0x0.ffffffffffffffffffffffffffcp0
catan 0x1p-54 -0x0.ffffffffffffffffffffffffffcp0
catan -0x1p-54 0x0.ffffffffffffffffffffffffffcp0
catan -0x1p-54 -0x0.ffffffffffffffffffffffffffcp0
catan 0x0.ffffffffffffffffffffffffffcp0 0x1p-54
catan -0x0.ffffffffffffffffffffffffffcp0 0x1p-54
catan 0x0.ffffffffffffffffffffffffffcp0 -0x1p-54
catan -0x0.ffffffffffffffffffffffffffcp0 -0x1p-54
catan 0x1p-57 0x1.0000000000000000000000000001p0
catan 0x1p-57 -0x1.0000000000000000000000000001p0
catan -0x1p-57 0x1.0000000000000000000000000001p0
catan -0x1p-57 -0x1.0000000000000000000000000001p0
catan 0x1.0000000000000000000000000001p0 0x1p-57
catan -0x1.0000000000000000000000000001p0 0x1p-57
catan 0x1.0000000000000000000000000001p0 -0x1p-57
catan -0x1.0000000000000000000000000001p0 -0x1p-57
catan 0x1p-57 0x0.ffffffffffffffffffffffffffff8p0
catan 0x1p-57 -0x0.ffffffffffffffffffffffffffff8p0
catan -0x1p-57 0x0.ffffffffffffffffffffffffffff8p0
catan -0x1p-57 -0x0.ffffffffffffffffffffffffffff8p0
catan 0x0.ffffffffffffffffffffffffffff8p0 0x1p-57
catan -0x0.ffffffffffffffffffffffffffff8p0 0x1p-57
catan 0x0.ffffffffffffffffffffffffffff8p0 -0x1p-57
catan -0x0.ffffffffffffffffffffffffffff8p0 -0x1p-57
catan 1.0 0x1p-64
catan 1.0 -0x1p-64
catan -1.0 0x1p-64
catan -1.0 -0x1p-64
catan 0x1p-64 1.0
catan -0x1p-64 1.0
catan 0x1p-64 -1.0
catan -0x1p-64 -1.0
catan 1.0 0x1.3p-73
catan 1.0 -0x1.3p-73
catan -1.0 0x1.3p-73
catan -1.0 -0x1.3p-73
catan 0x1.3p-73 1.0
catan -0x1.3p-73 1.0
catan 0x1.3p-73 -1.0
catan -0x1.3p-73 -1.0
catan 1.0 0x1p-124
catan 1.0 -0x1p-124
catan -1.0 0x1p-124
catan -1.0 -0x1p-124
catan 0x1p-124 1.0
catan -0x1p-124 1.0
catan 0x1p-124 -1.0
catan -0x1p-124 -1.0
catan 0x1.000002p0 0x1p-126
catan 0x1.000002p0 -0x1p-126
catan -0x1.000002p0 0x1p-126
catan -0x1.000002p0 -0x1p-126
catan 0x1p-126 0x1.000002p0
catan -0x1p-126 0x1.000002p0
catan 0x1p-126 -0x1.000002p0
catan -0x1p-126 -0x1.000002p0
catan 0x0.ffffffp0 0x1p-126
catan 0x0.ffffffp0 -0x1p-126
catan -0x0.ffffffp0 0x1p-126
catan -0x0.ffffffp0 -0x1p-126
catan 0x1p-126 0x0.ffffffp0
catan -0x1p-126 0x0.ffffffp0
catan 0x1p-126 -0x0.ffffffp0
catan -0x1p-126 -0x0.ffffffp0
catan 1.0 0x1p-512
catan 1.0 -0x1p-512
catan -1.0 0x1p-512
catan -1.0 -0x1p-512
catan 0x1p-512 1.0
catan -0x1p-512 1.0
catan 0x1p-512 -1.0
catan -0x1p-512 -1.0
catan 1.0 0x1.3p-536
catan 1.0 -0x1.3p-536
catan -1.0 0x1.3p-536
catan -1.0 -0x1.3p-536
catan 0x1.3p-536 1.0
catan -0x1.3p-536 1.0
catan 0x1.3p-536 -1.0
catan -0x1.3p-536 -1.0
catan 1.0 0x1p-1020
catan 1.0 -0x1p-1020
catan -1.0 0x1p-1020
catan -1.0 -0x1p-1020
catan 0x1p-1020 1.0
catan -0x1p-1020 1.0
catan 0x1p-1020 -1.0
catan -0x1p-1020 -1.0
catan 0x1.0000000000001p0 0x1p-1022
catan 0x1.0000000000001p0 -0x1p-1022
catan -0x1.0000000000001p0 0x1p-1022
catan -0x1.0000000000001p0 -0x1p-1022
catan 0x1p-1022 0x1.0000000000001p0
catan -0x1p-1022 0x1.0000000000001p0
catan 0x1p-1022 -0x1.0000000000001p0
catan -0x1p-1022 -0x1.0000000000001p0
catan 0x0.fffffffffffff8p0 0x1p-1022
catan 0x0.fffffffffffff8p0 -0x1p-1022
catan -0x0.fffffffffffff8p0 0x1p-1022
catan -0x0.fffffffffffff8p0 -0x1p-1022
catan 0x1p-1022 0x0.fffffffffffff8p0
catan -0x1p-1022 0x0.fffffffffffff8p0
catan 0x1p-1022 -0x0.fffffffffffff8p0
catan -0x1p-1022 -0x0.fffffffffffff8p0
catan 0x1.0000000000000002p0 0x1p-1022
catan 0x1.0000000000000002p0 -0x1p-1022
catan -0x1.0000000000000002p0 0x1p-1022
catan -0x1.0000000000000002p0 -0x1p-1022
catan 0x1p-1022 0x1.0000000000000002p0
catan -0x1p-1022 0x1.0000000000000002p0
catan 0x1p-1022 -0x1.0000000000000002p0
catan -0x1p-1022 -0x1.0000000000000002p0
catan 0x0.ffffffffffffffffp0 0x1p-1022
catan 0x0.ffffffffffffffffp0 -0x1p-1022
catan -0x0.ffffffffffffffffp0 0x1p-1022
catan -0x0.ffffffffffffffffp0 -0x1p-1022
catan 0x1p-1022 0x0.ffffffffffffffffp0
catan -0x1p-1022 0x0.ffffffffffffffffp0
catan 0x1p-1022 -0x0.ffffffffffffffffp0
catan -0x1p-1022 -0x0.ffffffffffffffffp0
catan 0x1.000000000000000000000000008p0 0x1p-1022
catan 0x1.000000000000000000000000008p0 -0x1p-1022
catan -0x1.000000000000000000000000008p0 0x1p-1022
catan -0x1.000000000000000000000000008p0 -0x1p-1022
catan 0x1p-1022 0x1.000000000000000000000000008p0
catan -0x1p-1022 0x1.000000000000000000000000008p0
catan 0x1p-1022 -0x1.000000000000000000000000008p0
catan -0x1p-1022 -0x1.000000000000000000000000008p0
catan 0x0.ffffffffffffffffffffffffffcp0 0x1p-1022
catan 0x0.ffffffffffffffffffffffffffcp0 -0x1p-1022
catan -0x0.ffffffffffffffffffffffffffcp0 0x1p-1022
catan -0x0.ffffffffffffffffffffffffffcp0 -0x1p-1022
catan 0x1p-1022 0x0.ffffffffffffffffffffffffffcp0
catan -0x1p-1022 0x0.ffffffffffffffffffffffffffcp0
catan 0x1p-1022 -0x0.ffffffffffffffffffffffffffcp0
catan -0x1p-1022 -0x0.ffffffffffffffffffffffffffcp0
catan 1.0 0x1p-8192
catan 1.0 -0x1p-8192
catan -1.0 0x1p-8192
catan -1.0 -0x1p-8192
catan 0x1p-8192 1.0
catan -0x1p-8192 1.0
catan 0x1p-8192 -1.0
catan -0x1p-8192 -1.0
catan 1.0 0x1.3p-8221
catan 1.0 -0x1.3p-8221
catan -1.0 0x1.3p-8221
catan -1.0 -0x1.3p-8221
catan 0x1.3p-8221 1.0
catan -0x1.3p-8221 1.0
catan 0x1.3p-8221 -1.0
catan -0x1.3p-8221 -1.0
catan 1.0 0x1.3p-8246
catan 1.0 -0x1.3p-8246
catan -1.0 0x1.3p-8246
catan -1.0 -0x1.3p-8246
catan 0x1.3p-8246 1.0
catan -0x1.3p-8246 1.0
catan 0x1.3p-8246 -1.0
catan -0x1.3p-8246 -1.0
catan 1.0 0x1p-16380
catan 1.0 -0x1p-16380
catan -1.0 0x1p-16380
catan -1.0 -0x1p-16380
catan 0x1p-16380 1.0
catan -0x1p-16380 1.0
catan 0x1p-16380 -1.0
catan -0x1p-16380 -1.0
catan 0x1.0000000000000002p0 0x1p-16382
catan 0x1.0000000000000002p0 -0x1p-16382
catan -0x1.0000000000000002p0 0x1p-16382
catan -0x1.0000000000000002p0 -0x1p-16382
catan 0x1p-16382 0x1.0000000000000002p0
catan -0x1p-16382 0x1.0000000000000002p0
catan 0x1p-16382 -0x1.0000000000000002p0
catan -0x1p-16382 -0x1.0000000000000002p0
catan 0x0.ffffffffffffffffp0 0x1p-16382
catan 0x0.ffffffffffffffffp0 -0x1p-16382
catan -0x0.ffffffffffffffffp0 0x1p-16382
catan -0x0.ffffffffffffffffp0 -0x1p-16382
catan 0x1p-16382 0x0.ffffffffffffffffp0
catan -0x1p-16382 0x0.ffffffffffffffffp0
catan 0x1p-16382 -0x0.ffffffffffffffffp0
catan -0x1p-16382 -0x0.ffffffffffffffffp0
catan 0x1.0000000000000000000000000001p0 0x1p-16382
catan 0x1.0000000000000000000000000001p0 -0x1p-16382
catan -0x1.0000000000000000000000000001p0 0x1p-16382
catan -0x1.0000000000000000000000000001p0 -0x1p-16382
catan 0x1p-16382 0x1.0000000000000000000000000001p0
catan -0x1p-16382 0x1.0000000000000000000000000001p0
catan 0x1p-16382 -0x1.0000000000000000000000000001p0
catan -0x1p-16382 -0x1.0000000000000000000000000001p0
catan 0x0.ffffffffffffffffffffffffffff8p0 0x1p-16382
catan 0x0.ffffffffffffffffffffffffffff8p0 -0x1p-16382
catan -0x0.ffffffffffffffffffffffffffff8p0 0x1p-16382
catan -0x0.ffffffffffffffffffffffffffff8p0 -0x1p-16382
catan 0x1p-16382 0x0.ffffffffffffffffffffffffffff8p0
catan -0x1p-16382 0x0.ffffffffffffffffffffffffffff8p0
catan 0x1p-16382 -0x0.ffffffffffffffffffffffffffff8p0
catan -0x1p-16382 -0x0.ffffffffffffffffffffffffffff8p0
catan 0.75 1.25
catan -2 -3
catanh 0 0
catanh -0 0
catanh 0 -0
catanh -0 -0
catanh 0 1.0
catanh -0 1.0
catanh 0 -1.0
catanh -0 -1.0
catanh 0x1p50 0.0
catanh 0x1p50 -0.0
catanh -0x1p50 0.0
catanh -0x1p50 -0.0
catanh 0.0 0x1p50
catanh -0.0 0x1p50
catanh 0.0 -0x1p50
catanh -0.0 -0x1p50
catanh 0x1p500 0.0
catanh 0x1p500 -0.0
catanh -0x1p500 0.0
catanh -0x1p500 -0.0
catanh 0.0 0x1p500
catanh -0.0 0x1p500
catanh 0.0 -0x1p500
catanh -0.0 -0x1p500
catanh 0x1p5000 0.0
catanh 0x1p5000 -0.0
catanh -0x1p5000 0.0
catanh -0x1p5000 -0.0
catanh 0.0 0x1p5000
catanh -0.0 0x1p5000
catanh 0.0 -0x1p5000
catanh -0.0 -0x1p5000
catanh 0x1p63 0.5
catanh 0x1p63 -0.5
catanh -0x1p63 0.5
catanh -0x1p63 -0.5
catanh 0.5 0x1p63
catanh -0.5 0x1p63
catanh 0.5 -0x1p63
catanh -0.5 -0x1p63
catanh 0x1p511 0.5
catanh 0x1p511 -0.5
catanh -0x1p511 0.5
catanh -0x1p511 -0.5
catanh 0.5 0x1p511
catanh -0.5 0x1p511
catanh 0.5 -0x1p511
catanh -0.5 -0x1p511
catanh 0x1p8191 0.5
catanh 0x1p8191 -0.5
catanh -0x1p8191 0.5
catanh -0x1p8191 -0.5
catanh 0.5 0x1p8191
catanh -0.5 0x1p8191
catanh 0.5 -0x1p8191
catanh -0.5 -0x1p8191
catanh 0x1p100 0.0
catanh 0x1p100 -0.0
catanh -0x1p100 0.0
catanh -0x1p100 -0.0
catanh 0.0 0x1p100
catanh -0.0 0x1p100
catanh 0.0 -0x1p100
catanh -0.0 -0x1p100
catanh 0x1.fp127 0.0
catanh 0x1.fp127 -0.0
catanh -0x1.fp127 0.0
catanh -0x1.fp127 -0.0
catanh 0.0 0x1.fp127
catanh -0.0 0x1.fp127
catanh 0.0 -0x1.fp127
catanh -0.0 -0x1.fp127
catanh 0x1.fp127 0x1.fp127
catanh 0x1.fp127 -0x1.fp127
catanh -0x1.fp127 0x1.fp127
catanh -0x1.fp127 -0x1.fp127
catanh 0x1p900 0.0
catanh 0x1p900 -0.0
catanh -0x1p900 0.0
catanh -0x1p900 -0.0
catanh 0.0 0x1p900
catanh -0.0 0x1p900
catanh 0.0 -0x1p900
catanh -0.0 -0x1p900
catanh 0x1.fp1023 0.0
catanh 0x1.fp1023 -0.0
catanh -0x1.fp1023 0.0
catanh -0x1.fp1023 -0.0
catanh 0.0 0x1.fp1023
catanh -0.0 0x1.fp1023
catanh 0.0 -0x1.fp1023
catanh -0.0 -0x1.fp1023
catanh 0x1.fp1023 0x1.fp1023
catanh 0x1.fp1023 -0x1.fp1023
catanh -0x1.fp1023 0x1.fp1023
catanh -0x1.fp1023 -0x1.fp1023
catanh 0x1p10000 0.0
catanh 0x1p10000 -0.0
catanh -0x1p10000 0.0
catanh -0x1p10000 -0.0
catanh 0.0 0x1p10000
catanh -0.0 0x1p10000
catanh 0.0 -0x1p10000
catanh -0.0 -0x1p10000
catanh 0x1.fp16383 0.0
catanh 0x1.fp16383 -0.0
catanh -0x1.fp16383 0.0
catanh -0x1.fp16383 -0.0
catanh 0.0 0x1.fp16383
catanh -0.0 0x1.fp16383
catanh 0.0 -0x1.fp16383
catanh -0.0 -0x1.fp16383
XFAIL catan and catanh tests on ibm128 Mark 4 catan and catanh tests as xfail-rounding:ibm128-libgcc. After this patch all catan and catanh tests pass on ibm128. Regenerated auto-libm-test-out using gmp 6.1.2, mpfr 3.1.5 and upstream mpc (9ef8030e50), Tested on powerpc, powerpc64 and powerpc64le. * math/auto-libm-test-in: Mark some catan and catanh as xfail-rounding:ibm128-libgcc. * math/auto-libm-test-out-catan: Regenerate. * math/auto-libm-test-out-catanh: Likewise.
2017-03-16 13:18:46 +00:00
catanh 0x1.fp16383 0x1.fp16383 xfail-rounding:ibm128-libgcc
catanh 0x1.fp16383 -0x1.fp16383 xfail-rounding:ibm128-libgcc
catanh -0x1.fp16383 0x1.fp16383 xfail-rounding:ibm128-libgcc
catanh -0x1.fp16383 -0x1.fp16383 xfail-rounding:ibm128-libgcc
Move tests of catan, catanh to auto-libm-test-*. This patch moves tests of catan and catanh with finite inputs (other than the divide-by-zero cases producing an exact infinity) to using the auto-libm-test machinery. Each of auto-libm-test-out-catan and auto-libm-test-out-catanh takes about three seconds to generate on my system (so in fact it wasn't necessary after all to defer the move to auto-libm-test-* until the output files were split up by function). Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of catan and catanh. * math/auto-libm-test-out-catan: New generated file. * math/auto-libm-test-out-catanh: Likewise. * math/libm-test-catan.inc (catan_test_data): Use AUTO_TESTS_c_c. Move tests with finite inputs, except divide-by-zero cases, to auto-libm-test-in. * math/libm-test-catanh.inc (catanh_test_data): Likewise. * math/Makefile (libm-test-funcs-auto): Add catan and catanh. (libm-test-funcs-noauto): Remove catan and catanh. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2017-02-17 18:42:37 +00:00
catanh 0x1p-13 1.0
catanh 0x1p-13 -1.0
catanh -0x1p-13 1.0
catanh -0x1p-13 -1.0
catanh 1.0 0x1p-13
catanh -1.0 0x1p-13
catanh 1.0 -0x1p-13
catanh -1.0 -0x1p-13
catanh 0x1p-27 1.0
catanh 0x1p-27 -1.0
catanh -0x1p-27 1.0
catanh -0x1p-27 -1.0
catanh 1.0 0x1p-27
catanh -1.0 0x1p-27
catanh 1.0 -0x1p-27
catanh -1.0 -0x1p-27
catanh 0x1p-33 1.0
catanh 0x1p-33 -1.0
catanh -0x1p-33 1.0
catanh -0x1p-33 -1.0
catanh 1.0 0x1p-33
catanh -1.0 0x1p-33
catanh 1.0 -0x1p-33
catanh -1.0 -0x1p-33
catanh 0x1p-54 1.0
catanh 0x1p-54 -1.0
catanh -0x1p-54 1.0
catanh -0x1p-54 -1.0
catanh 1.0 0x1p-54
catanh -1.0 0x1p-54
catanh 1.0 -0x1p-54
catanh -1.0 -0x1p-54
catanh 0x1p-57 1.0
catanh 0x1p-57 -1.0
catanh -0x1p-57 1.0
catanh -0x1p-57 -1.0
catanh 1.0 0x1p-57
catanh -1.0 0x1p-57
catanh 1.0 -0x1p-57
catanh -1.0 -0x1p-57
catanh 0x1p-13 0x1.000002p0
catanh 0x1p-13 -0x1.000002p0
catanh -0x1p-13 0x1.000002p0
catanh -0x1p-13 -0x1.000002p0
catanh 0x1.000002p0 0x1p-13
catanh -0x1.000002p0 0x1p-13
catanh 0x1.000002p0 -0x1p-13
catanh -0x1.000002p0 -0x1p-13
catanh 0x1p-13 0x0.ffffffp0
catanh 0x1p-13 -0x0.ffffffp0
catanh -0x1p-13 0x0.ffffffp0
catanh -0x1p-13 -0x0.ffffffp0
catanh 0x0.ffffffp0 0x1p-13
catanh -0x0.ffffffp0 0x1p-13
catanh 0x0.ffffffp0 -0x1p-13
catanh -0x0.ffffffp0 -0x1p-13
catanh 0x1p-27 0x1.0000000000001p0
catanh 0x1p-27 -0x1.0000000000001p0
catanh -0x1p-27 0x1.0000000000001p0
catanh -0x1p-27 -0x1.0000000000001p0
catanh 0x1.0000000000001p0 0x1p-27
catanh -0x1.0000000000001p0 0x1p-27
catanh 0x1.0000000000001p0 -0x1p-27
catanh -0x1.0000000000001p0 -0x1p-27
catanh 0x1p-27 0x0.fffffffffffff8p0
catanh 0x1p-27 -0x0.fffffffffffff8p0
catanh -0x1p-27 0x0.fffffffffffff8p0
catanh -0x1p-27 -0x0.fffffffffffff8p0
catanh 0x0.fffffffffffff8p0 0x1p-27
catanh -0x0.fffffffffffff8p0 0x1p-27
catanh 0x0.fffffffffffff8p0 -0x1p-27
catanh -0x0.fffffffffffff8p0 -0x1p-27
catanh 0x1p-33 0x1.0000000000000002p0
catanh 0x1p-33 -0x1.0000000000000002p0
catanh -0x1p-33 0x1.0000000000000002p0
catanh -0x1p-33 -0x1.0000000000000002p0
catanh 0x1.0000000000000002p0 0x1p-33
catanh -0x1.0000000000000002p0 0x1p-33
catanh 0x1.0000000000000002p0 -0x1p-33
catanh -0x1.0000000000000002p0 -0x1p-33
catanh 0x1p-33 0x0.ffffffffffffffffp0
catanh 0x1p-33 -0x0.ffffffffffffffffp0
catanh -0x1p-33 0x0.ffffffffffffffffp0
catanh -0x1p-33 -0x0.ffffffffffffffffp0
catanh 0x0.ffffffffffffffffp0 0x1p-33
catanh -0x0.ffffffffffffffffp0 0x1p-33
catanh 0x0.ffffffffffffffffp0 -0x1p-33
catanh -0x0.ffffffffffffffffp0 -0x1p-33
catanh 0x1p-54 0x1.000000000000000000000000008p0
catanh 0x1p-54 -0x1.000000000000000000000000008p0
catanh -0x1p-54 0x1.000000000000000000000000008p0
catanh -0x1p-54 -0x1.000000000000000000000000008p0
catanh 0x1.000000000000000000000000008p0 0x1p-54
catanh -0x1.000000000000000000000000008p0 0x1p-54
catanh 0x1.000000000000000000000000008p0 -0x1p-54
catanh -0x1.000000000000000000000000008p0 -0x1p-54
catanh 0x1p-54 0x0.ffffffffffffffffffffffffffcp0
catanh 0x1p-54 -0x0.ffffffffffffffffffffffffffcp0
catanh -0x1p-54 0x0.ffffffffffffffffffffffffffcp0
catanh -0x1p-54 -0x0.ffffffffffffffffffffffffffcp0
catanh 0x0.ffffffffffffffffffffffffffcp0 0x1p-54
catanh -0x0.ffffffffffffffffffffffffffcp0 0x1p-54
catanh 0x0.ffffffffffffffffffffffffffcp0 -0x1p-54
catanh -0x0.ffffffffffffffffffffffffffcp0 -0x1p-54
catanh 0x1p-57 0x1.0000000000000000000000000001p0
catanh 0x1p-57 -0x1.0000000000000000000000000001p0
catanh -0x1p-57 0x1.0000000000000000000000000001p0
catanh -0x1p-57 -0x1.0000000000000000000000000001p0
catanh 0x1.0000000000000000000000000001p0 0x1p-57
catanh -0x1.0000000000000000000000000001p0 0x1p-57
catanh 0x1.0000000000000000000000000001p0 -0x1p-57
catanh -0x1.0000000000000000000000000001p0 -0x1p-57
catanh 0x1p-57 0x0.ffffffffffffffffffffffffffff8p0
catanh 0x1p-57 -0x0.ffffffffffffffffffffffffffff8p0
catanh -0x1p-57 0x0.ffffffffffffffffffffffffffff8p0
catanh -0x1p-57 -0x0.ffffffffffffffffffffffffffff8p0
catanh 0x0.ffffffffffffffffffffffffffff8p0 0x1p-57
catanh -0x0.ffffffffffffffffffffffffffff8p0 0x1p-57
catanh 0x0.ffffffffffffffffffffffffffff8p0 -0x1p-57
catanh -0x0.ffffffffffffffffffffffffffff8p0 -0x1p-57
catanh 1.0 0x1p-64
catanh 1.0 -0x1p-64
catanh -1.0 0x1p-64
catanh -1.0 -0x1p-64
catanh 0x1p-64 1.0
catanh -0x1p-64 1.0
catanh 0x1p-64 -1.0
catanh -0x1p-64 -1.0
catanh 1.0 0x1.3p-73
catanh 1.0 -0x1.3p-73
catanh -1.0 0x1.3p-73
catanh -1.0 -0x1.3p-73
catanh 0x1.3p-73 1.0
catanh -0x1.3p-73 1.0
catanh 0x1.3p-73 -1.0
catanh -0x1.3p-73 -1.0
catanh 1.0 0x1p-124
catanh 1.0 -0x1p-124
catanh -1.0 0x1p-124
catanh -1.0 -0x1p-124
catanh 0x1p-124 1.0
catanh -0x1p-124 1.0
catanh 0x1p-124 -1.0
catanh -0x1p-124 -1.0
catanh 0x1.000002p0 0x1p-126
catanh 0x1.000002p0 -0x1p-126
catanh -0x1.000002p0 0x1p-126
catanh -0x1.000002p0 -0x1p-126
catanh 0x1p-126 0x1.000002p0
catanh -0x1p-126 0x1.000002p0
catanh 0x1p-126 -0x1.000002p0
catanh -0x1p-126 -0x1.000002p0
catanh 0x0.ffffffp0 0x1p-126
catanh 0x0.ffffffp0 -0x1p-126
catanh -0x0.ffffffp0 0x1p-126
catanh -0x0.ffffffp0 -0x1p-126
catanh 0x1p-126 0x0.ffffffp0
catanh -0x1p-126 0x0.ffffffp0
catanh 0x1p-126 -0x0.ffffffp0
catanh -0x1p-126 -0x0.ffffffp0
catanh 1.0 0x1p-512
catanh 1.0 -0x1p-512
catanh -1.0 0x1p-512
catanh -1.0 -0x1p-512
catanh 0x1p-512 1.0
catanh -0x1p-512 1.0
catanh 0x1p-512 -1.0
catanh -0x1p-512 -1.0
catanh 1.0 0x1.3p-536
catanh 1.0 -0x1.3p-536
catanh -1.0 0x1.3p-536
catanh -1.0 -0x1.3p-536
catanh 0x1.3p-536 1.0
catanh -0x1.3p-536 1.0
catanh 0x1.3p-536 -1.0
catanh -0x1.3p-536 -1.0
catanh 1.0 0x1p-1020
catanh 1.0 -0x1p-1020
catanh -1.0 0x1p-1020
catanh -1.0 -0x1p-1020
catanh 0x1p-1020 1.0
catanh -0x1p-1020 1.0
catanh 0x1p-1020 -1.0
catanh -0x1p-1020 -1.0
catanh 0x1.0000000000001p0 0x1p-1022
catanh 0x1.0000000000001p0 -0x1p-1022
catanh -0x1.0000000000001p0 0x1p-1022
catanh -0x1.0000000000001p0 -0x1p-1022
catanh 0x1p-1022 0x1.0000000000001p0
catanh -0x1p-1022 0x1.0000000000001p0
catanh 0x1p-1022 -0x1.0000000000001p0
catanh -0x1p-1022 -0x1.0000000000001p0
catanh 0x0.fffffffffffff8p0 0x1p-1022
catanh 0x0.fffffffffffff8p0 -0x1p-1022
catanh -0x0.fffffffffffff8p0 0x1p-1022
catanh -0x0.fffffffffffff8p0 -0x1p-1022
catanh 0x1p-1022 0x0.fffffffffffff8p0
catanh -0x1p-1022 0x0.fffffffffffff8p0
catanh 0x1p-1022 -0x0.fffffffffffff8p0
catanh -0x1p-1022 -0x0.fffffffffffff8p0
catanh 0x1.0000000000000002p0 0x1p-1022
catanh 0x1.0000000000000002p0 -0x1p-1022
catanh -0x1.0000000000000002p0 0x1p-1022
catanh -0x1.0000000000000002p0 -0x1p-1022
catanh 0x1p-1022 0x1.0000000000000002p0
catanh -0x1p-1022 0x1.0000000000000002p0
catanh 0x1p-1022 -0x1.0000000000000002p0
catanh -0x1p-1022 -0x1.0000000000000002p0
catanh 0x0.ffffffffffffffffp0 0x1p-1022
catanh 0x0.ffffffffffffffffp0 -0x1p-1022
catanh -0x0.ffffffffffffffffp0 0x1p-1022
catanh -0x0.ffffffffffffffffp0 -0x1p-1022
catanh 0x1p-1022 0x0.ffffffffffffffffp0
catanh -0x1p-1022 0x0.ffffffffffffffffp0
catanh 0x1p-1022 -0x0.ffffffffffffffffp0
catanh -0x1p-1022 -0x0.ffffffffffffffffp0
catanh 0x1.000000000000000000000000008p0 0x1p-1022
catanh 0x1.000000000000000000000000008p0 -0x1p-1022
catanh -0x1.000000000000000000000000008p0 0x1p-1022
catanh -0x1.000000000000000000000000008p0 -0x1p-1022
catanh 0x1p-1022 0x1.000000000000000000000000008p0
catanh -0x1p-1022 0x1.000000000000000000000000008p0
catanh 0x1p-1022 -0x1.000000000000000000000000008p0
catanh -0x1p-1022 -0x1.000000000000000000000000008p0
catanh 0x0.ffffffffffffffffffffffffffcp0 0x1p-1022
catanh 0x0.ffffffffffffffffffffffffffcp0 -0x1p-1022
catanh -0x0.ffffffffffffffffffffffffffcp0 0x1p-1022
catanh -0x0.ffffffffffffffffffffffffffcp0 -0x1p-1022
catanh 0x1p-1022 0x0.ffffffffffffffffffffffffffcp0
catanh -0x1p-1022 0x0.ffffffffffffffffffffffffffcp0
catanh 0x1p-1022 -0x0.ffffffffffffffffffffffffffcp0
catanh -0x1p-1022 -0x0.ffffffffffffffffffffffffffcp0
catanh 1.0 0x1p-8192
catanh 1.0 -0x1p-8192
catanh -1.0 0x1p-8192
catanh -1.0 -0x1p-8192
catanh 0x1p-8192 1.0
catanh -0x1p-8192 1.0
catanh 0x1p-8192 -1.0
catanh -0x1p-8192 -1.0
catanh 1.0 0x1.3p-8221
catanh 1.0 -0x1.3p-8221
catanh -1.0 0x1.3p-8221
catanh -1.0 -0x1.3p-8221
catanh 0x1.3p-8221 1.0
catanh -0x1.3p-8221 1.0
catanh 0x1.3p-8221 -1.0
catanh -0x1.3p-8221 -1.0
catanh 1.0 0x1.3p-8246
catanh 1.0 -0x1.3p-8246
catanh -1.0 0x1.3p-8246
catanh -1.0 -0x1.3p-8246
catanh 0x1.3p-8246 1.0
catanh -0x1.3p-8246 1.0
catanh 0x1.3p-8246 -1.0
catanh -0x1.3p-8246 -1.0
catanh 1.0 0x1p-16380
catanh 1.0 -0x1p-16380
catanh -1.0 0x1p-16380
catanh -1.0 -0x1p-16380
catanh 0x1p-16380 1.0
catanh -0x1p-16380 1.0
catanh 0x1p-16380 -1.0
catanh -0x1p-16380 -1.0
catanh 0x1.0000000000000002p0 0x1p-16382
catanh 0x1.0000000000000002p0 -0x1p-16382
catanh -0x1.0000000000000002p0 0x1p-16382
catanh -0x1.0000000000000002p0 -0x1p-16382
catanh 0x1p-16382 0x1.0000000000000002p0
catanh -0x1p-16382 0x1.0000000000000002p0
catanh 0x1p-16382 -0x1.0000000000000002p0
catanh -0x1p-16382 -0x1.0000000000000002p0
catanh 0x0.ffffffffffffffffp0 0x1p-16382
catanh 0x0.ffffffffffffffffp0 -0x1p-16382
catanh -0x0.ffffffffffffffffp0 0x1p-16382
catanh -0x0.ffffffffffffffffp0 -0x1p-16382
catanh 0x1p-16382 0x0.ffffffffffffffffp0
catanh -0x1p-16382 0x0.ffffffffffffffffp0
catanh 0x1p-16382 -0x0.ffffffffffffffffp0
catanh -0x1p-16382 -0x0.ffffffffffffffffp0
catanh 0x1.0000000000000000000000000001p0 0x1p-16382
catanh 0x1.0000000000000000000000000001p0 -0x1p-16382
catanh -0x1.0000000000000000000000000001p0 0x1p-16382
catanh -0x1.0000000000000000000000000001p0 -0x1p-16382
catanh 0x1p-16382 0x1.0000000000000000000000000001p0
catanh -0x1p-16382 0x1.0000000000000000000000000001p0
catanh 0x1p-16382 -0x1.0000000000000000000000000001p0
catanh -0x1p-16382 -0x1.0000000000000000000000000001p0
catanh 0x0.ffffffffffffffffffffffffffff8p0 0x1p-16382
catanh 0x0.ffffffffffffffffffffffffffff8p0 -0x1p-16382
catanh -0x0.ffffffffffffffffffffffffffff8p0 0x1p-16382
catanh -0x0.ffffffffffffffffffffffffffff8p0 -0x1p-16382
catanh 0x1p-16382 0x0.ffffffffffffffffffffffffffff8p0
catanh -0x1p-16382 0x0.ffffffffffffffffffffffffffff8p0
catanh 0x1p-16382 -0x0.ffffffffffffffffffffffffffff8p0
catanh -0x1p-16382 -0x0.ffffffffffffffffffffffffffff8p0
catanh 0.75 1.25
catanh -2 -3
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
cbrt 0.0
cbrt -0
cbrt -0.001
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
cbrt 2
cbrt 4
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
cbrt 8
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
cbrt -10
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
cbrt -27.0
cbrt 0.9921875
cbrt 0.75
cbrt 0x1p16383
cbrt 0x1p-16383
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
cbrt 1e5
Add more tests of cbrt. This patch adds some randomly-generated tests of cbrt that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cbrt. * math/auto-libm-test-out: Regenerated. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-04-08 17:56:15 +00:00
cbrt 0x3.132634p+0
cbrt -0xc.8d0442f2f0d1p-492
cbrt -0xa.6b142p+40
cbrt -0x1.f28ab85f3580ap-128
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
cbrt -0x2.b5cd28p-36
cbrt -0x1.d6a8bep-20
cbrt -0x3.593ed8p-72
cbrt 0x1.bd0098p-104
cbrt -0x3.300d34p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
cbrt 0x6.247f5p-4
Add more randomly-generated libm tests. This patch adds more libm test inputs found through random test generation to increase observed ulps on x86_64. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, atanh, cbrt, cosh, csqrt, erfc, expm1 and lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-11 15:03:10 +00:00
cbrt -0x3.48648028cb464p+0
add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)
2020-11-25 18:03:33 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
cbrt -0x1.7cec1dab41272p-230
Update libm-test XFAILs for ibm128 format. This patch cleans up and updates the libm-test XFAILs for the ibm128 format. More of them are changed to use a new ibm128-libgcc conditional, to reflect that they are not in fact needed if you've patched libgcc to fix the known issues (at substantial performance cost). Many additional XFAILs are added for tests that fail with unpatched libgcc (most but not all of them xfail-rounding). Note that further such fixes will be needed for test-ldouble actually to pass with default libgcc (in particular, XFAILs for pow tests and for various affected tests directly embedded in libm-test.inc). With patched libgcc, there may be a few XFAILs needed but the results are already substantially clean apart from a few ulps differences. Tested for powerpc. * math/libm-test.inc (TEST_COND_ibm128_libgcc): New macro. (init_max_error) [TEST_COND_ibm128]: Increase maximum error allowed to 16 ulps. * math/auto-libm-test-in: Change most XFAILs for ibm128 to use ibm128-libgcc. XFAIL more tests for ibm128-libgcc. * math/auto-libm-test-out: Regenerated.
2017-01-04 23:29:42 +00:00
cbrt max xfail:ibm128-libgcc
cbrt -max xfail:ibm128-libgcc
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
cbrt min
cbrt -min
cbrt min_subnorm
cbrt -min_subnorm
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ccos 0.0 0.0
ccos -0 0.0
ccos 0.0 -0
ccos -0 -0
ccos 0.75 1.25
ccos -2 -3
ccos 0.75 89.5
ccos 0.75 -89.5
ccos -0.75 89.5
ccos -0.75 -89.5
ccos 0.75 710.5
ccos 0.75 -710.5
ccos -0.75 710.5
ccos -0.75 -710.5
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
ccos 0.75 11357.25
ccos 0.75 -11357.25
ccos -0.75 11357.25
ccos -0.75 -11357.25
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ccos 0x1p-149 180
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
ccos 0x1p-1074 1440
ccos 0x1p-16434 22730
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ccos min_subnorm_p120 0x1p-120
ccos 0x1p-120 min_subnorm_p120
Fix cexp, ccos, ccosh, csin, csinh spurious underflows (bug 18594). cexp, ccos, ccosh, csin and csinh have spurious underflows in cases where they compute sin of the smallest normal, that produces an underflow exception (depending on which sin implementation is in use) but the final result does not underflow. ctan and ctanh may also have such underflows, or they may be latent (the issue there is that e.g. ctan (DBL_MIN) should, rounded upwards, be the next double value above DBL_MIN, which under glibc's accuracy goals may not have an underflow exception, but the intermediate computation of sin (DBL_MIN) would legitimately underflow on before-rounding architectures). This patch fixes all those functions so they use plain comparisons (> DBL_MIN etc.) instead of comparing the result of fpclassify with FP_SUBNORMAL (in all these cases, we already know the number being compared is finite). Note that in the case of csin / csinf / csinl, there is no need for fabs calls in the comparison because the real part has already been reduced to its absolute value. As the patch fixes the failures that previously obstructed moving tests of cexp to use ALL_RM_TEST, those tests are moved to ALL_RM_TEST by the patch (two functions remain yet to be converted). Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18594] * math/s_ccosh.c (__ccosh): Compare with least normal value instead of comparing class with FP_SUBNORMAL. * math/s_ccoshf.c (__ccoshf): Likewise. * math/s_ccoshl.c (__ccoshl): Likewise. * math/s_cexp.c (__cexp): Likewise. * math/s_cexpf.c (__cexpf): Likewise. * math/s_cexpl.c (__cexpl): Likewise. * math/s_csin.c (__csin): Likewise. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/s_ctan.c (__ctan): Likewise. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Add more tests of ccos, ccosh, cexp, csin, csinh, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cexp_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-24 21:04:51 +00:00
ccos min 1
ccos -min 1
ccos min_subnorm 80
ccos -min_subnorm 80
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ccosh 0.0 0.0
ccosh -0 0.0
ccosh 0.0 -0
ccosh -0 -0
ccosh 0.75 1.25
ccosh -2 -3
ccosh 89.5 0.75
ccosh -89.5 0.75
ccosh 89.5 -0.75
ccosh -89.5 -0.75
ccosh 710.5 0.75
ccosh -710.5 0.75
ccosh 710.5 -0.75
ccosh -710.5 -0.75
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
ccosh 11357.25 0.75
ccosh -11357.25 0.75
ccosh 11357.25 -0.75
ccosh -11357.25 -0.75
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ccosh 180 0x1p-149
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
ccosh 1440 0x1p-1074
ccosh 22730 0x1p-16434
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ccosh min_subnorm_p120 0x1p-120
ccosh 0x1p-120 min_subnorm_p120
Fix cexp, ccos, ccosh, csin, csinh spurious underflows (bug 18594). cexp, ccos, ccosh, csin and csinh have spurious underflows in cases where they compute sin of the smallest normal, that produces an underflow exception (depending on which sin implementation is in use) but the final result does not underflow. ctan and ctanh may also have such underflows, or they may be latent (the issue there is that e.g. ctan (DBL_MIN) should, rounded upwards, be the next double value above DBL_MIN, which under glibc's accuracy goals may not have an underflow exception, but the intermediate computation of sin (DBL_MIN) would legitimately underflow on before-rounding architectures). This patch fixes all those functions so they use plain comparisons (> DBL_MIN etc.) instead of comparing the result of fpclassify with FP_SUBNORMAL (in all these cases, we already know the number being compared is finite). Note that in the case of csin / csinf / csinl, there is no need for fabs calls in the comparison because the real part has already been reduced to its absolute value. As the patch fixes the failures that previously obstructed moving tests of cexp to use ALL_RM_TEST, those tests are moved to ALL_RM_TEST by the patch (two functions remain yet to be converted). Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18594] * math/s_ccosh.c (__ccosh): Compare with least normal value instead of comparing class with FP_SUBNORMAL. * math/s_ccoshf.c (__ccoshf): Likewise. * math/s_ccoshl.c (__ccoshl): Likewise. * math/s_cexp.c (__cexp): Likewise. * math/s_cexpf.c (__cexpf): Likewise. * math/s_cexpl.c (__cexpl): Likewise. * math/s_csin.c (__csin): Likewise. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/s_ctan.c (__ctan): Likewise. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Add more tests of ccos, ccosh, cexp, csin, csinh, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cexp_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-24 21:04:51 +00:00
ccosh 1 min
ccosh 1 -min
ccosh 80 min_subnorm
ccosh 80 -min_subnorm
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
cexp 0 0
cexp -0 0
cexp 0 -0
cexp -0 -0
cexp 0.75 1.25
cexp -2.0 -3.0
cexp 0 0x1p65
cexp 0 -0x1p65
cexp 50 0x1p127
cexp 0 1e22
cexp 0 0x1p1023
cexp 500 0x1p1023
cexp 0 0x1p16383
cexp -10000 0x1p16383
cexp 88.75 0.75
cexp -95 0.75
cexp 709.8125 0.75
cexp -720 0.75
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
cexp 11356.5625 0.75
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
cexp -11370 0.75
cexp 180 0x1p-149
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
cexp 1440 0x1p-1074
cexp 22730 0x1p-16434
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
cexp 1e6 0
cexp 1e6 min
cexp 1e6 -min
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
Fix cexp, ccos, ccosh, csin, csinh spurious underflows (bug 18594). cexp, ccos, ccosh, csin and csinh have spurious underflows in cases where they compute sin of the smallest normal, that produces an underflow exception (depending on which sin implementation is in use) but the final result does not underflow. ctan and ctanh may also have such underflows, or they may be latent (the issue there is that e.g. ctan (DBL_MIN) should, rounded upwards, be the next double value above DBL_MIN, which under glibc's accuracy goals may not have an underflow exception, but the intermediate computation of sin (DBL_MIN) would legitimately underflow on before-rounding architectures). This patch fixes all those functions so they use plain comparisons (> DBL_MIN etc.) instead of comparing the result of fpclassify with FP_SUBNORMAL (in all these cases, we already know the number being compared is finite). Note that in the case of csin / csinf / csinl, there is no need for fabs calls in the comparison because the real part has already been reduced to its absolute value. As the patch fixes the failures that previously obstructed moving tests of cexp to use ALL_RM_TEST, those tests are moved to ALL_RM_TEST by the patch (two functions remain yet to be converted). Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18594] * math/s_ccosh.c (__ccosh): Compare with least normal value instead of comparing class with FP_SUBNORMAL. * math/s_ccoshf.c (__ccoshf): Likewise. * math/s_ccoshl.c (__ccoshl): Likewise. * math/s_cexp.c (__cexp): Likewise. * math/s_cexpf.c (__cexpf): Likewise. * math/s_cexpl.c (__cexpl): Likewise. * math/s_csin.c (__csin): Likewise. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/s_ctan.c (__ctan): Likewise. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Add more tests of ccos, ccosh, cexp, csin, csinh, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cexp_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-24 21:04:51 +00:00
cexp 1 min
cexp 1 -min
cexp 80 min_subnorm
cexp 80 -min_subnorm
Remove stray spurious-underflow markings from cexp test. I noticed that I'd left a spurious-underflow allowance behind in auto-libm-test-in for a bug that was fixed some time ago. This patch removes it. Tested for x86_64 and x86. * math/auto-libm-test-in: Remove spurious underflow allowance for tests of cexp. * math/auto-libm-test-out: Regenerated.
2015-06-18 20:51:17 +00:00
cexp min min_subnorm
cexp min -min_subnorm
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
clog 0.75 1.25
clog -2 -3
Fix dbl-64 atan2 in non-default rounding modes (bug 18210, bug 18211). The dbl-64 implementation of atan2 does computations that expect to run in round-to-nearest mode, and in other modes the errors can accumulate to more than the maximum accepted 9ulp. This patch makes it use FE_TONEAREST internally, similar to other functions with such issues. Tests that previously produced large errors are added for atan2 and the closely related carg, clog and clog10 functions. Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18210] [BZ #18211] * sysdeps/ieee754/dbl-64/e_atan2.c: Include <fenv.h>. (__ieee754_atan2): Set FE_TONEAREST mode for internal computations. * math/auto-libm-test-in: Add more tests of atan2, carg, clog and clog10. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-08 17:32:17 +00:00
clog 0x2.f2f308p+0 0x4.c3841p-4
clog 0xd.3de7ap-36 -0xe.cf143p-40
clog 0x2.21e65p+0 0x5.576cf8p-4
clog 0x1.f4755cp+0 -0x4.29411p-4
clog -0xf.9c4c8p-4 -0xa.b4101p+20
clog 0x7.40ac68p+0 0x4.251bb8p-4
clog 0xa.3ac3cp+68 0x1.47239ep+68
clog 0x3.8ff10cp+0 -0x6.b0794p-4
Add more tests of clog and clog10. This patch adds some randomly-generated tests of clog and clog10 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of clog and clog10. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-09 22:14:00 +00:00
clog 0xa.a39ffp-4 -0x2.360c38p-4
clog 0x6.9a4569067b6ecp-4 0xb.0a30d15e7d798p-4
clog -0x1.105436p+0 -0x6.66396df3cc7ap-4
clog -0x2.c90b952282392dep-4 0x1.43cda16634cc7046p+0
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
clog -0x9.93d164127d9fp-4 0x7.c5c8d8p-4
clog -0xa.5920ap-4 -0x6.2cda5p-4
clog 0xd.d05c38ebb1b4p+60 -0x3.c22fdp+44
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
clog -0xa.19f8ec252c58d5p-4 0x7.d10cdec29a141538p-4
clog -0xa.7ac41a0b417cb8fp-4 -0x6.c5a32eaeedd4p-4
clog 0x3.c16p-136 0x8p-152
clog -0x1.0a69de710590dp+0 -0x7.bc7e121e2b0d1088p-4
Fix clog, clog10 inaccuracy (bug 19016). For arguments with X^2 + Y^2 close to 1, clog and clog10 avoid large errors from log(hypot) by computing X^2 + Y^2 - 1 in a way that avoids cancellation error and then using log1p. However, the thresholds for using that approach still result in log being used on argument as large as sqrt(13/16) > 0.9, leading to significant errors, in some cases above the 9ulp maximum allowed in glibc libm. This patch arranges for the approach using log1p to be used in any cases where |X|, |Y| < 1 and X^2 + Y^2 >= 0.5 (with the existing allowance for cases where one of X and Y is very small), adjusting the __x2y2m1 functions to work with the wider range of inputs. This way, log only gets used on arguments below sqrt(1/2) (or substantially above 1), where the error involved is much less. Tested for x86_64, x86, mips64 and powerpc. For the ulps regeneration I removed the existing clog and clog10 ulps before regenerating to allow any reduced ulps to appear. Tests added include those found by random test generation to produce large ulps either before or after the patch, and some found by trying inputs close to the (0.75, 0.5) threshold where the potential errors from using log are largest. [BZ #19016] * sysdeps/generic/math_private.h (__x2y2m1f): Update comment to allow more cases with X^2 + Y^2 >= 0.5. * sysdeps/ieee754/dbl-64/x2y2m1.c (__x2y2m1): Likewise. Add -1 as normal element in sum instead of special-casing based on values of arguments. * sysdeps/ieee754/dbl-64/x2y2m1f.c (__x2y2m1f): Update comment. * sysdeps/ieee754/ldbl-128/x2y2m1l.c (__x2y2m1l): Likewise. Add -1 as normal element in sum instead of special-casing based on values of arguments. * sysdeps/ieee754/ldbl-128ibm/x2y2m1l.c (__x2y2m1l): Likewise. * sysdeps/ieee754/ldbl-96/x2y2m1.c [FLT_EVAL_METHOD != 0] (__x2y2m1): Update comment. * sysdeps/ieee754/ldbl-96/x2y2m1l.c (__x2y2m1l): Likewise. Add -1 as normal element in sum instead of special-casing based on values of arguments. * math/s_clog.c (__clog): Handle more cases using log1p without hypot. * math/s_clog10.c (__clog10): Likewise. * math/s_clog10f.c (__clog10f): Likewise. * math/s_clog10l.c (__clog10l): Likewise. * math/s_clogf.c (__clogf): Likewise. * math/s_clogl.c (__clogl): Likewise. * math/auto-libm-test-in: Add more tests of clog and clog10. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-28 22:11:22 +00:00
clog -0x2.7bdep-4 0x5.ab7a4p-4
clog -0xb.e1d3d0ff44358p-4 -0x7.54785e1b143f8p-4
clog 0x3.ba473p+0 0x7.eea9ap-4
clog 0x9.d02220baee4ep+36 0x2.b9a29cp+0
clog -0x5.1a5cf8p-4 -0xb.73012p-4
clog -0xa.ff292a609dbb8p-4 0x6.f73d4cp-4
clog -0x5.1a5cfc2301114p-4 -0xb.730118p-4
clog 0xb.ffffcp-4 0x7.ffff1p-4
clog 0xb.ffffp-4 0x7.ffffap-4
clog 0xb.ffffp-4 0x7.fffff8p-4
clog 0xb.ffffp-4 0x7.ffffp-4
clog 0xb.fffffp-4 0x7.ffff68p-4
clog 0xb.fffffp-4 0x7.ffffp-4
clog 0xb.ffff8p-4 0x7.ffffcp-4
clog 0xb.ffffp-4 0x7.ffffcp-4
clog 0xb.ffffp-4 0x7.ffffb8p-4
clog 0xb.ffffp-4 0x7.ffff7p-4
clog 0xb.ffffp-4 0x7.ffff5p-4
clog 0xb.fffffffffff7p-4 0x7.fffff8p-4
clog 0xb.fffffffffff08p-4 0x7.fffffffffffdp-4
clog 0xb.fffffffffff08p-4 0x7.fffffffffff9p-4
clog 0xb.fffffffffffp-4 0x7.fffffffffffdcp-4
clog 0xb.fffffp-4 0x7.ffffffffffff4p-4
clog 0xb.fffffffffffp-4 0x7.fffffffffffecp-4
clog 0xb.fffffffffff8p-4 0x7.fffff8p-4
clog 0x8p-152 -0x1.10233ap+0
clog 0xa.03634p-4 -0x4.7bb918p-20
clog -0x5.e23d2p-4 0x8.525df889c21ap-4
clog 0x9.8ce58p-4 -0x8p-152
clog 0x8p-152 0x9.2af75p-4
clog 0x9.97a15de8e59d8p-4 -0
clog -0x4.74556ec92eb4746p-4 0x1.1e7aa1d936f6efe6p+0
clog 0x9.97a15de8e59d8p-4 -0
clog -0x9.7f1d7p-64 0x9.db37dp-4
clog -0x8.5efc4p-4 -0x5.40310cp-4
clog -0x9.0b459p-4 0
clog -0x6.a9419e9b30e68p-4 -0x6.262c7p-4
clog 0x5.2767cdfdfbf2p-4 0x7.69ee98p-4
clog -0x9.f5563cb3227d8p-4 0
clog -0x9.5a284p-4 0x6.899578p-8
clog 0xa.3e62bp-4 0x1.18c03p-100
clog 0 -0x9.22a99p-4
clog 0 0x9.7915bp-4
clog 0x3.00d1ap-12 0x1.23ff6ap+0
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
clog 0x1.fffffep+127 0x1.fffffep+127
clog 0x1.fffffep+127 1.0
clog 0x1p-149 0x1p-149
clog 0x1p-147 0x1p-147
clog 0x1.fffffffffffffp+1023 0x1.fffffffffffffp+1023
clog 0x1.fffffffffffffp+1023 0x1p+1023
clog 0x1p-1074 0x1p-1074
clog 0x1p-1073 0x1p-1073
clog 0x1.fp+16383 0x1.fp+16383
clog 0x1.fp+16383 0x1p+16383
clog 0x1p-16440 0x1p-16441
clog 0x1p-149 0x1.fp+127
clog -0x1p-149 0x1.fp+127
clog 0x1p-149 -0x1.fp+127
clog -0x1p-149 -0x1.fp+127
clog -0x1.fp+127 0x1p-149
clog -0x1.fp+127 -0x1p-149
clog 0x1.fp+127 0x1p-149
clog 0x1.fp+127 -0x1p-149
clog 0x1p-1074 0x1.fp+1023
clog -0x1p-1074 0x1.fp+1023
clog 0x1p-1074 -0x1.fp+1023
clog -0x1p-1074 -0x1.fp+1023
clog -0x1.fp+1023 0x1p-1074
clog -0x1.fp+1023 -0x1p-1074
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
clog 0x1.fp+1023 0x1p-1074
clog 0x1.fp+1023 -0x1p-1074
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
clog 0x1p-16445 0x1.fp+16383
clog -0x1p-16445 0x1.fp+16383
clog 0x1p-16445 -0x1.fp+16383
clog -0x1p-16445 -0x1.fp+16383
clog -0x1.fp+16383 0x1p-16445
clog -0x1.fp+16383 -0x1p-16445
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
clog 0x1.fp+16383 0x1p-16445
clog 0x1.fp+16383 -0x1p-16445
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
clog 0x1p-16494 0x1.fp+16383
clog -0x1p-16494 0x1.fp+16383
clog 0x1p-16494 -0x1.fp+16383
clog -0x1p-16494 -0x1.fp+16383
clog -0x1.fp+16383 0x1p-16494
clog -0x1.fp+16383 -0x1p-16494
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
clog 0x1.fp+16383 0x1p-16494
clog 0x1.fp+16383 -0x1p-16494
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
clog 1.0 0x1.234566p-10
clog -1.0 0x1.234566p-20
clog 0x1.234566p-30 1.0
clog -0x1.234566p-40 -1.0
clog 0x1.234566p-50 1.0
clog 0x1.234566p-60 1.0
clog 0x1p-62 1.0
clog 0x1p-63 1.0
clog 0x1p-64 1.0
clog 0x1p-510 1.0
clog 0x1p-511 1.0
clog 0x1p-512 1.0
clog 0x1p-8190 1.0
clog 0x1p-8191 1.0
clog 0x1p-8192 1.0
clog 0x1.000566p0 0x1.234p-10
clog 0x1.000566p0 0x1.234p-100
clog -0x1.0000000123456p0 0x1.2345678p-30
clog -0x1.0000000123456p0 0x1.2345678p-1000
clog 0x1.00000000000000123456789abcp0 0x1.23456789p-60
clog 0x1.00000000000000123456789abcp0 0x1.23456789p-1000
clog 0x0.ffffffp0 0x0.ffffffp-100
clog 0x0.fffffffffffff8p0 0x0.fffffffffffff8p-1000
clog 0x0.ffffffffffffffffp0 0x0.ffffffffffffffffp-15000
clog 0x1a6p-10 0x3a5p-10
clog 0xf2p-10 0x3e3p-10
clog 0x4d4ep-15 0x6605p-15
clog 0x2818p-15 0x798fp-15
clog 0x9b57bp-20 0xcb7b4p-20
clog 0x2731p-20 0xfffd0p-20
clog 0x2ede88p-23 0x771c3fp-23
clog 0x11682p-23 0x7ffed1p-23
clog 0xa1f2c1p-24 0xc643aep-24
clog 0x659feap-24 0xeaf6f9p-24
clog 0x4447d7175p-35 0x6c445e00ap-35
clog 0x2dd46725bp-35 0x7783a1284p-35
clog 0x164c74eea876p-45 0x16f393482f77p-45
clog 0xfe961079616p-45 0x1bc37e09e6d1p-45
clog 0xa4722f19346cp-51 0x7f9631c5e7f07p-51
clog 0x10673dd0f2481p-51 0x7ef1d17cefbd2p-51
clog 0x8ecbf810c4ae6p-52 0xd479468b09a37p-52
clog 0x5b06b680ea2ccp-52 0xef452b965da9fp-52
clog 0x659b70ab7971bp-53 0x1f5d111e08abecp-53
clog 0x15cfbd1990d1ffp-53 0x176a3973e09a9ap-53
clog 0x1367a310575591p-54 0x3cfcc0a0541f60p-54
clog 0x55cb6d0c83af5p-55 0x7fe33c0c7c4e90p-55
clog 0x298c62cb546588a7p-63 0x7911b1dfcc4ecdaep-63
clog 0x4d9c37e2b5cb4533p-63 0x65c98be2385a042ep-63
clog 0x602fd5037c4792efp-64 0xed3e2086dcca80b8p-64
clog 0x6b10b4f3520217b6p-64 0xe8893cbb449253a1p-64
clog 0x81b7efa81fc35ad1p-65 0x1ef4b835f1c79d812p-65
clog 0x3f96469050f650869c2p-75 0x6f16b2c9c8b05988335p-75
clog 0x3157fc1d73233e580c8p-75 0x761b52ccd435d7c7f5fp-75
clog 0x155f8afc4c48685bf63610p-85 0x17d0cf2652cdbeb1294e19p-85
clog 0x13836d58a13448d750b4b9p-85 0x195ca7bc3ab4f9161edbe6p-85
clog 0x1df515eb171a808b9e400266p-95 0x7c71eb0cd4688dfe98581c77p-95
clog 0xe33f66c9542ca25cc43c867p-95 0x7f35a68ebd3704a43c465864p-95
clog 0x6771f22c64ed551b857c128b4cp-105 0x1f570e7a13cc3cf2f44fd793ea1p-105
clog 0x15d8ab6ed05ca514086ac3a1e84p-105 0x1761e480aa094c0b10b34b09ce9p-105
clog 0x187190c1a334497bdbde5a95f48p-106 0x3b25f08062d0a095c4cfbbc338dp-106
clog 0x6241ef0da53f539f02fad67dabp-106 0x3fb46641182f7efd9caa769dac0p-106
clog 0x3e1d0a105ac4ebeacd9c6952d34cp-112 0xf859b3d1b06d005dcbb5516d5479p-112
clog 0x47017a2e36807acb1e5214b209dep-112 0xf5f4a550c9d75e3bb1839d865f0dp-112
clog 0x148f818cb7a9258fca942ade2a0cap-113 0x18854a34780b8333ec53310ad7001p-113
clog 0xfd95243681c055c2632286921092p-113 0x1bccabcd29ca2152860ec29e34ef7p-113
clog 0xdb85c467ee2aadd5f425fe0f4b8dp-114 0x3e83162a0f95f1dcbf97dddf410eap-114
clog 0x1415bcaf2105940d49a636e98ae59p-115 0x7e6a150adfcd1b0921d44b31f40f4p-115
Move tests of clog10 from libm-test.inc to auto-libm-test-in. This patch moves tests of clog10 to auto-libm-test-in. Note that this means gen-auto-libm-tests will now depend on the recent MPC 1.0.2 release which added a fix for a bug that made gen-auto-libm-tests hang for clog10. (It still can't conveniently be used for cacos cacosh casin casinh catan catanh csin csinh because of extreme slowness of those functions for special cases in MPC; at least some slow cases of csin / csinh are fixed in MPC trunk, but not in a release.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of clog10. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (clog10_test_data): Use AUTO_TESTS_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-02-19 14:26:29 +00:00
clog10 0.75 1.25
clog10 -2 -3
Fix dbl-64 atan2 in non-default rounding modes (bug 18210, bug 18211). The dbl-64 implementation of atan2 does computations that expect to run in round-to-nearest mode, and in other modes the errors can accumulate to more than the maximum accepted 9ulp. This patch makes it use FE_TONEAREST internally, similar to other functions with such issues. Tests that previously produced large errors are added for atan2 and the closely related carg, clog and clog10 functions. Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18210] [BZ #18211] * sysdeps/ieee754/dbl-64/e_atan2.c: Include <fenv.h>. (__ieee754_atan2): Set FE_TONEAREST mode for internal computations. * math/auto-libm-test-in: Add more tests of atan2, carg, clog and clog10. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-08 17:32:17 +00:00
clog10 0x2.f2f308p+0 0x4.c3841p-4
clog10 0xd.3de7ap-36 -0xe.cf143p-40
clog10 0x2.21e65p+0 0x5.576cf8p-4
clog10 0x1.f4755cp+0 -0x4.29411p-4
clog10 -0xf.9c4c8p-4 -0xa.b4101p+20
clog10 0x7.40ac68p+0 0x4.251bb8p-4
clog10 0xa.3ac3cp+68 0x1.47239ep+68
clog10 0x3.8ff10cp+0 -0x6.b0794p-4
Add more tests of clog and clog10. This patch adds some randomly-generated tests of clog and clog10 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of clog and clog10. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-09 22:14:00 +00:00
clog10 0x2.83f8ap+0 -0xb.0b529p-4
clog10 -0x2.eb21fcp-4 -0x6.59bbc8p-4
clog10 -0x3.3f7fc4p-4 0xb.ba599p-4
clog10 0x1.cd1ab2p-124 -0x8p-152
clog10 0xa.32054p-4 0x2.c7e71cp-4
clog10 -0x5.9ecf8c7b5a0f4p-4 0xa.a945e5f8761c8p-4
clog10 0x1.7a858p+0 -0x6.d940dp-4
clog10 -0x2.51320d99da5a2p-4 0x3.b8176p-4
clog10 -0x1.25c2d3e172df8p+0 0
clog10 0x1.0c684e35d0b2ap+0 -0x7.37df8a65c28fp-4
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
clog10 -0x9.93d164127d9fp-4 0x7.c5c8d8p-4
clog10 -0xa.5920ap-4 -0x6.2cda5p-4
clog10 0xd.d05c38ebb1b4p+60 -0x3.c22fdp+44
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
clog10 -0xa.19f8ec252c58d5p-4 0x7.d10cdec29a141538p-4
clog10 -0xa.7ac41a0b417cb8fp-4 -0x6.c5a32eaeedd4p-4
clog10 0x3.c16p-136 0x8p-152
clog10 -0x1.0a69de710590dp+0 -0x7.bc7e121e2b0d1088p-4
Fix clog, clog10 inaccuracy (bug 19016). For arguments with X^2 + Y^2 close to 1, clog and clog10 avoid large errors from log(hypot) by computing X^2 + Y^2 - 1 in a way that avoids cancellation error and then using log1p. However, the thresholds for using that approach still result in log being used on argument as large as sqrt(13/16) > 0.9, leading to significant errors, in some cases above the 9ulp maximum allowed in glibc libm. This patch arranges for the approach using log1p to be used in any cases where |X|, |Y| < 1 and X^2 + Y^2 >= 0.5 (with the existing allowance for cases where one of X and Y is very small), adjusting the __x2y2m1 functions to work with the wider range of inputs. This way, log only gets used on arguments below sqrt(1/2) (or substantially above 1), where the error involved is much less. Tested for x86_64, x86, mips64 and powerpc. For the ulps regeneration I removed the existing clog and clog10 ulps before regenerating to allow any reduced ulps to appear. Tests added include those found by random test generation to produce large ulps either before or after the patch, and some found by trying inputs close to the (0.75, 0.5) threshold where the potential errors from using log are largest. [BZ #19016] * sysdeps/generic/math_private.h (__x2y2m1f): Update comment to allow more cases with X^2 + Y^2 >= 0.5. * sysdeps/ieee754/dbl-64/x2y2m1.c (__x2y2m1): Likewise. Add -1 as normal element in sum instead of special-casing based on values of arguments. * sysdeps/ieee754/dbl-64/x2y2m1f.c (__x2y2m1f): Update comment. * sysdeps/ieee754/ldbl-128/x2y2m1l.c (__x2y2m1l): Likewise. Add -1 as normal element in sum instead of special-casing based on values of arguments. * sysdeps/ieee754/ldbl-128ibm/x2y2m1l.c (__x2y2m1l): Likewise. * sysdeps/ieee754/ldbl-96/x2y2m1.c [FLT_EVAL_METHOD != 0] (__x2y2m1): Update comment. * sysdeps/ieee754/ldbl-96/x2y2m1l.c (__x2y2m1l): Likewise. Add -1 as normal element in sum instead of special-casing based on values of arguments. * math/s_clog.c (__clog): Handle more cases using log1p without hypot. * math/s_clog10.c (__clog10): Likewise. * math/s_clog10f.c (__clog10f): Likewise. * math/s_clog10l.c (__clog10l): Likewise. * math/s_clogf.c (__clogf): Likewise. * math/s_clogl.c (__clogl): Likewise. * math/auto-libm-test-in: Add more tests of clog and clog10. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-28 22:11:22 +00:00
clog10 -0x2.7bdep-4 0x5.ab7a4p-4
clog10 -0xb.e1d3d0ff44358p-4 -0x7.54785e1b143f8p-4
clog10 0x3.ba473p+0 0x7.eea9ap-4
clog10 0x9.d02220baee4ep+36 0x2.b9a29cp+0
clog10 -0x5.1a5cf8p-4 -0xb.73012p-4
clog10 -0xa.ff292a609dbb8p-4 0x6.f73d4cp-4
clog10 -0x5.1a5cfc2301114p-4 -0xb.730118p-4
clog10 0xb.ffffcp-4 0x7.ffff1p-4
clog10 0xb.ffffp-4 0x7.ffffap-4
clog10 0xb.ffffp-4 0x7.fffff8p-4
clog10 0xb.ffffp-4 0x7.ffffp-4
clog10 0xb.fffffp-4 0x7.ffff68p-4
clog10 0xb.fffffp-4 0x7.ffffp-4
clog10 0xb.ffff8p-4 0x7.ffffcp-4
clog10 0xb.ffffp-4 0x7.ffffcp-4
clog10 0xb.ffffp-4 0x7.ffffb8p-4
clog10 0xb.ffffp-4 0x7.ffff7p-4
clog10 0xb.ffffp-4 0x7.ffff5p-4
clog10 0xb.fffffffffff7p-4 0x7.fffff8p-4
clog10 0xb.fffffffffff08p-4 0x7.fffffffffffdp-4
clog10 0xb.fffffffffff08p-4 0x7.fffffffffff9p-4
clog10 0xb.fffffffffffp-4 0x7.fffffffffffdcp-4
clog10 0xb.fffffp-4 0x7.ffffffffffff4p-4
clog10 0xb.fffffffffffp-4 0x7.fffffffffffecp-4
clog10 0xb.fffffffffff8p-4 0x7.fffff8p-4
clog10 0x8p-152 -0x1.10233ap+0
clog10 0xa.03634p-4 -0x4.7bb918p-20
clog10 -0x5.e23d2p-4 0x8.525df889c21ap-4
clog10 0x9.8ce58p-4 -0x8p-152
clog10 0x8p-152 0x9.2af75p-4
clog10 0x9.97a15de8e59d8p-4 -0
clog10 -0x4.74556ec92eb4746p-4 0x1.1e7aa1d936f6efe6p+0
clog10 0x9.97a15de8e59d8p-4 -0
clog10 -0x9.7f1d7p-64 0x9.db37dp-4
clog10 -0x8.5efc4p-4 -0x5.40310cp-4
clog10 -0x9.0b459p-4 0
clog10 -0x6.a9419e9b30e68p-4 -0x6.262c7p-4
clog10 0x5.2767cdfdfbf2p-4 0x7.69ee98p-4
clog10 -0x9.f5563cb3227d8p-4 0
clog10 -0x9.5a284p-4 0x6.899578p-8
clog10 0xa.3e62bp-4 0x1.18c03p-100
clog10 0 -0x9.22a99p-4
clog10 0 0x9.7915bp-4
clog10 0x3.00d1ap-12 0x1.23ff6ap+0
Move tests of clog10 from libm-test.inc to auto-libm-test-in. This patch moves tests of clog10 to auto-libm-test-in. Note that this means gen-auto-libm-tests will now depend on the recent MPC 1.0.2 release which added a fix for a bug that made gen-auto-libm-tests hang for clog10. (It still can't conveniently be used for cacos cacosh casin casinh catan catanh csin csinh because of extreme slowness of those functions for special cases in MPC; at least some slow cases of csin / csinh are fixed in MPC trunk, but not in a release.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of clog10. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (clog10_test_data): Use AUTO_TESTS_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-02-19 14:26:29 +00:00
clog10 0x1.fffffep+127 0x1.fffffep+127
clog10 0x1.fffffep+127 1.0
clog10 0x1p-149 0x1p-149
clog10 0x1p-147 0x1p-147
clog10 0x1.fffffffffffffp+1023 0x1.fffffffffffffp+1023
clog10 0x1.fffffffffffffp+1023 0x1p+1023
clog10 0x1p-1074 0x1p-1074
clog10 0x1p-1073 0x1p-1073
clog10 0x1.fp+16383 0x1.fp+16383
clog10 0x1.fp+16383 0x1p+16383
clog10 0x1p-16440 0x1p-16441
clog10 0x1p-149 0x1.fp+127
clog10 -0x1p-149 0x1.fp+127
clog10 0x1p-149 -0x1.fp+127
clog10 -0x1p-149 -0x1.fp+127
clog10 -0x1.fp+127 0x1p-149
clog10 -0x1.fp+127 -0x1p-149
clog10 0x1.fp+127 0x1p-149
clog10 0x1.fp+127 -0x1p-149
clog10 0x1p-1074 0x1.fp+1023
clog10 -0x1p-1074 0x1.fp+1023
clog10 0x1p-1074 -0x1.fp+1023
clog10 -0x1p-1074 -0x1.fp+1023
clog10 -0x1.fp+1023 0x1p-1074
clog10 -0x1.fp+1023 -0x1p-1074
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
clog10 0x1.fp+1023 0x1p-1074
clog10 0x1.fp+1023 -0x1p-1074
Move tests of clog10 from libm-test.inc to auto-libm-test-in. This patch moves tests of clog10 to auto-libm-test-in. Note that this means gen-auto-libm-tests will now depend on the recent MPC 1.0.2 release which added a fix for a bug that made gen-auto-libm-tests hang for clog10. (It still can't conveniently be used for cacos cacosh casin casinh catan catanh csin csinh because of extreme slowness of those functions for special cases in MPC; at least some slow cases of csin / csinh are fixed in MPC trunk, but not in a release.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of clog10. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (clog10_test_data): Use AUTO_TESTS_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-02-19 14:26:29 +00:00
clog10 0x1p-16445 0x1.fp+16383
clog10 -0x1p-16445 0x1.fp+16383
clog10 0x1p-16445 -0x1.fp+16383
clog10 -0x1p-16445 -0x1.fp+16383
clog10 -0x1.fp+16383 0x1p-16445
clog10 -0x1.fp+16383 -0x1p-16445
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
clog10 0x1.fp+16383 0x1p-16445
clog10 0x1.fp+16383 -0x1p-16445
Move tests of clog10 from libm-test.inc to auto-libm-test-in. This patch moves tests of clog10 to auto-libm-test-in. Note that this means gen-auto-libm-tests will now depend on the recent MPC 1.0.2 release which added a fix for a bug that made gen-auto-libm-tests hang for clog10. (It still can't conveniently be used for cacos cacosh casin casinh catan catanh csin csinh because of extreme slowness of those functions for special cases in MPC; at least some slow cases of csin / csinh are fixed in MPC trunk, but not in a release.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of clog10. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (clog10_test_data): Use AUTO_TESTS_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-02-19 14:26:29 +00:00
clog10 0x1p-16494 0x1.fp+16383
clog10 -0x1p-16494 0x1.fp+16383
clog10 0x1p-16494 -0x1.fp+16383
clog10 -0x1p-16494 -0x1.fp+16383
clog10 -0x1.fp+16383 0x1p-16494
clog10 -0x1.fp+16383 -0x1p-16494
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
clog10 0x1.fp+16383 0x1p-16494
clog10 0x1.fp+16383 -0x1p-16494
Move tests of clog10 from libm-test.inc to auto-libm-test-in. This patch moves tests of clog10 to auto-libm-test-in. Note that this means gen-auto-libm-tests will now depend on the recent MPC 1.0.2 release which added a fix for a bug that made gen-auto-libm-tests hang for clog10. (It still can't conveniently be used for cacos cacosh casin casinh catan catanh csin csinh because of extreme slowness of those functions for special cases in MPC; at least some slow cases of csin / csinh are fixed in MPC trunk, but not in a release.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of clog10. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (clog10_test_data): Use AUTO_TESTS_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-02-19 14:26:29 +00:00
clog10 1.0 0x1.234566p-10
clog10 -1.0 0x1.234566p-20
clog10 0x1.234566p-30 1.0
clog10 -0x1.234566p-40 -1.0
clog10 0x1.234566p-50 1.0
clog10 0x1.234566p-60 1.0
clog10 0x1p-61 1.0
clog10 0x1p-62 1.0
clog10 0x1p-63 1.0
clog10 0x1p-509 1.0
clog10 0x1p-510 1.0
clog10 0x1p-511 1.0
clog10 0x1p-8189 1.0
clog10 0x1p-8190 1.0
clog10 0x1p-8191 1.0
clog10 0x1.000566p0 0x1.234p-10
clog10 0x1.000566p0 0x1.234p-100
clog10 -0x1.0000000123456p0 0x1.2345678p-30
clog10 -0x1.0000000123456p0 0x1.2345678p-1000
clog10 0x1.00000000000000123456789abcp0 0x1.23456789p-60
clog10 0x1.00000000000000123456789abcp0 0x1.23456789p-1000
clog10 0x0.ffffffp0 0x0.ffffffp-100
clog10 0x0.fffffffffffff8p0 0x0.fffffffffffff8p-1000
clog10 0x0.ffffffffffffffffp0 0x0.ffffffffffffffffp-15000
clog10 0x1a6p-10 0x3a5p-10
clog10 0xf2p-10 0x3e3p-10
clog10 0x4d4ep-15 0x6605p-15
clog10 0x2818p-15 0x798fp-15
clog10 0x9b57bp-20 0xcb7b4p-20
clog10 0x2731p-20 0xfffd0p-20
clog10 0x2ede88p-23 0x771c3fp-23
clog10 0x11682p-23 0x7ffed1p-23
clog10 0xa1f2c1p-24 0xc643aep-24
clog10 0x659feap-24 0xeaf6f9p-24
clog10 0x4447d7175p-35 0x6c445e00ap-35
clog10 0x2dd46725bp-35 0x7783a1284p-35
clog10 0x164c74eea876p-45 0x16f393482f77p-45
clog10 0xfe961079616p-45 0x1bc37e09e6d1p-45
clog10 0xa4722f19346cp-51 0x7f9631c5e7f07p-51
clog10 0x10673dd0f2481p-51 0x7ef1d17cefbd2p-51
clog10 0x8ecbf810c4ae6p-52 0xd479468b09a37p-52
clog10 0x5b06b680ea2ccp-52 0xef452b965da9fp-52
clog10 0x659b70ab7971bp-53 0x1f5d111e08abecp-53
clog10 0x15cfbd1990d1ffp-53 0x176a3973e09a9ap-53
clog10 0x1367a310575591p-54 0x3cfcc0a0541f60p-54
clog10 0x55cb6d0c83af5p-55 0x7fe33c0c7c4e90p-55
clog10 0x298c62cb546588a7p-63 0x7911b1dfcc4ecdaep-63
clog10 0x4d9c37e2b5cb4533p-63 0x65c98be2385a042ep-63
clog10 0x602fd5037c4792efp-64 0xed3e2086dcca80b8p-64
clog10 0x6b10b4f3520217b6p-64 0xe8893cbb449253a1p-64
clog10 0x81b7efa81fc35ad1p-65 0x1ef4b835f1c79d812p-65
clog10 0x3f96469050f650869c2p-75 0x6f16b2c9c8b05988335p-75
clog10 0x3157fc1d73233e580c8p-75 0x761b52ccd435d7c7f5fp-75
clog10 0x155f8afc4c48685bf63610p-85 0x17d0cf2652cdbeb1294e19p-85
clog10 0x13836d58a13448d750b4b9p-85 0x195ca7bc3ab4f9161edbe6p-85
clog10 0x1df515eb171a808b9e400266p-95 0x7c71eb0cd4688dfe98581c77p-95
clog10 0xe33f66c9542ca25cc43c867p-95 0x7f35a68ebd3704a43c465864p-95
clog10 0x6771f22c64ed551b857c128b4cp-105 0x1f570e7a13cc3cf2f44fd793ea1p-105
clog10 0x15d8ab6ed05ca514086ac3a1e84p-105 0x1761e480aa094c0b10b34b09ce9p-105
clog10 0x187190c1a334497bdbde5a95f48p-106 0x3b25f08062d0a095c4cfbbc338dp-106
clog10 0x6241ef0da53f539f02fad67dabp-106 0x3fb46641182f7efd9caa769dac0p-106
clog10 0x3e1d0a105ac4ebeacd9c6952d34cp-112 0xf859b3d1b06d005dcbb5516d5479p-112
clog10 0x47017a2e36807acb1e5214b209dep-112 0xf5f4a550c9d75e3bb1839d865f0dp-112
clog10 0x148f818cb7a9258fca942ade2a0cap-113 0x18854a34780b8333ec53310ad7001p-113
clog10 0xfd95243681c055c2632286921092p-113 0x1bccabcd29ca2152860ec29e34ef7p-113
clog10 0xdb85c467ee2aadd5f425fe0f4b8dp-114 0x3e83162a0f95f1dcbf97dddf410eap-114
clog10 0x1415bcaf2105940d49a636e98ae59p-115 0x7e6a150adfcd1b0921d44b31f40f4p-115
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
cos 0
cos -0
cos pi/3
cos 2pi/3
cos pi/2
cos 0.75
cos 0x1p65
cos -0x1p65
cos 0.80190127184058835
cos 0x1.442f74p+15
cos 1e22
cos 0x1p1023
cos 0x1p16383
cos 0x1p+120
cos 0x1p+127
cos 0x1.fffff8p+127
cos 0x1.fffffep+127
cos 0x1p+50
cos 0x1p+28
cos 0x1.000000cf4a2a2p0
cos 0x1.0000010b239a9p0
cos 0x1.00000162a932bp0
cos 0x1.000002d452a10p0
cos 0x1.000005bc7d86dp0
Fix sign of input to bsloww1 (BZ #16623) In 84ba214c, I removed some redundant sign computations and in the process, I incorrectly got rid of a temporary variable, thus passing the absolute value of the input to bsloww1. This caused #16623. This fix undoes the incorrect change.
2014-02-27 15:42:09 +00:00
cos 0x1.200145a975ce6p32
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
cos 1
cos 2
cos 3
cos 4
cos 5
cos 6
cos 7
cos 8
cos 9
cos 10
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
cos 0x1p-5
cos 0x1p-10
cos 0x1p-15
cos 0x1p-20
cos 0x1p-25
cos 0x1p-30
cos 0x1p-35
cos 0x1p-40
cos 0x1p-45
cos 0x1p-50
cos 0x1p-55
cos 0x1p-60
cos 0x1p-100
cos 0x1p-600
cos 0x1p-10000
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
cos max
cos -max
cos min
cos -min
cos min_subnorm
cos -min_subnorm
Add more tests of cos, sin, sincos. This patch adds some randomly-generated tests of cos, sin and sincos that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cos, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 22:41:00 +00:00
cos -0x3.3de320f6be87ep+1020
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
cos 0xe.9f1e5bc3bb88p+112
Add more tests of acosh, atanh, cos, csqrt, erfc, sin, sincos. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, cos, csqrt, erfc, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-06 17:30:18 +00:00
cos 0x4.7857dp+68
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
cos -0x1.02e34cp+0
cos 0xf.f0274p+4
cos 0x3.042d88p+0
Fix cos computation for multiple precision fallback (bz #20357) During the sincos consolidation I made two mistakes, one was a logical error due to which cos(0x1.8475e5afd4481p+0) returned sin(0x1.8475e5afd4481p+0) instead. The second issue was an error in negating inputs for the correct quadrants for sine. I could not find a suitable test case for this despite running a program to search for such an input for a couple of hours. Following patch fixes both issues. Tested on x86_64. Thanks to Matt Clay for identifying the issue. [BZ #20357] * sysdeps/ieee754/dbl-64/s_sin.c (sloww): Fix up condition to call __mpsin/__mpcos and to negate values. * math/auto-libm-test-in: Add test. * math/auto-libm-test-out: Regenerate.
2016-07-18 17:03:09 +00:00
cos 0x1.8475e5afd4481p+0
Improve float range reduction accuracy near pi/2 (bug 21094). Bug 21094 reports 3ulp errors of cosf and tanf for certain arguments near pi/2 arising from the use of an insufficiently accurate range reduction. (To be clear, this is a quality-of-implementation issue relating to the apparent intent of those particular cosf and tanf implementations; 3ulp is within the general glibc accuracy goals, so not inherently a bug.) This patch fixes that error by making a wider range of cases use the existing more accurate range reduction for arguments close to pi/2. The wider range of values is still narrow enough for the "z -= pio2_2;" in the more accurate case to be exact, as the code expects. Tested for x86_64, x86 and mips64; no ulps updates needed (but at least on mips64, the larger ulps were seen if the tests were added without the substantive fix). [BZ #21094] * sysdeps/ieee754/flt-32/e_rem_pio2f.c (__ieee754_rem_pio2f): Use 24+24+24-bit pi for wider range of values around pi/2. * math/auto-libm-test-in: Add more tests of cos and tan. * math/auto-libm-test-out-cos: Regenerated. * math/auto-libm-test-out-tan: Likewise.
2017-03-15 22:00:54 +00:00
cos 1.57079697
cos -1.57079697
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
cos 0x2.3c6ef4p-12
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary128)
cos 0xe.6672d458b05edf50af4fab1a42p+40
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
cosh 0
cosh -0
cosh 0.75
cosh 709.8893558127259666434838436543941497802734375
cosh -709.8893558127259666434838436543941497802734375
cosh 22
cosh 23
cosh 24
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
cosh 0x1p-5
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
cosh 0x1p-10
cosh 0x1p-15
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
cosh 0x1p-20
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
cosh 0x1p-25
cosh 0x1p-30
cosh 0x1p-35
cosh 0x1p-40
cosh 0x1p-45
cosh 0x1p-50
cosh 0x1p-100
cosh 0x1p-600
cosh 0x1p-10000
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
cosh -1
cosh 50
Add more tests of cosh, sinh. This patch adds some randomly-generated tests of cosh and sinh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cosh and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-24 23:48:04 +00:00
cosh -0xb.60713p+0
Add further tests of cosh and sinh. This patch adds some further randomly-generated tests of cosh and sinh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cosh and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-30 22:32:08 +00:00
cosh -0x3.cee48p+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
cosh 0x2.f5d128p+0
cosh -0xd.0c03p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
cosh -0x3.d04328728b72cp-4
Add more randomly-generated libm tests. This patch adds more libm test inputs found through random test generation to increase observed ulps on x86_64. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, atanh, cbrt, cosh, csqrt, erfc, expm1 and lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-11 15:03:10 +00:00
cosh 0x1.629188p+4
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
cosh max
cosh -max
Fix cosh spurious underflows from expm1 (bug 16354), inaccurate results near 0 (bug 17061). This patch fixes bug 16354, spurious underflows from cosh when a tiny argument is passed to expm1 and expm1 correctly underflows although the final result of cosh should be 1. As noted in that bug, some cases are latent because of expm1 implementations not raising underflow (bug 16353), but all the implementations are fixed similarly. They already contained checks for tiny arguments, but the checks were too late to avoid underflow from expm1 (although they would avoid underflow from subsequent squaring of the result of expm1); they are moved before the expm1 calls. The thresholds used for considering arguments tiny are not particularly consistent in how they relate to the precision of the floating-point format in question. They are, however, all sufficient to ensure that the round-to-nearest result of cosh is indeed 1 below the threshold (although sometimes they are smaller than necessary). But the previous logic did not return 1, but the previously computed 1 + expm1(abs(x)) value. And the thresholds in the ldbl-128 and ldbl-128ibm code (0x1p-71L - I suspect 0x3f8b was intended in the code instead of 0x3fb8 - and (roughly) 0x1p-55L) are not sufficient for that value to be 1. So by moving the test for tiny arguments, and consequently returning 1 directly now the expm1 value hasn't been computed by that point, this patch also fixes bug 17061, the (large number of ulps) inaccuracy for small arguments in those implementations. Tests for that bug are duly added. Tested x86_64 and x86 and ulps updated accordingly. Also tested for mips64 and powerpc32 to validate the ldbl-128 and ldbl-128ibm changes. [BZ #16354] [BZ #17061] * sysdeps/ieee754/dbl-64/e_cosh.c (__ieee754_cosh): Check for small arguments before calling __expm1. * sysdeps/ieee754/flt-32/e_coshf.c (__ieee754_coshf): Check for small arguments before calling __expm1f. * sysdeps/ieee754/ldbl-128/e_coshl.c (__ieee754_coshl): Check for small arguments before calling __expm1l. * sysdeps/ieee754/ldbl-128ibm/e_coshl.c (__ieee754_coshl): Likewise. * sysdeps/ieee754/ldbl-96/e_coshl.c (__ieee754_coshl): Likewise. * math/auto-libm-test-in: Add more cosh tests. Do not allow spurious underflow for some cosh tests. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update.
2014-06-23 20:20:10 +00:00
cosh min
cosh -min
cosh min_subnorm
cosh -min_subnorm
cosh 0x1p-56
cosh -0x1p-56
cosh 0x1p-72
cosh -0x1p-72
Add more libm-test coverage of [a-c]* real functions. Various libm functions have inadequate test coverage in libm-test.inc / auto-libm-test-in - failing to cover all the usual special cases (infinities, NaNs, zero, large and small finite values, subnormals) as well as a reasonable range of ordinary inputs and, where appropriate, inputs close to the thresholds for underflow and overflow. This patch improves test coverage for real functions [a-c]* (with the expectation of adding more coverage for other functions later). Tested x86_64 and x86 and ulps updated accordingly (and eight glibc bugs and one C11 DR filed for issues found in the process). * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cbrt, cos and cosh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acosh_test_data): Add more tests. (atanh_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 21:03:39 +00:00
# Test values either side of overflow for each floating-point format.
cosh 0x5.96a7ep+4
cosh 0x5.96a7e8p+4
cosh -0x5.96a7ep+4
cosh -0x5.96a7e8p+4
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
cosh 0x2.c679d1f73f0fap+8
cosh 0x2.c679d1f73f0fcp+8
cosh -0x2.c679d1f73f0fap+8
cosh -0x2.c679d1f73f0fcp+8
Update libm-test XFAILs for ibm128 format. This patch cleans up and updates the libm-test XFAILs for the ibm128 format. More of them are changed to use a new ibm128-libgcc conditional, to reflect that they are not in fact needed if you've patched libgcc to fix the known issues (at substantial performance cost). Many additional XFAILs are added for tests that fail with unpatched libgcc (most but not all of them xfail-rounding). Note that further such fixes will be needed for test-ldouble actually to pass with default libgcc (in particular, XFAILs for pow tests and for various affected tests directly embedded in libm-test.inc). With patched libgcc, there may be a few XFAILs needed but the results are already substantially clean apart from a few ulps differences. Tested for powerpc. * math/libm-test.inc (TEST_COND_ibm128_libgcc): New macro. (init_max_error) [TEST_COND_ibm128]: Increase maximum error allowed to 16 ulps. * math/auto-libm-test-in: Change most XFAILs for ibm128 to use ibm128-libgcc. XFAIL more tests for ibm128-libgcc. * math/auto-libm-test-out: Regenerated.
2017-01-04 23:29:42 +00:00
cosh 0x2.c679d1f73f0fb624d358b213a7p+8 xfail-rounding:ibm128-libgcc
cosh 0x2.c679d1f73f0fb624d358b213a8p+8 xfail-rounding:ibm128-libgcc
cosh -0x2.c679d1f73f0fb624d358b213a7p+8 xfail-rounding:ibm128-libgcc
cosh -0x2.c679d1f73f0fb624d358b213a8p+8 xfail-rounding:ibm128-libgcc
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
cosh 0x2.c5d37700c6bb03a4p+12
cosh 0x2.c5d37700c6bb03a8p+12
cosh -0x2.c5d37700c6bb03a4p+12
cosh -0x2.c5d37700c6bb03a8p+12
cosh 0x2.c5d37700c6bb03a6c24b6c9b494cp+12
cosh 0x2.c5d37700c6bb03a6c24b6c9b494ep+12
cosh -0x2.c5d37700c6bb03a6c24b6c9b494cp+12
cosh -0x2.c5d37700c6bb03a6c24b6c9b494ep+12
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
cosh 0x5.8bfe6p+4
add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)
2020-11-25 18:03:33 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
cosh 0x1.633c457f1e66bp+9
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (ldbl-96)
cosh 0x2.c5d376167f4052f4p+12
# the next value generates larger error bounds on x86_64 (binary128)
cosh -0x2.c5d376eefcd4bbeb000452d84662p+12
Move TEST_f_f tests for [a-c]* functions from libm-test.inc to auto-libm-test-in.
2013-11-30 22:04:13 +00:00
Test cpow in all rounding modes. This patch enables testing of cpow in all rounding modes using ALL_RM_TEST. There were two reasons this was previously deferred: * MPC has complicated rounding-mode-dependent rules for the signs of exact zero real or imaginary parts in the result of mpc_pow. Annex G does not impose any such requirements and I don't think glibc should try to implement any particular logic here. This patch adds support for gen-auto-libm-tests passing the IGNORE_ZERO_INF_SIGN flag to libm-test.inc. * Error accumulations in some tests in non-default rounding modes exceed the maximum error permitted in libm-test.inc. This patch marks the problem tests with xfail-rounding. (It might be possible to reduce the accumulations a bit by using round-to-nearest when cpow calls clog, but I don't think there's much point; the implementation approach for cpow is fundamentally deficient, as discussed in the existing bug for cpow inaccuracy which can reasonably be considered to cover these less-inaccurate cases as well. It's possible that the test "cpow 2 0 10 0" will also need xfail-rounding on some platforms.) Tested x86_64 and x86 and ulps updated accordingly. * math/gen-auto-libm-tests.c: Document use of ignore-zero-inf-sign. (input_flag_type): Add value flag_ignore_zero_inf_sign. (input_flags): Add ignore-zero-inf-sign. (output_for_one_input_case): Handle flag_ignore_zero_inf_sign. * math/gen-libm-test.pl (generate_testfile): Handle ignore-zero-inf-sign. * math/auto-libm-test-in: Mark some cpow tests with ignore-zero-inf-sign and some with xfail-rounding. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cpow_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-06-23 20:15:14 +00:00
cpow 1 0 0 0 ignore-zero-inf-sign
cpow 2 0 10 0 ignore-zero-inf-sign
Move tests of cpow from libm-test.inc to auto-libm-test-in. This patch moves tests of cpow to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of cpow. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cpow_test_data): Use AUTO_TESTS_cc_c. * * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_cc_c. (func_calc_desc): Add mpc_cc_c union field. (test_functions): Add cpow. (special_fill_2pi): New function. (special_real_inputs): Add 2pi. (calc_generic_results): Handle mpc_cc_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:35:10 +00:00
# Bug 14473: cpow results inaccurate.
cpow e 0 0 2pi xfail
Test cpow in all rounding modes. This patch enables testing of cpow in all rounding modes using ALL_RM_TEST. There were two reasons this was previously deferred: * MPC has complicated rounding-mode-dependent rules for the signs of exact zero real or imaginary parts in the result of mpc_pow. Annex G does not impose any such requirements and I don't think glibc should try to implement any particular logic here. This patch adds support for gen-auto-libm-tests passing the IGNORE_ZERO_INF_SIGN flag to libm-test.inc. * Error accumulations in some tests in non-default rounding modes exceed the maximum error permitted in libm-test.inc. This patch marks the problem tests with xfail-rounding. (It might be possible to reduce the accumulations a bit by using round-to-nearest when cpow calls clog, but I don't think there's much point; the implementation approach for cpow is fundamentally deficient, as discussed in the existing bug for cpow inaccuracy which can reasonably be considered to cover these less-inaccurate cases as well. It's possible that the test "cpow 2 0 10 0" will also need xfail-rounding on some platforms.) Tested x86_64 and x86 and ulps updated accordingly. * math/gen-auto-libm-tests.c: Document use of ignore-zero-inf-sign. (input_flag_type): Add value flag_ignore_zero_inf_sign. (input_flags): Add ignore-zero-inf-sign. (output_for_one_input_case): Handle flag_ignore_zero_inf_sign. * math/gen-libm-test.pl (generate_testfile): Handle ignore-zero-inf-sign. * math/auto-libm-test-in: Mark some cpow tests with ignore-zero-inf-sign and some with xfail-rounding. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cpow_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-06-23 20:15:14 +00:00
cpow 2 3 4 0 xfail-rounding
Move tests of cpow from libm-test.inc to auto-libm-test-in. This patch moves tests of cpow to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of cpow. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cpow_test_data): Use AUTO_TESTS_cc_c. * * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_cc_c. (func_calc_desc): Add mpc_cc_c union field. (test_functions): Add cpow. (special_fill_2pi): New function. (special_real_inputs): Add 2pi. (calc_generic_results): Handle mpc_cc_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:35:10 +00:00
Test cpow in all rounding modes. This patch enables testing of cpow in all rounding modes using ALL_RM_TEST. There were two reasons this was previously deferred: * MPC has complicated rounding-mode-dependent rules for the signs of exact zero real or imaginary parts in the result of mpc_pow. Annex G does not impose any such requirements and I don't think glibc should try to implement any particular logic here. This patch adds support for gen-auto-libm-tests passing the IGNORE_ZERO_INF_SIGN flag to libm-test.inc. * Error accumulations in some tests in non-default rounding modes exceed the maximum error permitted in libm-test.inc. This patch marks the problem tests with xfail-rounding. (It might be possible to reduce the accumulations a bit by using round-to-nearest when cpow calls clog, but I don't think there's much point; the implementation approach for cpow is fundamentally deficient, as discussed in the existing bug for cpow inaccuracy which can reasonably be considered to cover these less-inaccurate cases as well. It's possible that the test "cpow 2 0 10 0" will also need xfail-rounding on some platforms.) Tested x86_64 and x86 and ulps updated accordingly. * math/gen-auto-libm-tests.c: Document use of ignore-zero-inf-sign. (input_flag_type): Add value flag_ignore_zero_inf_sign. (input_flags): Add ignore-zero-inf-sign. (output_for_one_input_case): Handle flag_ignore_zero_inf_sign. * math/gen-libm-test.pl (generate_testfile): Handle ignore-zero-inf-sign. * math/auto-libm-test-in: Mark some cpow tests with ignore-zero-inf-sign and some with xfail-rounding. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cpow_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-06-23 20:15:14 +00:00
cpow 0.75 1.25 0.75 1.25 xfail-rounding
cpow 0.75 1.25 1.0 1.0 xfail-rounding
Move tests of cpow from libm-test.inc to auto-libm-test-in. This patch moves tests of cpow to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of cpow. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cpow_test_data): Use AUTO_TESTS_cc_c. * * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_cc_c. (func_calc_desc): Add mpc_cc_c union field. (test_functions): Add cpow. (special_fill_2pi): New function. (special_real_inputs): Add 2pi. (calc_generic_results): Handle mpc_cc_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:35:10 +00:00
cpow 0.75 1.25 1.0 0.0
cpow 0.75 1.25 0.0 1.0
Move csin, csinh tests to auto-libm-test-in. This patch moves most tests of csin and csinh with finite inputs from libm-test.inc to auto-libm-test-in. The remaining two tests of each function with small arguments are not moved because moving them causes the time required by gen-auto-libm-tests to go up from under 8 seconds to over 11 minutes for me. (The current development version of MPC has had speed improvements for mpc_sin for some time, but there hasn't been a release containing those improvements yet.) Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Remove tests moved to auto-libm-test-in. (csinh_test_data): Likewise.
2015-06-24 16:37:44 +00:00
csin 0.0 0.0
csin -0 0.0
csin 0.0 -0
csin -0 -0
csin 0.75 1.25
csin -2 -3
csin 0.75 89.5
csin 0.75 -89.5
csin -0.75 89.5
csin -0.75 -89.5
csin 0.75 710.5
csin 0.75 -710.5
csin -0.75 710.5
csin -0.75 -710.5
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
csin 0.75 11357.25
csin 0.75 -11357.25
csin -0.75 11357.25
csin -0.75 -11357.25
Move csin, csinh tests to auto-libm-test-in. This patch moves most tests of csin and csinh with finite inputs from libm-test.inc to auto-libm-test-in. The remaining two tests of each function with small arguments are not moved because moving them causes the time required by gen-auto-libm-tests to go up from under 8 seconds to over 11 minutes for me. (The current development version of MPC has had speed improvements for mpc_sin for some time, but there hasn't been a release containing those improvements yet.) Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Remove tests moved to auto-libm-test-in. (csinh_test_data): Likewise.
2015-06-24 16:37:44 +00:00
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
csin 0.75 1e6
csin 0.75 -1e6
csin -0.75 1e6
csin -0.75 -1e6
Fix csin, csinh overflow in directed rounding modes (bug 18593). csin and csinh can produce bad results when overflowing in directed rounding modes, because a multiplication that can overflow is followed by a possible negation. This patch fixes this by negating one of the arguments of the multiplication before the multiplication instead of negating the result. The new tests for this issue are added to auto-libm-test-in, starting use of that file for csin and csinh. The issue was found in the course of moving existing tests for csin and csinh (existing tests, by being enabled in more cases than previously, showed the issue for float and double but not for long double); that move will now be done separately. Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18593] * math/s_csin.c (__csin): Negate before rather than after possibly overflowing multiplication. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/auto-libm-test-in: Add some tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Use AUTO_TESTS_c_c. (csinh_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-06-24 16:20:48 +00:00
Move csin, csinh tests to auto-libm-test-in. This patch moves most tests of csin and csinh with finite inputs from libm-test.inc to auto-libm-test-in. The remaining two tests of each function with small arguments are not moved because moving them causes the time required by gen-auto-libm-tests to go up from under 8 seconds to over 11 minutes for me. (The current development version of MPC has had speed improvements for mpc_sin for some time, but there hasn't been a release containing those improvements yet.) Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Remove tests moved to auto-libm-test-in. (csinh_test_data): Likewise.
2015-06-24 16:37:44 +00:00
csin 0x1p-149 180
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
csin 0x1p-1074 1440
csin 0x1p-16434 22730
Move csin, csinh tests to auto-libm-test-in. This patch moves most tests of csin and csinh with finite inputs from libm-test.inc to auto-libm-test-in. The remaining two tests of each function with small arguments are not moved because moving them causes the time required by gen-auto-libm-tests to go up from under 8 seconds to over 11 minutes for me. (The current development version of MPC has had speed improvements for mpc_sin for some time, but there hasn't been a release containing those improvements yet.) Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Remove tests moved to auto-libm-test-in. (csinh_test_data): Likewise.
2015-06-24 16:37:44 +00:00
Fix cexp, ccos, ccosh, csin, csinh spurious underflows (bug 18594). cexp, ccos, ccosh, csin and csinh have spurious underflows in cases where they compute sin of the smallest normal, that produces an underflow exception (depending on which sin implementation is in use) but the final result does not underflow. ctan and ctanh may also have such underflows, or they may be latent (the issue there is that e.g. ctan (DBL_MIN) should, rounded upwards, be the next double value above DBL_MIN, which under glibc's accuracy goals may not have an underflow exception, but the intermediate computation of sin (DBL_MIN) would legitimately underflow on before-rounding architectures). This patch fixes all those functions so they use plain comparisons (> DBL_MIN etc.) instead of comparing the result of fpclassify with FP_SUBNORMAL (in all these cases, we already know the number being compared is finite). Note that in the case of csin / csinf / csinl, there is no need for fabs calls in the comparison because the real part has already been reduced to its absolute value. As the patch fixes the failures that previously obstructed moving tests of cexp to use ALL_RM_TEST, those tests are moved to ALL_RM_TEST by the patch (two functions remain yet to be converted). Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18594] * math/s_ccosh.c (__ccosh): Compare with least normal value instead of comparing class with FP_SUBNORMAL. * math/s_ccoshf.c (__ccoshf): Likewise. * math/s_ccoshl.c (__ccoshl): Likewise. * math/s_cexp.c (__cexp): Likewise. * math/s_cexpf.c (__cexpf): Likewise. * math/s_cexpl.c (__cexpl): Likewise. * math/s_csin.c (__csin): Likewise. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/s_ctan.c (__ctan): Likewise. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Add more tests of ccos, ccosh, cexp, csin, csinh, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cexp_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-24 21:04:51 +00:00
csin min 1
csin -min 1
csin min_subnorm 80
csin -min_subnorm 80
Move more csin, csinh tests to auto-libm-test-in. When I moved tests of csin and csinh to auto-libm-test-in, I didn't move a few tests for which gen-auto-libm-tests was very slow because of MPC slowness on certain inputs. Now that auto-libm-test-out has been split up, such slowness only affects regenerating the test expectations for the individual functions in question, rather than any addition of tests for any function to auto-libm-test-in. Thus, I no longer consider it a problem to have these inputs in auto-libm-test-in, and this patch moves them there. This results in test generation for csin and csinh taking 5m43s (for each of csin and csinh) on my system, while other functions are unaffected. I expect the test generation to be much faster in MPC 1.1 (the relevant performance improvements went in MPC mainline in Dec 2013, but there hasn't been a release from mainline since then). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out-csin: Regenerated. * math/auto-libm-test-out-csinh: Likewise. * math/libm-test-csin.inc (csin_test_data): Remove tests moved to auto-libm-test-in. * math/libm-test-csinh.inc (csinh_test_data): Likewise.
2017-02-09 18:30:41 +00:00
csin min_subnorm min
csin min min_subnorm
Move csin, csinh tests to auto-libm-test-in. This patch moves most tests of csin and csinh with finite inputs from libm-test.inc to auto-libm-test-in. The remaining two tests of each function with small arguments are not moved because moving them causes the time required by gen-auto-libm-tests to go up from under 8 seconds to over 11 minutes for me. (The current development version of MPC has had speed improvements for mpc_sin for some time, but there hasn't been a release containing those improvements yet.) Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Remove tests moved to auto-libm-test-in. (csinh_test_data): Likewise.
2015-06-24 16:37:44 +00:00
csinh 0.0 0.0
csinh -0 0.0
csinh 0.0 -0
csinh -0 -0
csinh 0.75 1.25
csinh -2 -3
csinh 89.5 0.75
csinh -89.5 0.75
csinh 89.5 -0.75
csinh -89.5 -0.75
csinh 710.5 0.75
csinh -710.5 0.75
csinh 710.5 -0.75
csinh -710.5 -0.75
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
csinh 11357.25 0.75
csinh -11357.25 0.75
csinh 11357.25 -0.75
csinh -11357.25 -0.75
Move csin, csinh tests to auto-libm-test-in. This patch moves most tests of csin and csinh with finite inputs from libm-test.inc to auto-libm-test-in. The remaining two tests of each function with small arguments are not moved because moving them causes the time required by gen-auto-libm-tests to go up from under 8 seconds to over 11 minutes for me. (The current development version of MPC has had speed improvements for mpc_sin for some time, but there hasn't been a release containing those improvements yet.) Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Remove tests moved to auto-libm-test-in. (csinh_test_data): Likewise.
2015-06-24 16:37:44 +00:00
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
csinh 1e6 0.75
csinh -1e6 0.75
csinh 1e6 -0.75
csinh -1e6 -0.75
Fix csin, csinh overflow in directed rounding modes (bug 18593). csin and csinh can produce bad results when overflowing in directed rounding modes, because a multiplication that can overflow is followed by a possible negation. This patch fixes this by negating one of the arguments of the multiplication before the multiplication instead of negating the result. The new tests for this issue are added to auto-libm-test-in, starting use of that file for csin and csinh. The issue was found in the course of moving existing tests for csin and csinh (existing tests, by being enabled in more cases than previously, showed the issue for float and double but not for long double); that move will now be done separately. Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18593] * math/s_csin.c (__csin): Negate before rather than after possibly overflowing multiplication. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/auto-libm-test-in: Add some tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Use AUTO_TESTS_c_c. (csinh_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-06-24 16:20:48 +00:00
Move csin, csinh tests to auto-libm-test-in. This patch moves most tests of csin and csinh with finite inputs from libm-test.inc to auto-libm-test-in. The remaining two tests of each function with small arguments are not moved because moving them causes the time required by gen-auto-libm-tests to go up from under 8 seconds to over 11 minutes for me. (The current development version of MPC has had speed improvements for mpc_sin for some time, but there hasn't been a release containing those improvements yet.) Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Remove tests moved to auto-libm-test-in. (csinh_test_data): Likewise.
2015-06-24 16:37:44 +00:00
csinh 180 0x1p-149
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
csinh 1440 0x1p-1074
csinh 22730 0x1p-16434
Move csin, csinh tests to auto-libm-test-in. This patch moves most tests of csin and csinh with finite inputs from libm-test.inc to auto-libm-test-in. The remaining two tests of each function with small arguments are not moved because moving them causes the time required by gen-auto-libm-tests to go up from under 8 seconds to over 11 minutes for me. (The current development version of MPC has had speed improvements for mpc_sin for some time, but there hasn't been a release containing those improvements yet.) Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (csin_test_data): Remove tests moved to auto-libm-test-in. (csinh_test_data): Likewise.
2015-06-24 16:37:44 +00:00
Fix cexp, ccos, ccosh, csin, csinh spurious underflows (bug 18594). cexp, ccos, ccosh, csin and csinh have spurious underflows in cases where they compute sin of the smallest normal, that produces an underflow exception (depending on which sin implementation is in use) but the final result does not underflow. ctan and ctanh may also have such underflows, or they may be latent (the issue there is that e.g. ctan (DBL_MIN) should, rounded upwards, be the next double value above DBL_MIN, which under glibc's accuracy goals may not have an underflow exception, but the intermediate computation of sin (DBL_MIN) would legitimately underflow on before-rounding architectures). This patch fixes all those functions so they use plain comparisons (> DBL_MIN etc.) instead of comparing the result of fpclassify with FP_SUBNORMAL (in all these cases, we already know the number being compared is finite). Note that in the case of csin / csinf / csinl, there is no need for fabs calls in the comparison because the real part has already been reduced to its absolute value. As the patch fixes the failures that previously obstructed moving tests of cexp to use ALL_RM_TEST, those tests are moved to ALL_RM_TEST by the patch (two functions remain yet to be converted). Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18594] * math/s_ccosh.c (__ccosh): Compare with least normal value instead of comparing class with FP_SUBNORMAL. * math/s_ccoshf.c (__ccoshf): Likewise. * math/s_ccoshl.c (__ccoshl): Likewise. * math/s_cexp.c (__cexp): Likewise. * math/s_cexpf.c (__cexpf): Likewise. * math/s_cexpl.c (__cexpl): Likewise. * math/s_csin.c (__csin): Likewise. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/s_ctan.c (__ctan): Likewise. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Add more tests of ccos, ccosh, cexp, csin, csinh, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cexp_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-24 21:04:51 +00:00
csinh 1 min
csinh 1 -min
csinh 80 min_subnorm
csinh 80 -min_subnorm
Move more csin, csinh tests to auto-libm-test-in. When I moved tests of csin and csinh to auto-libm-test-in, I didn't move a few tests for which gen-auto-libm-tests was very slow because of MPC slowness on certain inputs. Now that auto-libm-test-out has been split up, such slowness only affects regenerating the test expectations for the individual functions in question, rather than any addition of tests for any function to auto-libm-test-in. Thus, I no longer consider it a problem to have these inputs in auto-libm-test-in, and this patch moves them there. This results in test generation for csin and csinh taking 5m43s (for each of csin and csinh) on my system, while other functions are unaffected. I expect the test generation to be much faster in MPC 1.1 (the relevant performance improvements went in MPC mainline in Dec 2013, but there hasn't been a release from mainline since then). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of csin and csinh. * math/auto-libm-test-out-csin: Regenerated. * math/auto-libm-test-out-csinh: Likewise. * math/libm-test-csin.inc (csin_test_data): Remove tests moved to auto-libm-test-in. * math/libm-test-csinh.inc (csinh_test_data): Likewise.
2017-02-09 18:30:41 +00:00
csinh min_subnorm min
csinh min min_subnorm
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
csqrt 0 0
csqrt 0 -0
csqrt -0 0
csqrt -0 -0
csqrt 16.0 -30.0
csqrt -1 0
csqrt 0 2
csqrt 119 120
csqrt 0.75 1.25
csqrt -2 -3
csqrt -2 3
# Principal square root should be returned (i.e., non-negative real part).
csqrt 0 -1
Add more tests of csqrt. This patch adds some randomly-generated tests of csqrt that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of csqrt. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-30 22:51:29 +00:00
csqrt -0xe.6432ap-4 0xe.8175p-4
csqrt -0x4.d01448p-4 -0x7.c1915p+0
csqrt -0xd.e1d5fp-4 -0x1.054226p+4
csqrt 0x5.39e238p+0 -0x4.576278p-4
csqrt -0xe.735dbp+0 -0x5.26cb98p+40
csqrt -0x7.915fafbe9f588p-4 -0x2.5e01bcp+0
csqrt 0xe.229827fe17d08p-4 0xd.849ecp-4
csqrt -0x4.d0144005d7af4p-4 -0x7.c19148p+0
csqrt 0x8p-152 0x7.8p-148
csqrt -0x4.82773b736291p-4 -0x1.bcb7cep+0
csqrt 0xf.fffffp+124 0xe.7e0c2p+116
csqrt -0x4.15ca1p+0 -0x8p-152
csqrt 0xf.a24adp+28 0x8.0f148p+36
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
csqrt 0x1.f9610ap+4 0x9.87716p+4
csqrt 0x5.9cc21p-4 -0x1.fb1ec91b40dcdp+0
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
csqrt -0x7.31291c9fdae04p-160 -0x8p-152
csqrt 0x1.d60caep+0 0x7.a7d468p+0
csqrt -0xb.e2bc1cd6eaa7p-180 0x8p-152
Add more tests of acosh, atanh, cos, csqrt, erfc, sin, sincos. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, cos, csqrt, erfc, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-06 17:30:18 +00:00
csqrt 0xd.25d559ac5baap-168 0x8p-152
Add more tests of csqrt, lgamma, log10, sinh. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of csqrt, lgamma, log10 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-08 17:55:11 +00:00
csqrt -0x9.0a61a7b482d28p-168 -0x8p-152
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
csqrt 0x3.f768f58949e3fe6cp-4 0x2.0c2e89a5cff98p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
csqrt 0x6.b1a2e79e9c9acp-164 0x8p-152
csqrt -0x8.ec8932bf5603p-172 0x8p-152
Add more randomly-generated libm tests. This patch adds more libm test inputs found through random test generation to increase observed ulps on x86_64. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, atanh, cbrt, cosh, csqrt, erfc, expm1 and lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-11 15:03:10 +00:00
csqrt -0x1.9edb24c83e22cp-172 -0x8p-152
csqrt -0x1.65c7ac7c97a25p-176 -0x8p-152
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
csqrt 0x1.0221e29d5a3cdp-188 -0x8p-152
csqrt -0x2.927275f6febb6p-184 0x8p-152
csqrt -0x8p-152 -0x2.c832ff5b163af0dcp-2444
csqrt -0x8p-152 -0x1.818bccp+0
Add more tests of csqrt. This patch adds some randomly-generated tests of csqrt that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of csqrt. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-30 22:51:29 +00:00
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
csqrt 0x1.fffffep+127 0x1.fffffep+127
csqrt 0x1.fffffep+127 1.0
csqrt 0x1p-149 0x1p-149
csqrt 0x1p-147 0x1p-147
csqrt 0 0x1p-149
csqrt 0x1p-50 0x1p-149
csqrt 0x1p+127 0x1p-149
csqrt 0x1p-149 0x1p+127
csqrt 0x1.000002p-126 0x1.000002p-126
csqrt -0x1.000002p-126 -0x1.000002p-126
csqrt 0x1.fffffffffffffp+1023 0x1.fffffffffffffp+1023
csqrt 0x1.fffffffffffffp+1023 0x1p+1023
csqrt 0x1p-1074 0x1p-1074
csqrt 0x1p-1073 0x1p-1073
csqrt 0 0x1p-1074
csqrt 0x1p-500 0x1p-1074
csqrt 0x1p+1023 0x1p-1074
csqrt 0x1p-1074 0x1p+1023
csqrt 0x1.0000000000001p-1022 0x1.0000000000001p-1022
csqrt -0x1.0000000000001p-1022 -0x1.0000000000001p-1022
csqrt 0x1.fp+16383 0x1.fp+16383
csqrt 0x1.fp+16383 0x1p+16383
csqrt 0x1p-16440 0x1p-16441
csqrt 0 0x1p-16445
csqrt 0x1p-5000 0x1p-16445
csqrt 0x1p+16383 0x1p-16445
csqrt 0x1p-16445 0x1p+16383
csqrt 0x1.0000000000000002p-16382 0x1.0000000000000002p-16382
csqrt -0x1.0000000000000002p-16382 -0x1.0000000000000002p-16382
csqrt 0 0x1p-16494
csqrt 0x1p-5000 0x1p-16494
csqrt 0x1p+16383 0x1p-16494
csqrt 0x1p-16494 0x1p+16383
csqrt 0x1.0000000000000000000000000001p-16382 0x1.0000000000000000000000000001p-16382
csqrt -0x1.0000000000000000000000000001p-16382 -0x1.0000000000000000000000000001p-16382
Fix csqrt spurious underflows (bug 18371). The csqrt implementations in glibc can cause spurious underflows in some cases as a side-effect of the scaling for large arguments (when underflow is correct for the square root of the argument that was scaled down to avoid overflow, but not for the original argument). This patch arranges to avoid the underflowing intermediate computation (eliminating a multiplication in 0.5 in the problem cases where a subsequent scaling by 2 would follow). Tested for x86_64 and x86 and ulps updated accordingly (only needed for x86). [BZ #18371] * math/s_csqrt.c (__csqrt): Avoid multiplication by 0.5 where intermediate but not final result might underflow. * math/s_csqrtf.c (__csqrtf): Likewise. * math/s_csqrtl.c (__csqrtl): Likewise. * math/auto-libm-test-in: Add more tests of csqrt. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-06-23 16:01:54 +00:00
csqrt 0x0.ffp128 0x1.1p-61
csqrt -0x0.ffp128 0x1.1p-61
csqrt 0x0.ffp1024 0x1.1p-509
csqrt -0x0.ffp1024 0x1.1p-509
csqrt 0x0.ffp16384 0x1.1p-8189
csqrt -0x0.ffp16384 0x1.1p-8189
Fix csqrt spurious underflows (bug 18823). The csqrt functions scale up small arguments to avoid underflows when calling hypot functions. However, even when hypot does not underflow, a subsequent calculation of 0.5 * hypot can underflow. This patch duly increases the threshold and scale factor to avoid such underflows as well. Tested for x86_64, x86 and mips64. [BZ #18823] * math/s_csqrt.c (__csqrt): Increase threshold and scale factor for scaling up small arguments. * math/s_csqrtf.c (__csqrtf): Likewise. * math/s_csqrtl.c (__csqrtl): Likewise. * math/auto-libm-test-in: Add more tests of csqrt. * math/auto-libm-test-out: Regenerated.
2015-08-17 23:02:54 +00:00
csqrt 0x1p-149 0x1.000002p-126
csqrt 0x1p-149 0x1.000004p-126
csqrt 0x1p-1074 0x1.0000000000001p-1022
csqrt 0x1p-1074 0x1.0000000000002p-1022
csqrt 0x1p-16445 0x1.0000000000000002p-16382
csqrt 0x1p-16445 0x1.0000000000000004p-16382
csqrt 0x1p-16494 0x1.0000000000000000000000000001p-16382
csqrt 0x1p-16494 0x1.0000000000000000000000000002p-16382
Fix csqrt missing underflows (bug 18370). The csqrt implementations in glibc can miss underflow exceptions when the real or imaginary part of the result becomes tiny in the course of scaling down (in particular, multiplication by 0.5) and that scaling is exact although the relevant part of the mathematical result isn't. This patch forces the exception in a similar way to previous fixes. Tested for x86_64 and x86. [BZ #18370] * math/s_csqrt.c (__csqrt): Force underflow exception for results whose real or imaginary part has small absolute value. * math/s_csqrtf.c (__csqrtf): Likewise. * math/s_csqrtl.c (__csqrtl): Likewise. * math/auto-libm-test-in: Add more tests of csqrt. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-08-19 22:42:01 +00:00
csqrt 1 min
csqrt 1 -min
csqrt -1 min
csqrt -1 -min
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ctan 0 0
ctan 0 -0
ctan -0 0
ctan -0 -0
ctan 0.75 1.25
ctan -2 -3
ctan 1 45
ctan 1 47
ctan 1 355
ctan 1 365
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
ctan 1 5680
ctan 1 5690
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ctan 0x3.243f6cp-1 0
ctan 0x1p127 1
ctan 0x1p1023 1
ctan 0x1p16383 1
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
ctan 50000 50000
ctan 50000 -50000
ctan -50000 50000
ctan -50000 -50000
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ctan 0x1.921fb6p+0 0x1p-149
ctan 0x1.921fb54442d18p+0 0x1p-1074
ctan 0x1.921fb54442d1846ap+0 0x1p-16445
Fix ctan, ctanh missing underflows (bug 18595). Similar to various other bugs in this area, ctan and ctanh can fail to raise the underflow exception for some cases of results that are tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64 and x86. [BZ #18595] * math/s_ctan.c (__ctan): Force underflow exception for results whose real or imaginary part has small absolute value. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Do not allow missing underflow for ctan and ctanh. Add more tests of ctan and ctanh.
2015-09-15 17:46:08 +00:00
ctan min 0
ctan -min 0
ctan min_subnorm 0
ctan -min_subnorm 0
ctan 0 min
ctan 0 -min
ctan 0 min_subnorm
ctan 0 -min_subnorm
Fix cexp, ccos, ccosh, csin, csinh spurious underflows (bug 18594). cexp, ccos, ccosh, csin and csinh have spurious underflows in cases where they compute sin of the smallest normal, that produces an underflow exception (depending on which sin implementation is in use) but the final result does not underflow. ctan and ctanh may also have such underflows, or they may be latent (the issue there is that e.g. ctan (DBL_MIN) should, rounded upwards, be the next double value above DBL_MIN, which under glibc's accuracy goals may not have an underflow exception, but the intermediate computation of sin (DBL_MIN) would legitimately underflow on before-rounding architectures). This patch fixes all those functions so they use plain comparisons (> DBL_MIN etc.) instead of comparing the result of fpclassify with FP_SUBNORMAL (in all these cases, we already know the number being compared is finite). Note that in the case of csin / csinf / csinl, there is no need for fabs calls in the comparison because the real part has already been reduced to its absolute value. As the patch fixes the failures that previously obstructed moving tests of cexp to use ALL_RM_TEST, those tests are moved to ALL_RM_TEST by the patch (two functions remain yet to be converted). Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18594] * math/s_ccosh.c (__ccosh): Compare with least normal value instead of comparing class with FP_SUBNORMAL. * math/s_ccoshf.c (__ccoshf): Likewise. * math/s_ccoshl.c (__ccoshl): Likewise. * math/s_cexp.c (__cexp): Likewise. * math/s_cexpf.c (__cexpf): Likewise. * math/s_cexpl.c (__cexpl): Likewise. * math/s_csin.c (__csin): Likewise. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/s_ctan.c (__ctan): Likewise. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Add more tests of ccos, ccosh, cexp, csin, csinh, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cexp_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-24 21:04:51 +00:00
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ctanh 0 0
ctanh 0 -0
ctanh -0 0
ctanh -0 -0
ctanh 0 pi/4
ctanh 0.75 1.25
ctanh -2 -3
ctanh 45 1
ctanh 47 1
ctanh 355 1
ctanh 365 1
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
ctanh 5680 1
ctanh 5690 1
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ctanh 0 0x3.243f6cp-1
ctanh 1 0x1p127
ctanh 1 0x1p1023
ctanh 1 0x1p16383
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
ctanh 50000 50000
ctanh 50000 -50000
ctanh -50000 50000
ctanh -50000 -50000
Move various TEST_c_c tests from libm-test.inc to auto-libm-test-inc. This patch moves tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Other TEST_c_c functions aren't moved for now (although the relevant table entries are put in gen-auto-libm-tests for it to know how to handle them): clog10 because of a known MPC bug causing it to hang for at least some pure imaginary inputs (fixed in SVN, but I'd rather not rely on unreleased versions of MPFR or MPC even if relying on very recent releases); the inverse trig and hyperbolic functions because of known slowness in special cases; and csin / csinh because of observed slowness that I need to investigate and report to the MPC maintainers. Slowness can be bypassed by moving to incremental generation (only for new / changed tests) rather than regenerating the whole of auto-libm-test-out every time, but that needs implementing. (This patch takes the time for running gen-auto-libm-tests from about one second to seven, on my system, which I think is reasonable. The slow functions would make it take several minutes at least, which seems unreasonable.) Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of ccos, ccosh, cexp, clog, csqrt, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): New macro. (TEST_COND_x86): Likewise. (ccos_test_data): Use AUTO_TESTS_c_c. (ccosh_test_data): Likewise. (cexp_test_data): Likewise. (clog_test_data): Likewise. (csqrt_test_data): Likewise. (ctan_test_data): Likewise. (ctan_tonearest_test_data): Likewise. (ctan_towardzero_test_data): Likewise. (ctan_downward_test_data): Likewise. (ctan_upward_test_data): Likewise. (ctanh_test_data): Likewise. (ctanh_tonearest_test_data): Likewise. (ctanh_towardzero_test_data): Likewise. (ctanh_downward_test_data): Likewise. (ctanh_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpc_c_c. (func_calc_desc): Add mpc_c_c union field. (FUNC_mpc_c_c): New macro. (test_functions): Add cacos, cacosh, casin, casinh, catan, catanh, ccos, ccosh, cexp, clog, clog10, csin, csinh, csqrt, ctan and ctanh. (special_fill_min_subnorm_p120): New function. (special_real_inputs): Add min_subnorm_p120. (calc_generic_results): Handle mpc_c_c. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-20 12:32:44 +00:00
ctanh 0x1p-149 0x1.921fb6p+0
ctanh 0x1p-1074 0x1.921fb54442d18p+0
ctanh 0x1p-16445 0x1.921fb54442d1846ap+0
Fix ctan, ctanh missing underflows (bug 18595). Similar to various other bugs in this area, ctan and ctanh can fail to raise the underflow exception for some cases of results that are tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64 and x86. [BZ #18595] * math/s_ctan.c (__ctan): Force underflow exception for results whose real or imaginary part has small absolute value. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Do not allow missing underflow for ctan and ctanh. Add more tests of ctan and ctanh.
2015-09-15 17:46:08 +00:00
ctanh 0 min
ctanh 0 -min
ctanh 0 min_subnorm
ctanh 0 -min_subnorm
ctanh min 0
ctanh -min 0
ctanh min_subnorm 0
ctanh -min_subnorm 0
Fix cexp, ccos, ccosh, csin, csinh spurious underflows (bug 18594). cexp, ccos, ccosh, csin and csinh have spurious underflows in cases where they compute sin of the smallest normal, that produces an underflow exception (depending on which sin implementation is in use) but the final result does not underflow. ctan and ctanh may also have such underflows, or they may be latent (the issue there is that e.g. ctan (DBL_MIN) should, rounded upwards, be the next double value above DBL_MIN, which under glibc's accuracy goals may not have an underflow exception, but the intermediate computation of sin (DBL_MIN) would legitimately underflow on before-rounding architectures). This patch fixes all those functions so they use plain comparisons (> DBL_MIN etc.) instead of comparing the result of fpclassify with FP_SUBNORMAL (in all these cases, we already know the number being compared is finite). Note that in the case of csin / csinf / csinl, there is no need for fabs calls in the comparison because the real part has already been reduced to its absolute value. As the patch fixes the failures that previously obstructed moving tests of cexp to use ALL_RM_TEST, those tests are moved to ALL_RM_TEST by the patch (two functions remain yet to be converted). Tested for x86_64 and x86 and ulps updated accordingly. [BZ #18594] * math/s_ccosh.c (__ccosh): Compare with least normal value instead of comparing class with FP_SUBNORMAL. * math/s_ccoshf.c (__ccoshf): Likewise. * math/s_ccoshl.c (__ccoshl): Likewise. * math/s_cexp.c (__cexp): Likewise. * math/s_cexpf.c (__cexpf): Likewise. * math/s_cexpl.c (__cexpl): Likewise. * math/s_csin.c (__csin): Likewise. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/s_ctan.c (__ctan): Likewise. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * math/auto-libm-test-in: Add more tests of ccos, ccosh, cexp, csin, csinh, ctan and ctanh. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (cexp_test): Use ALL_RM_TEST. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-24 21:04:51 +00:00
Add narrowing divide functions. This patch adds the narrowing divide functions from TS 18661-1 to glibc's libm: fdiv, fdivl, ddivl, f32divf64, f32divf32x, f32xdivf64 for all configurations; f32divf64x, f32divf128, f64divf64x, f64divf128, f32xdivf64x, f32xdivf128, f64xdivf128 for configurations with _Float64x and _Float128; __nldbl_ddivl for ldbl-opt. The changes are mostly essentially the same as for the other narrowing functions, so the description of those generally applies to this patch as well. Tested for x86_64, x86, mips64 (all three ABIs, both hard and soft float) and powerpc, and with build-many-glibcs.py. * math/Makefile (libm-narrow-fns): Add div. (libm-test-funcs-narrow): Likewise. * math/Versions (GLIBC_2.28): Add narrowing divide functions. * math/bits/mathcalls-narrow.h (div): Use __MATHCALL_NARROW. * math/gen-auto-libm-tests.c (test_functions): Add div. * math/math-narrow.h (CHECK_NARROW_DIV): New macro. (NARROW_DIV_ROUND_TO_ODD): Likewise. (NARROW_DIV_TRIVIAL): Likewise. * sysdeps/ieee754/float128/float128_private.h (__fdivl): New macro. (__ddivl): Likewise. * sysdeps/ieee754/ldbl-opt/Makefile (libnldbl-calls): Add fdiv and ddiv. (CFLAGS-nldbl-ddiv.c): New variable. (CFLAGS-nldbl-fdiv.c): Likewise. * sysdeps/ieee754/ldbl-opt/Versions (GLIBC_2.28): Add __nldbl_ddivl. * sysdeps/ieee754/ldbl-opt/nldbl-compat.h (__nldbl_ddivl): New prototype. * manual/arith.texi (Misc FP Arithmetic): Document fdiv, fdivl, ddivl, fMdivfN, fMdivfNx, fMxdivfN and fMxdivfNx. * math/auto-libm-test-in: Add tests of div. * math/auto-libm-test-out-narrow-div: New generated file. * math/libm-test-narrow-div.inc: New file. * sysdeps/i386/fpu/s_f32xdivf64.c: Likewise. * sysdeps/ieee754/dbl-64/s_f32xdivf64.c: Likewise. * sysdeps/ieee754/dbl-64/s_fdiv.c: Likewise. * sysdeps/ieee754/float128/s_f32divf128.c: Likewise. * sysdeps/ieee754/float128/s_f64divf128.c: Likewise. * sysdeps/ieee754/float128/s_f64xdivf128.c: Likewise. * sysdeps/ieee754/ldbl-128/s_ddivl.c: Likewise. * sysdeps/ieee754/ldbl-128/s_f64xdivf128.c: Likewise. * sysdeps/ieee754/ldbl-128/s_fdivl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_ddivl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_fdivl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_ddivl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_fdivl.c: Likewise. * sysdeps/ieee754/ldbl-opt/nldbl-ddiv.c: Likewise. * sysdeps/ieee754/ldbl-opt/nldbl-fdiv.c: Likewise. * sysdeps/ieee754/soft-fp/s_ddivl.c: Likewise. * sysdeps/ieee754/soft-fp/s_fdiv.c: Likewise. * sysdeps/ieee754/soft-fp/s_fdivl.c: Likewise. * sysdeps/powerpc/fpu/libm-test-ulps: Update. * sysdeps/mach/hurd/i386/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/aarch64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/alpha/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/arm/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/hppa/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/i386/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/m68k/coldfire/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/m68k/m680x0/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/microblaze/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/mips/mips32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/mips/mips64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/nios2/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/fpu/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/nofpu/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/libm-le.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/riscv/rv64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/s390/s390-32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/s390/s390-64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sh/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/x86_64/64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/x86_64/x32/libm.abilist: Likewise.
2018-05-17 00:40:52 +00:00
div 0 min
div 0 -min
div 0 min_subnorm
div 0 -min_subnorm
div 0 max
div 0 -max
div -0 min
div -0 -min
div -0 min_subnorm
div -0 -min_subnorm
div -0 max
div -0 -max
div max max xfail:ibm128-libgcc
div max -max xfail:ibm128-libgcc
div -max max xfail:ibm128-libgcc
div -max -max xfail:ibm128-libgcc
div min min
div min -min
div -min min
div -min -min
div min_subnorm min_subnorm
div min_subnorm -min_subnorm
div -min_subnorm min_subnorm
div -min_subnorm -min_subnorm
div max min xfail-rounding:ibm128-libgcc
div max -min xfail-rounding:ibm128-libgcc
div -max min xfail-rounding:ibm128-libgcc
div -max -min xfail-rounding:ibm128-libgcc
div max min_subnorm xfail-rounding:ibm128-libgcc
div max -min_subnorm xfail-rounding:ibm128-libgcc
div -max min_subnorm xfail-rounding:ibm128-libgcc
div -max -min_subnorm xfail-rounding:ibm128-libgcc
div min max xfail-rounding:ibm128-libgcc
div min -max xfail-rounding:ibm128-libgcc
div -min max xfail-rounding:ibm128-libgcc
div -min -max xfail-rounding:ibm128-libgcc
div min_subnorm max xfail-rounding:ibm128-libgcc
div min_subnorm -max xfail-rounding:ibm128-libgcc
div -min_subnorm max xfail-rounding:ibm128-libgcc
div -min_subnorm -max xfail-rounding:ibm128-libgcc
div 1 2
div 1 -2
div -1 2
div -1 -2
div 100.5 0.75
div 100.5 -0.75
div -100.5 0.75
div -100.5 -0.75
# Cases where the ratio of two values in a wider format is very close
# to half way between two representable values in a narrower format,
# so that double rounding would sometimes yield the wrong result. For
# a narrower format of width a and a wider format of width w, take
# b = w - a; these examples are of the form
# (1 + 2^-a + 2^-b + 2^(1-w)) / (1 + 2^-b). These examples have
# ratios of the form (1 + 2^-a) to the width of the wider format, but
# the mathematical value is slightly greater, so that rounding once
# should round the result up but double rounding would wrongly round
# the result down.
# a = 24, w = 53.
div 0x1.0000010800001p0 0x1.00000008p0
# a = 24, w = 64.
div 0x1.0000010001000002p0 0x1.0000000001p0
# a = 24, w = 113.
div 0x1.0000010000000000000000800001p0 0x1.00000000000000000000008p0
# a = 53, w = 64.
div 0x1.0020000000000802p0 0x1.002p0
# a = 53, w = 113.
div 0x1.0000000000000810000000000001p0 0x1.000000000000001p0
# a = 64, w = 113.
div 0x1.0000000000008001000000000001p0 0x1.0000000000008p0
# Similar, for double rounding to 64-bit of a division of 53-bit values.
div 0x1ffe1p0 0xfffp0
Fix f64xdivf128, f64xmulf128 spurious underflows (bug 28358) As described in bug 28358, the round-to-odd computations used in the libm functions that round their results to a narrower format can yield spurious underflow exceptions in the following circumstances: the narrowing only narrows the precision of the type and not the exponent range (i.e., it's narrowing _Float128 to _Float64x on x86_64, x86 or ia64), the architecture does after-rounding tininess detection (which applies to all those architectures), the result is inexact, tiny before rounding but not tiny after rounding (with the chosen rounding mode) for _Float64x (which is possible for narrowing mul, div and fma, not for narrowing add, sub or sqrt), so the underflow exception resulting from the toward-zero computation in _Float128 is spurious for _Float64x. Fixed by making ROUND_TO_ODD call feclearexcept (FE_UNDERFLOW) in the problem cases (as indicated by an extra argument to the macro); there is never any need to preserve underflow exceptions from this part of the computation, because the conversion of the round-to-odd value to the narrower type will underflow in exactly the cases in which the function should raise that exception, but it may be more efficient to avoid the extra manipulation of the floating-point environment when not needed. Tested for x86_64 and x86, and with build-many-glibcs.py.
2021-09-21 21:54:37 +00:00
# Cases where there is underflow before rounding (for some format) but
# might not be after rounding, depending on the rounding mode.
div 0x1p-126 0x1.0000001p0
div 0x1p-126 -0x1.0000001p0
div -0x1p-126 0x1.0000001p0
div -0x1p-126 -0x1.0000001p0
Adjust new narrowing div/mul tests for IBM long double, update powerpc ULPs Testing for powerpc shows some of the new narrowing div/mul tests need XFAILing for IBM long double and some ULPs updates are needed for those tests.
2021-09-22 12:35:44 +00:00
div 0x1p-1022 0x1.00000000000001p0 missing-underflow:arg-ibm128
div 0x1p-1022 -0x1.00000000000001p0 missing-underflow:arg-ibm128
div -0x1p-1022 0x1.00000000000001p0 missing-underflow:arg-ibm128
div -0x1p-1022 -0x1.00000000000001p0 missing-underflow:arg-ibm128
div 0x1p-16382 0x1.00000000000000001p0 missing-underflow:arg-ibm128 xfail-rounding:ibm128-libgcc
div 0x1p-16382 -0x1.00000000000000001p0 missing-underflow:arg-ibm128 xfail-rounding:ibm128-libgcc
div -0x1p-16382 0x1.00000000000000001p0 missing-underflow:arg-ibm128 xfail-rounding:ibm128-libgcc
div -0x1p-16382 -0x1.00000000000000001p0 missing-underflow:arg-ibm128 xfail-rounding:ibm128-libgcc
Fix f64xdivf128, f64xmulf128 spurious underflows (bug 28358) As described in bug 28358, the round-to-odd computations used in the libm functions that round their results to a narrower format can yield spurious underflow exceptions in the following circumstances: the narrowing only narrows the precision of the type and not the exponent range (i.e., it's narrowing _Float128 to _Float64x on x86_64, x86 or ia64), the architecture does after-rounding tininess detection (which applies to all those architectures), the result is inexact, tiny before rounding but not tiny after rounding (with the chosen rounding mode) for _Float64x (which is possible for narrowing mul, div and fma, not for narrowing add, sub or sqrt), so the underflow exception resulting from the toward-zero computation in _Float128 is spurious for _Float64x. Fixed by making ROUND_TO_ODD call feclearexcept (FE_UNDERFLOW) in the problem cases (as indicated by an extra argument to the macro); there is never any need to preserve underflow exceptions from this part of the computation, because the conversion of the round-to-odd value to the narrower type will underflow in exactly the cases in which the function should raise that exception, but it may be more efficient to avoid the extra manipulation of the floating-point environment when not needed. Tested for x86_64 and x86, and with build-many-glibcs.py.
2021-09-21 21:54:37 +00:00
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erf 0
erf -0
erf 0.125
erf 0.75
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erf 1
erf -1
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erf 1.25
erf 2.0
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erf -2
erf 3
erf -3
erf 4
erf -4
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erf 4.125
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erf 5
erf 6
erf 7
erf 8
erf 9
erf 10
erf -10
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erf 27.0
erf -27.0
erf -0x1.fffffffffffff8p-2
Fix erf underflow handling near 0 (bug 16516). Bug 16516 reports spurious underflows from erf (for all floating-point types), when the result is close to underflowing but does not actually underflow. erf (x) is about (2/sqrt(pi))*x for x close to 0, so there are subnormal arguments for which it does not underflow. The various implementations do (x + efx*x) (for efx = 2/sqrt(pi) - 1), for greater accuracy than if just using a single multiplication by an approximation to 2/sqrt(pi) (effectively, this way there are a few more bits in the approximation to 2/sqrt(pi)). This can introduce underflows when efx*x underflows even though the final result does not, so a scaled calculation with 8*efx is done in these cases - but 8 is not a big enough scale factor to avoid all such underflows. 16 is (any underflows with a scale factor of 16 would only occur when the final result underflows), so this patch changes the code to use that factor. Rather than recomputing all the values of the efx8 variable, it is removed, leaving it to the compiler's constant folding to compute 16*efx. As such scaling can also lose underflows when the final scaling down happens to be exact, appropriate checks are added to ensure underflow exceptions occur when required in such cases. Tested x86_64 and x86; no ulps updates needed. Also spot-checked for powerpc32 and mips64 to verify the changes to the ldbl-128ibm and ldbl-128 implementations. [BZ #16516] * sysdeps/ieee754/dbl-64/s_erf.c (efx8): Remove variable. (__erf): Scale by 16 instead of 8 in potentially underflowing case. Ensure exception if result actually underflows. * sysdeps/ieee754/flt-32/s_erff.c (efx8): Remove variable. (__erff): Scale by 16 instead of 8 in potentially underflowing case. Ensure exception if result actually underflows. * sysdeps/ieee754/ldbl-128/s_erfl.c: Include <float.h>. (efx8): Remove variable. (__erfl): Scale by 16 instead of 8 in potentially underflowing case. Ensure exception if result actually underflows. * sysdeps/ieee754/ldbl-128ibm/s_erfl.c: Include <float.h>. (efx8): Remove variable. (__erfl): Scale by 16 instead of 8 in potentially underflowing case. Ensure exception if result actually underflows. * sysdeps/ieee754/ldbl-96/s_erfl.c: Include <float.h>. (efx8): Remove variable. (__erfl): Scale by 16 instead of 8 in potentially underflowing case. Ensure exception if result actually underflows. * math/auto-libm-test-in: Add more tests of erf. * math/auto-libm-test-out: Regenerated.
2014-05-14 12:34:03 +00:00
erf 0x1.c5bf94p-127
erf 0x3.8b7fa8p-128
erf -0x3.8b7f12369ded8p-1024
erf 0x3.8b7f12369ded5518p-16384
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erf 0x3.8b7ee8p-128
erf 0x3.8b7f3cp-128
erf 0x3.8b7f12369decp-1024
erf 0x3.8b7f12369deeap-1024
erf 0x7.16fe246d3bdaa9e70ec1483562p-972
erf 0x7.16fe246d3bdaa9e70ec148358cp-972
erf 0x3.8b7f12369ded54c8p-16384
erf 0x3.8b7f12369ded551cp-16384
erf 0x1.c5bf891b4ef6aa64p-16384
erf 0x1.c5bf891b4ef6aa8ep-16384
erf 0x3.8b7f12369ded54f38760a41abb5cp-16384
erf 0x3.8b7f12369ded54f38760a41abb86p-16384
Fix ldbl-128 erfl spurious underflows (bug 16287). This patch fixes bug 16287, spurious underflows from ldbl-128 erfl arising from it calling erfcl for arguments with absolute value at least 1.0, although for large positive arguments erfcl correctly underflows but erfl shouldn't. The fix is simply to avoid calling erfcl, and just return 1, for arguments above a cut-off large enough that erfl correctly rounds to-nearest as 1 but not so large that erfcl underflows. Tested mips64. Also tested x86_64 and x86 to confirm the new tests (taken from the tests of erfc) don't cause any problems there; no ulps updates needed. [BZ #16287] * sysdeps/ieee754/ldbl-128/s_erfl.c (__erfl): Return 1 without calling __erfcl for arguments at least 16. * math/auto-libm-test-in: Add more tests of erf. * math/auto-libm-test-out: Regenerated.
2014-06-24 20:56:56 +00:00
erf 26.0
erf 28.0
erf 100
erf 106
erf 106.5
erf 106.625
erf 107
erf 108
erf 1000
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erf 0x1p-5
erf -0x1p-5
erf 0x1p-10
erf 0x1p-15
erf 0x1p-20
erf 0x1p-25
erf 0x1p-30
erf 0x1p-35
erf 0x1p-40
erf 0x1p-45
erf 0x1p-50
erf 0x1p-55
erf 0x1p-60
erf 0x1p-100
erf 0x1p-600
erf 0x1p-10000
erf min
erf -min
erf min_subnorm
erf -min_subnorm
Fix ldbl-128 erfl spurious underflows (bug 16287). This patch fixes bug 16287, spurious underflows from ldbl-128 erfl arising from it calling erfcl for arguments with absolute value at least 1.0, although for large positive arguments erfcl correctly underflows but erfl shouldn't. The fix is simply to avoid calling erfcl, and just return 1, for arguments above a cut-off large enough that erfl correctly rounds to-nearest as 1 but not so large that erfcl underflows. Tested mips64. Also tested x86_64 and x86 to confirm the new tests (taken from the tests of erfc) don't cause any problems there; no ulps updates needed. [BZ #16287] * sysdeps/ieee754/ldbl-128/s_erfl.c (__erfl): Return 1 without calling __erfcl for arguments at least 16. * math/auto-libm-test-in: Add more tests of erf. * math/auto-libm-test-out: Regenerated.
2014-06-24 20:56:56 +00:00
erf max
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erf -max
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
Add more tests of erf, erfc. This patch adds some randomly-generated tests of erf and erfc that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of erf and erfc. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 17:49:44 +00:00
erf -0x1.ddaea4p+0
erf -0x1.2b1f68p+0
erf 0x1.44e722p+0
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
erf -0x1.3a0d48p+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
erf -0x1.c975cap+0
erf -0x1.e6a006p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
erf -0x1.4d32f4p-12
Add more tests of erf, erfc. This patch adds some randomly-generated tests of erf and erfc that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of erf and erfc. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 17:49:44 +00:00
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erfc 0.0
erfc -0
Fix erfcf spurious underflows (bug 18217). The flt-32 implementation of erfcf produces spurious underflow exceptions for some arguments close to 0, because of calculations squaring the argument and then multiplying by small constants. This patch fixes this by adjusting the threshold for arguments for which the result is so close to 1 that 1 - x will give the right result from 2**-56 to 2**-26. (If 1 - x * 2/sqrt(pi) were used, the errors would be on the order of x^3 and a much larger threshold could be used.) Tested for x86_64 and x86. [BZ #18217] * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Use 2**-26 not 2**-56 as threshold for returning 1 - x. * math/auto-libm-test-in: Add more tests of erfc. * math/auto-libm-test-out: Regenerated.
2015-05-15 00:16:10 +00:00
erfc 0x1p-55
erfc -0x1p-55
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erfc 0.125
erfc 0.75
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erfc 1
erfc -1
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erfc 1.25
erfc 2.0
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erfc -2
erfc 3
erfc -3
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erfc 0x1.f7303cp+1
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erfc 4
erfc -4
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erfc 4.125
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erfc 5
erfc -5
erfc 6
erfc -6
erfc 7
erfc -7
erfc 8
erfc -8
erfc 9
erfc -9
erfc 10
erfc -10
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erfc 0x1.ffa002p+2
erfc 0x1.ffffc8p+2
erfc -0x1.fffffffffffff8p-2
Fix erfc errno setting on underflow (bug 6786).
2013-12-03 16:25:18 +00:00
erfc 26.0
erfc 27.0
erfc 28.0
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
erfc 0x1.ffff56789abcdef0123456789a8p+2
Fix erfc errno setting on underflow (bug 6786).
2013-12-03 16:25:18 +00:00
erfc 100
erfc 106
erfc 106.5
erfc 106.625
erfc 107
erfc 108
erfc 1000
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erfc 0x1p-5
erfc -0x1p-5
erfc 0x1p-10
erfc 0x1p-15
erfc 0x1p-20
erfc 0x1p-25
erfc 0x1p-30
erfc 0x1p-35
erfc 0x1p-40
erfc 0x1p-45
erfc 0x1p-50
erfc 0x1p-60
erfc 0x1p-100
erfc 0x1p-600
erfc 0x1p-10000
erfc 0x9.31cdfp+0
erfc 0x9.31cep+0
erfc 0x1.a8b12fc6e4891p+4
erfc 0x1.a8b12fc6e4892p+4
erfc 0x1.9d7adac608e8586300e6c8b99ep+4
erfc 0x1.9d7adac608e8586300e6c8b99e8p+4
erfc 0x6.a893032db905274p+4
erfc 0x6.a893032db9052748p+4
erfc 0x6.a8a0561d8bbe942p+4
erfc 0x6.a8a0561d8bbe9428p+4
erfc 0x6.a893032db905274042fb05c665dcp+4
erfc 0x6.a893032db905274042fb05c665ep+4
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
erfc 0xd.03d06p-4
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (ldbl-96)
erfc 0x1.5cf2167efe9207d2p+0
# the next value generates larger error bounds on x86_64 (binary128)
erfc 0x1.5166e0efc44a9dfc79b8c8873a99p+0
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erfc min
erfc -min
erfc min_subnorm
erfc -min_subnorm
Fix erfc errno setting on underflow (bug 6786).
2013-12-03 16:25:18 +00:00
erfc max
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
erfc -max
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
Add more tests of erf, erfc. This patch adds some randomly-generated tests of erf and erfc that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of erf and erfc. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 17:49:44 +00:00
erfc 0x1.8a0c64p+0
erfc 0x1.8a0c62p+0
erfc 0x1.64dafap+0
erfc 0x6.88fb08p+0
erfc 0xd.361d9p-4
erfc 0x8.c66b44ca40038p+0
erfc 0x2.586f1cp+0
erfc 0xb.acb72p+0
erfc 0xb.227499103357d84p+0
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
erfc 0xd.28abfp-4
erfc 0x1.5289fep+0
erfc 0x4.b48498p+0
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
erfc 0x2.f8646cp+0
Add more tests of acosh, atanh, cos, csqrt, erfc, sin, sincos. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, cos, csqrt, erfc, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-06 17:30:18 +00:00
erfc 0x1.514548p+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
erfc 0x2.36c504p+0
erfc 0x1.65e31p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
erfc 0xd.44cd3p-4
erfc 0xd.47425b3cafa48p-4
Add more randomly-generated libm tests. This patch adds more libm test inputs found through random test generation to increase observed ulps on x86_64. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, atanh, cbrt, cosh, csqrt, erfc, expm1 and lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-11 15:03:10 +00:00
erfc 0x1.2f644ep+0
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
erfc 0x2.56af04p+0
erfc 0x2.b7f8cb76737d2af98dead7c4c5eep+0
erfc 0x2.dfb9b4p+0
erfc 0x1.e33c9ep+0
add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)
2020-11-25 18:03:33 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
erfc 0x1.3ffcbf39febb4p+0
Add more tests of erf, erfc. This patch adds some randomly-generated tests of erf and erfc that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of erf and erfc. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 17:49:44 +00:00
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp 0
exp -0
exp 1
exp 2
exp 3
exp 0.75
exp 50.0
exp 88.72269439697265625
exp 709.75
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp 1000.0
exp 710
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp -1234
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp 0x2.c679d1f73f0fb628p+8
exp 1e5
exp max
Fix exp missing underflows (bug 15268, bug 15425).
2013-12-03 21:49:56 +00:00
exp -7.4444006192138124e+02
exp -0x1.75f113c30b1c8p+9
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp -max
BZ #16447: Fix ldbl-128 expl implementation. Extend the range of numbers handled via unsafe mode. Add expl testcase and regenerate ULPs for s390.
2014-02-11 12:47:47 +00:00
exp -11342.8125
Fix x86 / x86_64 expl, exp10l missing underflows (bug 16361). Similar to various other bugs in this area, the x86 and x86_64 implementations of expl / exp10l can fail to produce underflow exceptions when the unscaled result has trailing 0 bits so the scaling down to subnormal precision is exact. This patch fixes this by forcing the exception in the case of tiny results. Tested for x86_64 and x86. [BZ #16361] * sysdeps/i386/fpu/e_expl.S [!USE_AS_EXPM1L] (cmin): New object. [!USE_AS_EXPM1L] (IEEE754_EXPL): Force underflow exception for tiny results. * sysdeps/x86_64/fpu/e_expl.S [!USE_AS_EXPM1L] (cmin): New object. [!USE_AS_EXPM1L] (IEEE754_EXPL): Force underflow exception for tiny results. * math/auto-libm-test-in: Add more tests of exp and exp10. Do not mark underflow exceptions as possibly missing for bug 16361. * math/auto-libm-test-out: Regenerated.
2015-06-21 17:48:04 +00:00
exp -0x2.c5b2319c4843acc0p12
Fix i386 exp missing underflows (bug 18961). On i386, the double version of exp can miss underflow exceptions if the result is in the subnormal range for double but the last 11 bits of the 64-bit extended-precision mantissa happen to be zero. This patch forces the exception in a similar way to previous fixes. As with the exp2 fixes, the expf changes may in fact not be needed to ensure underflow exceptions, but are included for consistency and to fix the exp part of bug 18875 by ensuring that excess range and precision is removed from underflowing return values. Tested for x86_64 and x86. [BZ #18875] [BZ #18961] * sysdeps/i386/fpu/e_exp.S (dbl_min): New object. (MO): New macro. (__ieee754_exp): For small results, force underflow exception and remove excess range and precision from return value. (__exp_finite): Likewise. * sysdeps/i386/fpu/e_expf.S (flt_min): New object. (MO): New macro. (__ieee754_expf): For small results, force underflow exception and remove excess range and precision from return value. (__expf_finite): Likewise. * math/auto-libm-test-in: Add more tests of exp. * math/auto-libm-test-out: Regenerated.
2015-09-14 22:40:05 +00:00
exp -0x2.c469d9p+8
exp -0x2.c46d96p+8
exp -0x2.c46727p+8
exp -0x2.c469dep+8
exp -0x2.c46c04p+8
exp -0x2.c46adep+8
exp -0x2.c471b3p+8
exp -0x2.c46993p+8
exp -0x2.c49fap+8
exp -0x2.c4ac1p+8
exp -0x2.c4d89p+8
Fix x86/x86_64 expl/exp10l spurious underflows (bug 16348). This patch fixes bug 16348, spurious underflows from x86/x86_64 expl on arguments close to 0. These implementations effectively use expm1 (on the fractional part of the argument) internally, so resulting in spurious underflows when the result is very close to 1. For arguments small enough that the round-to-nearest correct result is 1, this patch uses 1+x instead. These implementations are also used for exp10l and so the patch fixes similar issues there (the 0x1p-67 threshold being small enough to be correct for exp10l as well as expl). But because of spurious underflows in other exp10 implementations (bug 16560), the tests aren't added for exp10 at this point - they can be added when the other exp10 parts of that bug are fixed. Tested x86_64 and x86; no ulps updates needed. [BZ #16348] * sysdeps/i386/fpu/e_expl.S (IEEE754_EXPL) [!USE_AS_EXPM1L]: Use 1+x for argument with exponent below -67. * sysdeps/x86_64/fpu/e_expl.S (IEEE754_EXPL) [!USE_AS_EXPM1L]: Likewise. * math/auto-libm-test-in: Add more tests of exp. * math/auto-libm-test-out: Regenerated.
2014-03-27 18:41:14 +00:00
exp 0x1p-10
exp -0x1p-10
exp 0x1p-20
exp -0x1p-20
exp 0x1p-30
exp -0x1p-30
exp 0x1p-40
exp -0x1p-40
exp 0x1p-50
exp -0x1p-50
exp 0x1p-60
exp -0x1p-60
exp 0x1p-100
exp -0x1p-100
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp 0x1p-600
exp -0x1p-600
exp 0x1p-10000
exp -0x1p-10000
exp 0x5.8b90b8p+4
exp 0x5.8b90cp+4
exp -0x5.75628p+4
exp -0x5.756278p+4
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp 0x2.c5c85fdf473dep+8
exp 0x2.c5c85fdf473ep+8
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp -0x2.c4657baf579a6p+8
exp -0x2.c4657baf579a4p+8
Update libm-test XFAILs for ibm128 format. This patch cleans up and updates the libm-test XFAILs for the ibm128 format. More of them are changed to use a new ibm128-libgcc conditional, to reflect that they are not in fact needed if you've patched libgcc to fix the known issues (at substantial performance cost). Many additional XFAILs are added for tests that fail with unpatched libgcc (most but not all of them xfail-rounding). Note that further such fixes will be needed for test-ldouble actually to pass with default libgcc (in particular, XFAILs for pow tests and for various affected tests directly embedded in libm-test.inc). With patched libgcc, there may be a few XFAILs needed but the results are already substantially clean apart from a few ulps differences. Tested for powerpc. * math/libm-test.inc (TEST_COND_ibm128_libgcc): New macro. (init_max_error) [TEST_COND_ibm128]: Increase maximum error allowed to 16 ulps. * math/auto-libm-test-in: Change most XFAILs for ibm128 to use ibm128-libgcc. XFAIL more tests for ibm128-libgcc. * math/auto-libm-test-out: Regenerated.
2017-01-04 23:29:42 +00:00
exp 0x2.c5c85fdf473de6ab278ece600fp+8 xfail-rounding:ibm128-libgcc
exp 0x2.c5c85fdf473de6ab278ece601p+8 xfail-rounding:ibm128-libgcc
exp -0x2.9fa8dcb9092a538b3f2ee2ca67p+8 xfail-rounding:ibm128-libgcc
exp -0x2.9fa8dcb9092a538b3f2ee2ca66p+8 xfail-rounding:ibm128-libgcc
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp 0x2.c5c85fdf473de6acp+12
exp 0x2.c5c85fdf473de6bp+12
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp -0x2.c5b2319c4843accp+12
exp -0x2.c5b2319c4843acbcp+12
exp -0x2.c5bd48bdc7c0c9b8p+12
exp -0x2.c5bd48bdc7c0c9b4p+12
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp 0x2.c5c85fdf473de6af278ece600fcap+12
exp 0x2.c5c85fdf473de6af278ece600fccp+12
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp -0x2.c5b2319c4843acbff21591e99cccp+12
exp -0x2.c5b2319c4843acbff21591e99ccap+12
Fix x86/x86_64 expl/exp10l spurious underflows (bug 16348). This patch fixes bug 16348, spurious underflows from x86/x86_64 expl on arguments close to 0. These implementations effectively use expm1 (on the fractional part of the argument) internally, so resulting in spurious underflows when the result is very close to 1. For arguments small enough that the round-to-nearest correct result is 1, this patch uses 1+x instead. These implementations are also used for exp10l and so the patch fixes similar issues there (the 0x1p-67 threshold being small enough to be correct for exp10l as well as expl). But because of spurious underflows in other exp10 implementations (bug 16560), the tests aren't added for exp10 at this point - they can be added when the other exp10 parts of that bug are fixed. Tested x86_64 and x86; no ulps updates needed. [BZ #16348] * sysdeps/i386/fpu/e_expl.S (IEEE754_EXPL) [!USE_AS_EXPM1L]: Use 1+x for argument with exponent below -67. * sysdeps/x86_64/fpu/e_expl.S (IEEE754_EXPL) [!USE_AS_EXPM1L]: Likewise. * math/auto-libm-test-in: Add more tests of exp. * math/auto-libm-test-out: Regenerated.
2014-03-27 18:41:14 +00:00
exp min
exp -min
exp min_subnorm
exp -min_subnorm
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
Add more tests of exp, exp10, exp2, expm1. This patch adds some randomly-generated tests of exp, exp10, exp2 and expm1 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 20:33:04 +00:00
exp -0x1.760cd14774bd9p+0
exp 0x1.4bed28p+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
exp -0x1.f1cf36p+8
exp 0x3.248524p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
exp 0x1.f0b362p+0
exp 0xd.89746a799ac4eedp+0
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
exp -0x6.58b64p-4
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
exp 0x1.fefe02p-16
add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)
2020-11-25 18:03:33 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
exp -0x1.0c4af1ad4ee36p+9
Add more tests of exp, exp10, exp2, expm1. This patch adds some randomly-generated tests of exp, exp10, exp2 and expm1 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 20:33:04 +00:00
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp10 0
exp10 -0
exp10 3
exp10 -1
exp10 36
exp10 -36
exp10 305
exp10 -305
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp10 4932
Fix x86 / x86_64 expl, exp10l missing underflows (bug 16361). Similar to various other bugs in this area, the x86 and x86_64 implementations of expl / exp10l can fail to produce underflow exceptions when the unscaled result has trailing 0 bits so the scaling down to subnormal precision is exact. This patch fixes this by forcing the exception in the case of tiny results. Tested for x86_64 and x86. [BZ #16361] * sysdeps/i386/fpu/e_expl.S [!USE_AS_EXPM1L] (cmin): New object. [!USE_AS_EXPM1L] (IEEE754_EXPL): Force underflow exception for tiny results. * sysdeps/x86_64/fpu/e_expl.S [!USE_AS_EXPM1L] (cmin): New object. [!USE_AS_EXPM1L] (IEEE754_EXPL): Force underflow exception for tiny results. * math/auto-libm-test-in: Add more tests of exp and exp10. Do not mark underflow exceptions as possibly missing for bug 16361. * math/auto-libm-test-out: Regenerated.
2015-06-21 17:48:04 +00:00
exp10 -4932
exp10 -0x1.343793004f503232p12
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp10 1e5
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp10 -1e5
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp10 1e6
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp10 -1e6
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp10 max
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp10 -max
exp10 0.75
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp10 0x1.348e45573a1dd72cp+8
Fix i386 exp10 missing underflows (bug 18966). On i386, the double version of exp10 can miss underflow exceptions if the result is in the subnormal range for double but the last 11 bits of the 64-bit extended-precision mantissa happen to be zero. This patch forces the exception in a similar way to previous fixes. As with the exp2 and exp fixes, the exp10f changes may in fact not be needed to ensure underflow exceptions, but are included for consistency and to fix the exp10 part of bug 18875 by ensuring that excess range and precision is removed from underflowing return values. Tested for x86_64 and x86. [BZ #18875] [BZ #18966] * sysdeps/i386/fpu/e_exp10.S (dbl_min): New object. (MO): New macro. (__ieee754_exp10): For small results, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/e_exp10f.S (flt_min): New object. (MO): New macro. (__ieee754_exp10f): For small results, force underflow exception and remove excess range and precision from return value. * math/auto-libm-test-in: Add more tests of exp10. * math/auto-libm-test-out: Regenerated.
2015-09-15 16:50:02 +00:00
exp10 -0x1.33aa03p+8
exp10 -0x1.33ad17p+8
exp10 -0x1.33afcap+8
Fix exp10 spurious underflows (bug 16560). This patch fixes spurious underflows from exp10 for arguments near 0 (part of bug 16560; that bug also includes spurious underflows from exp2, which are not fixed by this patch). The problem is underflows in the internal computation converting the exp10 argument to arguments for exp (with extra precision), and the fix is simply to return 1 early for arguments near enough to 0 (just as arguments with large enough magnitude have their own overflow / underflow logic at the start of the function). Tested x86_64 and x86 and ulps updated accordingly; also tested for powerpc32 and mips64 to validate the ldbl-128ibm and ldbl-128 changes. [BZ #16560] * sysdeps/ieee754/dbl-64/e_exp10.c (__ieee754_exp10): Return 1 for arguments close to 0. * sysdeps/ieee754/ldbl-128/e_exp10l.c (__ieee754_exp10l): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_exp10l.c (__ieee754_exp10l): Likewise. * math/auto-libm-test-in: Add more tests of exp10. * math/auto-libm-test-out: Regenerated. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2014-06-25 11:33:22 +00:00
exp10 0x1p-10
exp10 -0x1p-10
exp10 0x1p-20
exp10 -0x1p-20
exp10 0x1p-30
exp10 -0x1p-30
exp10 0x1p-40
exp10 -0x1p-40
exp10 0x1p-50
exp10 -0x1p-50
exp10 0x1p-60
exp10 -0x1p-60
exp10 0x1p-100
exp10 -0x1p-100
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp10 0x1p-600
exp10 -0x1p-600
exp10 0x1p-10000
exp10 -0x1p-10000
exp10 0x2.688268p+4
exp10 0x2.68826cp+4
exp10 -0x2.5ee064p+4
exp10 -0x2.5ee06p+4
exp10 0x1.34413509f79fep+8
exp10 0x1.34413509f79ffp+8
exp10 -0x1.33a7146f72a42p+8
exp10 -0x1.33a7146f72a41p+8
Update libm-test XFAILs for ibm128 format. This patch cleans up and updates the libm-test XFAILs for the ibm128 format. More of them are changed to use a new ibm128-libgcc conditional, to reflect that they are not in fact needed if you've patched libgcc to fix the known issues (at substantial performance cost). Many additional XFAILs are added for tests that fail with unpatched libgcc (most but not all of them xfail-rounding). Note that further such fixes will be needed for test-ldouble actually to pass with default libgcc (in particular, XFAILs for pow tests and for various affected tests directly embedded in libm-test.inc). With patched libgcc, there may be a few XFAILs needed but the results are already substantially clean apart from a few ulps differences. Tested for powerpc. * math/libm-test.inc (TEST_COND_ibm128_libgcc): New macro. (init_max_error) [TEST_COND_ibm128]: Increase maximum error allowed to 16 ulps. * math/auto-libm-test-in: Change most XFAILs for ibm128 to use ibm128-libgcc. XFAIL more tests for ibm128-libgcc. * math/auto-libm-test-out: Regenerated.
2017-01-04 23:29:42 +00:00
exp10 0x1.34413509f79fef2f625b0205a88p+8 xfail:ibm128-libgcc
exp10 0x1.34413509f79fef2f625b0205a9p+8 xfail:ibm128-libgcc
exp10 -0x1.23b2b470ae9318183ba772361cp+8 xfail-rounding:ibm128-libgcc
exp10 -0x1.23b2b470ae9318183ba772361b8p+8 xfail-rounding:ibm128-libgcc
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp10 0x1.34413509f79fef3p+12
exp10 0x1.34413509f79fef32p+12
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp10 -0x1.343793004f503232p+12
exp10 -0x1.343793004f50323p+12
exp10 -0x1.343c6405237810b2p+12
exp10 -0x1.343c6405237810bp+12
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp10 0x1.34413509f79fef311f12b35816f9p+12
exp10 0x1.34413509f79fef311f12b35816fap+12
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp10 -0x1.343793004f503231a589bac27c39p+12
exp10 -0x1.343793004f503231a589bac27c38p+12
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
exp10 -0xa.6f431p+0
Fix exp10 spurious underflows (bug 16560). This patch fixes spurious underflows from exp10 for arguments near 0 (part of bug 16560; that bug also includes spurious underflows from exp2, which are not fixed by this patch). The problem is underflows in the internal computation converting the exp10 argument to arguments for exp (with extra precision), and the fix is simply to return 1 early for arguments near enough to 0 (just as arguments with large enough magnitude have their own overflow / underflow logic at the start of the function). Tested x86_64 and x86 and ulps updated accordingly; also tested for powerpc32 and mips64 to validate the ldbl-128ibm and ldbl-128 changes. [BZ #16560] * sysdeps/ieee754/dbl-64/e_exp10.c (__ieee754_exp10): Return 1 for arguments close to 0. * sysdeps/ieee754/ldbl-128/e_exp10l.c (__ieee754_exp10l): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_exp10l.c (__ieee754_exp10l): Likewise. * math/auto-libm-test-in: Add more tests of exp10. * math/auto-libm-test-out: Regenerated. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2014-06-25 11:33:22 +00:00
exp10 min
exp10 -min
exp10 min_subnorm
exp10 -min_subnorm
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
Add more tests of exp, exp10, exp2, expm1. This patch adds some randomly-generated tests of exp, exp10, exp2 and expm1 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 20:33:04 +00:00
exp10 0xd.f73d6p-4
exp10 0x1.cc6776p+0
exp10 0x5.b00bcd891ffe56fp+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
exp10 0xe.8b349p+4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
exp10 0x3.495c78p+0
exp10 0xf.f33f6p+0
Add more tests of exp, exp10, exp2, expm1. This patch adds some randomly-generated tests of exp, exp10, exp2 and expm1 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 20:33:04 +00:00
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp2 0
exp2 -0
exp2 10
exp2 -1
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp2 1e6
Fix exp2 errno setting on underflow (bug 16283).
2013-12-03 20:50:51 +00:00
exp2 -1e6
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp2 max
Fix exp2 errno setting on underflow (bug 16283).
2013-12-03 20:50:51 +00:00
exp2 -max
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp2 0.75
exp2 100.5
Fix exp2, exp2f spurious underflows (bug 18219). The dbl-64 and flt-32 implementations of exp2 functions produce spurious underflow exceptions. The underlying reason is the same in both cases: the computation works as (2^a - 1)*2^b + 2^b for suitably chosen a and b, where a has small magnitude so 2^a - 1 can be computed with a low-degree polynomial approximation, and (2^a - 1)*2^b can underflow even when the final result does not. This patch fixes this by adjusting the threshold for when scaling is used to avoid intermediate underflow so it works for any possible value of a where the final result would not underflow. Tested for x86_64 and x86. [BZ #18219] * sysdeps/ieee754/dbl-64/e_exp2.c (__ieee754_exp2): Reduce threshold on absolute value of exponent for which scaling is used. * sysdeps/ieee754/flt-32/e_exp2f.c (__ieee754_exp2f): Likewise. * math/auto-libm-test-in: Add more tests of exp2. * math/auto-libm-test-out: Regenerated.
2015-06-23 14:35:18 +00:00
exp2 -116.5
exp2 -123.5
exp2 -124.5
exp2 -125.5
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp2 127
exp2 -149
exp2 1000.25
Fix exp2, exp2f spurious underflows (bug 18219). The dbl-64 and flt-32 implementations of exp2 functions produce spurious underflow exceptions. The underlying reason is the same in both cases: the computation works as (2^a - 1)*2^b + 2^b for suitably chosen a and b, where a has small magnitude so 2^a - 1 can be computed with a low-degree polynomial approximation, and (2^a - 1)*2^b can underflow even when the final result does not. This patch fixes this by adjusting the threshold for when scaling is used to avoid intermediate underflow so it works for any possible value of a where the final result would not underflow. Tested for x86_64 and x86. [BZ #18219] * sysdeps/ieee754/dbl-64/e_exp2.c (__ieee754_exp2): Reduce threshold on absolute value of exponent for which scaling is used. * sysdeps/ieee754/flt-32/e_exp2f.c (__ieee754_exp2f): Likewise. * math/auto-libm-test-in: Add more tests of exp2. * math/auto-libm-test-out: Regenerated.
2015-06-23 14:35:18 +00:00
exp2 -1019.5
exp2 -1020.5
exp2 -1021.5
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
exp2 1023
Fix exp2 errno setting on underflow (bug 16283).
2013-12-03 20:50:51 +00:00
exp2 -1074
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp2 16383
Fix exp2 errno setting on underflow (bug 16283).
2013-12-03 20:50:51 +00:00
exp2 -16400
Fix exp2 missing underflows (bug 16521). Various exp2 implementations in glibc can miss underflow exceptions when the scaling down part of the calculation is exact (or, in the x86 case, when the conversion from extended precision to the target precision is exact). This patch forces the exception in a similar way to previous fixes. The x86 exp2f changes may in fact not be needed for this purpose - it's likely to be the case that no argument of type float has an exp2 result so close to an exact subnormal float value that it equals that value when rounded to 64 bits (even taking account of variation between different x86 implementations). However, they are included for consistency with the changes to exp2 and so as to fix the exp2f part of bug 18875 by ensuring that excess range and precision is removed from underflowing return values. Tested for x86_64, x86 and mips64. [BZ #16521] [BZ #18875] * math/e_exp2l.c (__ieee754_exp2l): Force underflow exception for small results. * sysdeps/i386/fpu/e_exp2.S (dbl_min): New object. (MO): New macro. (__ieee754_exp2): For small results, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/e_exp2f.S (flt_min): New object. (MO): New macro. (__ieee754_exp2f): For small results, force underflow exception and remove excess range and precision from return value. * sysdeps/i386/fpu/e_exp2l.S (ldbl_min): New object. (MO): New macro. (__ieee754_exp2l): Force underflow exception for small results. * sysdeps/ieee754/dbl-64/e_exp2.c (__ieee754_exp2): Likewise. * sysdeps/ieee754/flt-32/e_exp2f.c (__ieee754_exp2f): Likewise. * sysdeps/x86_64/fpu/e_exp2l.S (ldbl_min): New object. (MO): New macro. (__ieee754_exp2l): Force underflow exception for small results. * math/auto-libm-test-in: Add more tests or exp2. * math/auto-libm-test-out: Regenerated.
2015-09-14 22:00:12 +00:00
exp2 -126.125
exp2 -126.25
exp2 -126.375
exp2 -126.5
exp2 -126.625
exp2 -126.75
exp2 -126.875
exp2 -1022.125
exp2 -1022.25
exp2 -1022.375
exp2 -1022.5
exp2 -1022.625
exp2 -1022.75
exp2 -1022.875
exp2 -0x3.fe4e8p+8
exp2 -0x3.fe513p+8
exp2 -16382.125
exp2 -16382.25
exp2 -16382.375
exp2 -16382.5
exp2 -16382.625
exp2 -16382.75
exp2 -16382.875
Fix exp2 spurious underflows (bug 16560). This patch fixes the remaining part of bug 16560, spurious underflows from exp2 of arguments close to 0 (when the result is close to 1, so should not underflow), by just using 1+x instead of a more complicated calculation when the argument is sufficiently small. Tested for x86_64, x86 and mips64. [BZ #16560] * math/e_exp2l.c [LDBL_MANT_DIG == 106] (LDBL_EPSILON): Undefine and redefine. (__ieee754_exp2l): Do not multiply small fractional parts by M_LN2l. * sysdeps/i386/fpu/e_exp2l.S (__ieee754_exp2l): Just add 1 to small argument. * sysdeps/ieee754/dbl-64/e_exp2.c (__ieee754_exp2): Likewise. * sysdeps/ieee754/flt-32/e_exp2f.c (__ieee754_exp2f): Likewise. * sysdeps/x86_64/fpu/e_exp2l.S (__ieee754_exp2l): Likewise. * math/auto-libm-test-in: Add more tests of exp2. * math/auto-libm-test-out: Regenerated.
2015-02-12 19:02:45 +00:00
exp2 0x1p-10
exp2 -0x1p-10
exp2 0x1p-20
exp2 -0x1p-20
exp2 0x1p-30
exp2 -0x1p-30
exp2 0x1p-40
exp2 -0x1p-40
exp2 0x1p-50
exp2 -0x1p-50
exp2 0x1p-60
exp2 -0x1p-60
exp2 0x1p-100
exp2 -0x1p-100
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp2 0x1p-600
exp2 -0x1p-600
exp2 0x1p-10000
exp2 -0x1p-10000
exp2 0x7.fffff8p+4
exp2 0x8.00001p+4
exp2 -0x7.e00008p+4
exp2 -0x7.dffff8p+4
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp2 0x3.ffffffffffffep+8
exp2 0x4.0000000000004p+8
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp2 -0x3.fe00000000002p+8
exp2 -0x3.fdffffffffffep+8
Update libm-test XFAILs for ibm128 format. This patch cleans up and updates the libm-test XFAILs for the ibm128 format. More of them are changed to use a new ibm128-libgcc conditional, to reflect that they are not in fact needed if you've patched libgcc to fix the known issues (at substantial performance cost). Many additional XFAILs are added for tests that fail with unpatched libgcc (most but not all of them xfail-rounding). Note that further such fixes will be needed for test-ldouble actually to pass with default libgcc (in particular, XFAILs for pow tests and for various affected tests directly embedded in libm-test.inc). With patched libgcc, there may be a few XFAILs needed but the results are already substantially clean apart from a few ulps differences. Tested for powerpc. * math/libm-test.inc (TEST_COND_ibm128_libgcc): New macro. (init_max_error) [TEST_COND_ibm128]: Increase maximum error allowed to 16 ulps. * math/auto-libm-test-in: Change most XFAILs for ibm128 to use ibm128-libgcc. XFAIL more tests for ibm128-libgcc. * math/auto-libm-test-out: Regenerated.
2017-01-04 23:29:42 +00:00
exp2 0x3.fffffffffffffffa3aae26b51fp+8 xfail-rounding:ibm128-libgcc
exp2 0x3.fffffffffffffffa3aae26b52p+8 xfail-rounding:ibm128-libgcc
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp2 -0x3.c9000000000000000000000001p+8
exp2 -0x3.c8ffffffffffffffffffffffffp+8
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp2 0x3.fffffffffffffffcp+12
exp2 0x4.0000000000000008p+12
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp2 -0x3.ffe0000000000004p+12
exp2 -0x3.ffdffffffffffffcp+12
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
exp2 -0x3.fff0000000000004p+12
exp2 -0x3.ffeffffffffffffcp+12
exp2 0x3.fffffffffffffffffffffffffffep+12
exp2 0x4.0000000000000000000000000004p+12
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
exp2 -0x3.ffe0000000000000000000000002p+12
exp2 -0x3.ffdffffffffffffffffffffffffep+12
Fix exp2 spurious underflows (bug 16560). This patch fixes the remaining part of bug 16560, spurious underflows from exp2 of arguments close to 0 (when the result is close to 1, so should not underflow), by just using 1+x instead of a more complicated calculation when the argument is sufficiently small. Tested for x86_64, x86 and mips64. [BZ #16560] * math/e_exp2l.c [LDBL_MANT_DIG == 106] (LDBL_EPSILON): Undefine and redefine. (__ieee754_exp2l): Do not multiply small fractional parts by M_LN2l. * sysdeps/i386/fpu/e_exp2l.S (__ieee754_exp2l): Just add 1 to small argument. * sysdeps/ieee754/dbl-64/e_exp2.c (__ieee754_exp2): Likewise. * sysdeps/ieee754/flt-32/e_exp2f.c (__ieee754_exp2f): Likewise. * sysdeps/x86_64/fpu/e_exp2l.S (__ieee754_exp2l): Likewise. * math/auto-libm-test-in: Add more tests of exp2. * math/auto-libm-test-out: Regenerated.
2015-02-12 19:02:45 +00:00
exp2 min
exp2 -min
exp2 min_subnorm
exp2 -min_subnorm
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
Add more tests of exp, exp10, exp2, expm1. This patch adds some randomly-generated tests of exp, exp10, exp2 and expm1 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 20:33:04 +00:00
exp2 0xb.71754p-4
exp2 0xd.d77dp+0
exp2 0xc.122c4p-4
exp2 -0x1.567cc8p+0
exp2 -0x1.bbbd76p+0
exp2 -0x1.3045fep+8
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
exp2 0xa.87b8bp+0
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
exp2 -0xe.2ce69p-4
exp2 -0xc.1bf12p-16
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
exp2 -0x4.8ce878p-4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
exp2 0xf.93d18bf7be8d272p-4
Add more tests of exp, exp10, exp2, expm1. This patch adds some randomly-generated tests of exp, exp10, exp2 and expm1 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 20:33:04 +00:00
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
expm1 0
expm1 -0
expm1 1
expm1 0.75
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
expm1 2
expm1 3
expm1 4
expm1 5
expm1 10
expm1 15
expm1 20
expm1 25
expm1 30
expm1 35
expm1 40
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
expm1 50.0
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
expm1 60
expm1 70
expm1 80
expm1 90
expm1 100
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
expm1 127.0
expm1 500.0
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
expm1 11356.25
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
expm1 -10.0
expm1 -16.0
expm1 -17.0
expm1 -18.0
expm1 -36.0
expm1 -37.0
expm1 -38.0
expm1 -44.0
expm1 -45.0
expm1 -46.0
expm1 -73.0
expm1 -74.0
expm1 -75.0
expm1 -78.0
expm1 -79.0
expm1 -80.0
expm1 -100.0
expm1 -1000.0
expm1 -10000.0
expm1 -100000.0
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
expm1 100000.0
expm1 max
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
expm1 -max
Fix x86/x86_64 expm1 inaccuracy near 0 in directed rounding modes (bug 16293). Bug 16293 is inaccuracy of x86/x86_64 versions of expm1, near 0 in directed rounding modes, that arises from frndint rounding the exponent to 1 or -1 instead of 0, resulting in large cancellation error. This inaccuracy in turn affects other functions such as sinh that use expm1. This patch fixes the problem by setting round-to-nearest mode temporarily around the affected calls to frndint. I don't think this is needed for other uses of frndint, such as in exp itself, as only for expm1 is the cancellation error significant. Tested x86_64 and x86 and ulps updated accordingly. * sysdeps/i386/fpu/e_expl.S (IEEE754_EXPL) [USE_AS_EXPM1L]: Set round-to-nearest mode when using frndint. * sysdeps/i386/fpu/s_expm1.S (__expm1): Likewise. * sysdeps/i386/fpu/s_expm1f.S (__expm1f): Likewise. * sysdeps/x86_64/fpu/e_expl.S (IEEE754_EXPL) [USE_AS_EXPM1L]: Likewise. * math/auto-libm-test-in: Add more tests of expm1. Do not expect sinh test to fail. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): Remove macro. (TEST_COND_x86): Likewise. (expm1_tonearest_test_data): New array. (expm1_test_tonearest): New function. (expm1_towardzero_test_data): New array. (expm1_test_towardzero): New function. (expm1_downward_test_data): New array. (expm1_test_downward): New function. (expm1_upward_test_data): New array. (expm1_test_upward): New function. (main): Run the new test functions. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-19 13:36:10 +00:00
expm1 0x1p-2
expm1 -0x1p-2
expm1 0x1p-10
expm1 -0x1p-10
expm1 0x1p-20
expm1 -0x1p-20
expm1 0x1p-29
expm1 -0x1p-29
expm1 0x1p-32
expm1 -0x1p-32
expm1 0x1p-50
expm1 -0x1p-50
expm1 0x1p-64
expm1 -0x1p-64
expm1 0x1p-100
expm1 -0x1p-100
Improve test coverage of real libm functions [a-e]*. This patch improves test coverage of the real libm functions [a-e]*, ensuring that special cases and ranges of input values of potential significance (such as close to overflow and underflow thresholds) are more systematically covered. This is a followup to <https://sourceware.org/ml/libc-alpha/2013-12/msg00757.html> which covered [a-c]* (however, I found more weaknesses in the coverage of those functions when preparing this patch, hence the additional tests being added for them here). Addition of a test for acosh (-qNaN) is temporarily deferred, to be included as part of a fix for bug 19032 which was discovered in the course of adding these tests (and which illustrates the use of testing -qNaN as well as +qNaN as input even to functions for which the sign of a NaN isn't meant to be significant). Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, atan, atan2, atanh, cbrt, cos, cosh, erf, erfc, exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (acos_test_data): Add more tests. (asin_test_data): Likewise. (asinh_test_data): Likewise. (atan_test_data): Likewise. (atanh_test_data): Likewise. (atan2_test_data): Likewise. (cbrt_test_data): Likewise. (ceil_test_data): Likewise. (copysign_test_data): Likewise. (cos_test_data): Likewise. (cosh_test_data): Likewise. (erf_test_data): Likewise. (erfc_test_data): Likewise. (exp_test_data): Likewise. (exp10_test_data): Likewise. (exp2_test_data): Likewise. (expm1_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-30 18:06:02 +00:00
expm1 0x1p-600
expm1 -0x1p-600
expm1 0x1p-10000
expm1 -0x1p-10000
Add more tests of expm1. This patch adds some randomly-generated tests of expm1 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of expm1. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-25 00:05:13 +00:00
expm1 0xe.4152ac57cd1ea7ap-60
expm1 0x6.660247486aed8p-4
expm1 0x6.289a78p-4
expm1 0x6.1b4d318238d4a2a8p-4
Add more tests of exp, exp10, exp2, expm1. This patch adds some randomly-generated tests of exp, exp10, exp2 and expm1 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of exp, exp10, exp2 and expm1. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 20:33:04 +00:00
expm1 0x5.fb8dc64e91a74p-4
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
expm1 0x3.735f497c4e67535cp-4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
expm1 -0x7.d6c50b469d404p+0
Add more randomly-generated libm tests. This patch adds more libm test inputs found through random test generation to increase observed ulps on x86_64. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, atanh, cbrt, cosh, csqrt, erfc, expm1 and lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-11 15:03:10 +00:00
expm1 0x4.857de8p+4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
expm1 0x5.dfeb68p-4
Fix expm1 missing underflows (bug 16353). Similar to various other bugs in this area, some expm1 implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. (The issue does not apply to the ldbl-* implementations or to those for x86 / x86_64 long double. The change to sysdeps/ieee754/dbl-64/wordsize-64/e_cosh.c is one I missed when previously fixing bug 16354; the bug in that implementation was previously latent, but the expm1 fixes stopped it being latent and so required it to be fixed to avoid spurious underflows from cosh.) Tested for x86_64 and x86. [BZ #16353] * sysdeps/i386/fpu/s_expm1.S (dbl_min): New object. (__expm1): Force underflow exception for arguments with small absolute value. * sysdeps/i386/fpu/s_expm1f.S (flt_min): New object. (__expm1f): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/dbl-64/s_expm1.c: Include <float.h>. (__expm1): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/s_expm1f.c: Include <float.h>. (__expm1f): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/dbl-64/wordsize-64/e_cosh.c (__ieee754_cosh): Check for small arguments before calling __expm1. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 16353. * math/auto-libm-test-out: Regenerated.
2015-06-22 21:06:19 +00:00
expm1 0x4.0000000000000028p-16384
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (ldbl-96)
expm1 0x5.8b911eb6733469c8p-4
# the next value generates larger error bounds on x86_64 (binary128)
expm1 0x5.a343df0d680099a7a1a873a751a8p-4
Fix expm1 missing underflows (bug 16353). Similar to various other bugs in this area, some expm1 implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. (The issue does not apply to the ldbl-* implementations or to those for x86 / x86_64 long double. The change to sysdeps/ieee754/dbl-64/wordsize-64/e_cosh.c is one I missed when previously fixing bug 16354; the bug in that implementation was previously latent, but the expm1 fixes stopped it being latent and so required it to be fixed to avoid spurious underflows from cosh.) Tested for x86_64 and x86. [BZ #16353] * sysdeps/i386/fpu/s_expm1.S (dbl_min): New object. (__expm1): Force underflow exception for arguments with small absolute value. * sysdeps/i386/fpu/s_expm1f.S (flt_min): New object. (__expm1f): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/dbl-64/s_expm1.c: Include <float.h>. (__expm1): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/s_expm1f.c: Include <float.h>. (__expm1f): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/dbl-64/wordsize-64/e_cosh.c (__ieee754_cosh): Check for small arguments before calling __expm1. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 16353. * math/auto-libm-test-out: Regenerated.
2015-06-22 21:06:19 +00:00
expm1 min
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
expm1 -min
Fix expm1 missing underflows (bug 16353). Similar to various other bugs in this area, some expm1 implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. (The issue does not apply to the ldbl-* implementations or to those for x86 / x86_64 long double. The change to sysdeps/ieee754/dbl-64/wordsize-64/e_cosh.c is one I missed when previously fixing bug 16354; the bug in that implementation was previously latent, but the expm1 fixes stopped it being latent and so required it to be fixed to avoid spurious underflows from cosh.) Tested for x86_64 and x86. [BZ #16353] * sysdeps/i386/fpu/s_expm1.S (dbl_min): New object. (__expm1): Force underflow exception for arguments with small absolute value. * sysdeps/i386/fpu/s_expm1f.S (flt_min): New object. (__expm1f): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/dbl-64/s_expm1.c: Include <float.h>. (__expm1): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/s_expm1f.c: Include <float.h>. (__expm1f): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/dbl-64/wordsize-64/e_cosh.c (__ieee754_cosh): Check for small arguments before calling __expm1. * math/auto-libm-test-in: Do not mark underflow exceptions as possibly missing for bug 16353. * math/auto-libm-test-out: Regenerated.
2015-06-22 21:06:19 +00:00
expm1 min_subnorm
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
expm1 -min_subnorm
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 1.0 2.0 3.0
fma 1.25 0.75 0.0625
fma 0 0 0
fma 0 0 -0
fma 0 -0 0
fma 0 -0 -0
fma -0 0 0
fma -0 0 -0
fma -0 -0 0
fma -0 -0 -0
fma 1.0 0 0
fma 1.0 0 -0
fma 1.0 -0 0
fma 1.0 -0 -0
fma -1.0 0 0
fma -1.0 0 -0
fma -1.0 -0 0
fma -1.0 -0 -0
fma 0 1.0 0
fma 0 1.0 -0
fma 0 -1.0 0
fma 0 -1.0 -0
fma -0 1.0 0
fma -0 1.0 -0
fma -0 -1.0 0
fma -0 -1.0 -0
fma 1.0 1.0 -1.0
fma 1.0 -1.0 1.0
fma -1.0 1.0 1.0
fma -1.0 -1.0 -1.0
soft-fp: Fix _FP_FMA when product is zero and third argument is finite (bug 17932). soft-fp's _FP_FMA fails to set the result's exponent for cases where the result of the multiplication is 0, yielding incorrect (arbitrary, depending on uninitialized values) results for those cases. This affects libm for architectures using soft-fp to implement fma. This patch adds the exponent setting and tests for this case. Tested for ARM soft-float (which uses soft-fp fma), x86_64 and x86 (to verify not introducing new libm test failures there). (This bug showed up in testing my patch to move the Linux kernel to current soft-fp. math/Makefile has "override CFLAGS += -Wno-uninitialized" which would have stopped compiler warnings from showing up this problem, although I wouldn't be surprised if removing that shows spurious warnings from this code, if the compiler fails to follow that various cases where the exponent is uninitialized don't need it initialized because the class is set to a value meaning the uninitialized exponent isn't used.) [BZ #17932] * soft-fp/op-common.h (_FP_FMA): Set exponent of result in case where multiplication results in zero and third argument is finite and nonzero. * math/auto-libm-test-in: Add more tests of fma. * math/auto-libm-test-out: Regenerated.
2015-02-06 15:44:07 +00:00
fma 0 0 1
fma 0 0 2
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0 0 max xfail-rounding:ibm128-libgcc missing-errno
soft-fp: Fix _FP_FMA when product is zero and third argument is finite (bug 17932). soft-fp's _FP_FMA fails to set the result's exponent for cases where the result of the multiplication is 0, yielding incorrect (arbitrary, depending on uninitialized values) results for those cases. This affects libm for architectures using soft-fp to implement fma. This patch adds the exponent setting and tests for this case. Tested for ARM soft-float (which uses soft-fp fma), x86_64 and x86 (to verify not introducing new libm test failures there). (This bug showed up in testing my patch to move the Linux kernel to current soft-fp. math/Makefile has "override CFLAGS += -Wno-uninitialized" which would have stopped compiler warnings from showing up this problem, although I wouldn't be surprised if removing that shows spurious warnings from this code, if the compiler fails to follow that various cases where the exponent is uninitialized don't need it initialized because the class is set to a value meaning the uninitialized exponent isn't used.) [BZ #17932] * soft-fp/op-common.h (_FP_FMA): Set exponent of result in case where multiplication results in zero and third argument is finite and nonzero. * math/auto-libm-test-in: Add more tests of fma. * math/auto-libm-test-out: Regenerated.
2015-02-06 15:44:07 +00:00
fma 0 1 1
fma 1 0 1
fma 0 1 2
fma 1 0 2
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0 1 max xfail-rounding:ibm128-libgcc missing-errno
fma 1 0 max xfail-rounding:ibm128-libgcc missing-errno
soft-fp: Fix _FP_FMA when product is zero and third argument is finite (bug 17932). soft-fp's _FP_FMA fails to set the result's exponent for cases where the result of the multiplication is 0, yielding incorrect (arbitrary, depending on uninitialized values) results for those cases. This affects libm for architectures using soft-fp to implement fma. This patch adds the exponent setting and tests for this case. Tested for ARM soft-float (which uses soft-fp fma), x86_64 and x86 (to verify not introducing new libm test failures there). (This bug showed up in testing my patch to move the Linux kernel to current soft-fp. math/Makefile has "override CFLAGS += -Wno-uninitialized" which would have stopped compiler warnings from showing up this problem, although I wouldn't be surprised if removing that shows spurious warnings from this code, if the compiler fails to follow that various cases where the exponent is uninitialized don't need it initialized because the class is set to a value meaning the uninitialized exponent isn't used.) [BZ #17932] * soft-fp/op-common.h (_FP_FMA): Set exponent of result in case where multiplication results in zero and third argument is finite and nonzero. * math/auto-libm-test-in: Add more tests of fma. * math/auto-libm-test-out: Regenerated.
2015-02-06 15:44:07 +00:00
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
# Bug 6801: errno setting may be missing. This applies to both plain
# and narrowing fma, and is only commented here once.
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma min min 0 missing-errno
fma min min -0 missing-errno
fma min -min 0 missing-errno
fma min -min -0 missing-errno
fma -min min 0 missing-errno
fma -min min -0 missing-errno
fma -min -min 0 missing-errno
fma -min -min -0 missing-errno
Implement proper fmal for ldbl-128ibm (bug 13304). ldbl-128ibm had an implementation of fmal that just did (x * y) + z in most cases, with no attempt at actually being a fused operation. This patch replaces it with a genuine fused operation. It is not necessarily correctly rounding, but should produce a result at least as accurate as the long double arithmetic operations in libgcc, which I think is all that can reasonably be expected for such a non-IEEE format where arithmetic is approximate rather than rounded according to any particular rule for determining the exact result. Like the libgcc arithmetic, it may produce spurious overflow and underflow results, and it falls back to the libgcc multiplication in the case of (finite, finite, zero). This concludes the fixes for bug 13304; any subsequently found fma issues should go in separate Bugzilla bugs. Various other pieces of bug 13304 were fixed in past releases over the past several years. Tested for powerpc. [BZ #13304] * sysdeps/ieee754/ldbl-128ibm/s_fmal.c: Include <fenv.h>, <float.h>, <math_private.h> and <stdlib.h>. (add_split): New function. (mul_split): Likewise. (ext_val): New typedef. (store_ext_val): New function. (mul_ext_val): New function. (compare): New function. (add_split_ext): New function. (__fmal): After checking for Inf, NaN and zero, compute result as an exact sum of scaled double values in round-to-nearest before adding those up and adjusting for other rounding modes. * math/auto-libm-test-in: Remove xfail-rounding:ldbl-128ibm from tests of fma. * math/auto-libm-test-out: Regenerated.
2016-05-19 20:10:56 +00:00
fma max max min missing-errno
fma max max -min missing-errno
fma max -max min missing-errno
fma max -max -min missing-errno
fma -max max min missing-errno
fma -max max -min missing-errno
fma -max -max min missing-errno
fma -max -max -min missing-errno
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x1.7ff8p+13 0x1.000002p+0 0x1.ffffp-24
fma 0x1.fffp+0 0x1.00001p+0 -0x1.fffp+0
fma 0x1.9abcdep+127 0x0.9abcdep-126 -0x1.f08948p+0
fma 0x1.9abcdep+100 0x0.9abcdep-126 -0x1.f08948p-27
fma 0x1.fffffep+127 0x1.001p+0 -0x1.fffffep+127
fma -0x1.fffffep+127 0x1.fffffep+0 0x1.fffffep+127
fma 0x1.fffffep+127 2.0 -0x1.fffffep+127
fma 0x1.4p-126 0x1.000004p-1 0x1p-128
fma -0x1.4p-126 0x1.000004p-1 -0x1p-128
fma 0x1.fffff8p-126 0x1.000002p-1 0x1p-149
fma -0x1.fffff8p-126 0x1.000002p-1 -0x1p-149
fma 0x1p-149 0x1p-1 0x0.fffffep-126
fma -0x1p-149 0x1p-1 -0x0.fffffep-126
fma 0x1p-149 0x1.1p-1 0x0.fffffep-126
fma -0x1p-149 0x1.1p-1 -0x0.fffffep-126
fma 0x1p-149 0x1p-149 0x1p127
fma 0x1p-149 -0x1p-149 0x1p127
fma 0x1p-149 0x1p-149 -0x1p127
fma 0x1p-149 -0x1p-149 -0x1p127
fma 0x1p-149 0x1p-149 0x1p-126
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1p-149 -0x1p-149 0x1p-126 missing-underflow:arg-ibm128
fma 0x1p-149 0x1p-149 -0x1p-126 missing-underflow:arg-ibm128
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x1p-149 -0x1p-149 -0x1p-126
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1p-149 0x1p-149 0x0.fffffep-126 missing-underflow:arg-ibm128
fma 0x1p-149 -0x1p-149 0x0.fffffep-126 missing-underflow:arg-ibm128
fma 0x1p-149 0x1p-149 -0x0.fffffep-126 missing-underflow:arg-ibm128
fma 0x1p-149 -0x1p-149 -0x0.fffffep-126 missing-underflow:arg-ibm128
fma 0x1p-149 0x1p-149 0x1p-149 missing-underflow:arg-ibm128
fma 0x1p-149 -0x1p-149 0x1p-149 missing-errno missing-underflow:arg-ibm128
fma 0x1p-149 0x1p-149 -0x1p-149 missing-errno missing-underflow:arg-ibm128
fma 0x1p-149 -0x1p-149 -0x1p-149 missing-underflow:arg-ibm128
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x0.fffp0 0x0.fffp0 -0x0.ffep0
fma 0x0.fffp0 -0x0.fffp0 0x0.ffep0
fma -0x0.fffp0 0x0.fffp0 0x0.ffep0
fma -0x0.fffp0 -0x0.fffp0 -0x0.ffep0
fma 0x1.000002p-126 0x1.000002p-26 0x1p127
fma 0x1.000002p-126 -0x1.000002p-26 0x1p127
fma 0x1.000002p-126 0x1.000002p-26 -0x1p127
fma 0x1.000002p-126 -0x1.000002p-26 -0x1p127
fma 0x1.000002p-126 0x1.000002p-26 0x1p103
fma 0x1.000002p-126 -0x1.000002p-26 0x1p103
fma 0x1.000002p-126 0x1.000002p-26 -0x1p103
fma 0x1.000002p-126 -0x1.000002p-26 -0x1p103
fma 0x1.7fp+13 0x1.0000000000001p+0 0x1.ffep-48
fma 0x1.fffp+0 0x1.0000000000001p+0 -0x1.fffp+0
fma 0x1.0000002p+0 0x1.ffffffcp-1 0x1p-300
fma 0x1.0000002p+0 0x1.ffffffcp-1 -0x1p-300
fma 0x1.deadbeef2feedp+1023 0x0.deadbeef2feedp-1022 -0x1.a05f8c01a4bfbp+1
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1.deadbeef2feedp+900 0x0.deadbeef2feedp-1022 -0x1.a05f8c01a4bfbp-122 missing-errno
fma 0x1.fffffffffffffp+1023 0x1.001p+0 -0x1.fffffffffffffp+1023 missing-errno
fma -0x1.fffffffffffffp+1023 0x1.fffffffffffffp+0 0x1.fffffffffffffp+1023 missing-errno
fma 0x1.fffffffffffffp+1023 2.0 -0x1.fffffffffffffp+1023 missing-errno
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x1.6a09e667f3bccp-538 0x1.6a09e667f3bccp-538 0.0 missing-errno
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1.deadbeef2feedp-495 0x1.deadbeef2feedp-495 -0x1.bf86a5786a574p-989 missing-errno
fma 0x1.deadbeef2feedp-503 0x1.deadbeef2feedp-503 -0x1.bf86a5786a574p-1005 missing-errno
fma 0x1p-537 0x1p-538 0x1p-1074 missing-errno
fma 0x1.7fffff8p-968 0x1p-106 0x0.000001p-1022 missing-errno
fma 0x1.4000004p-967 0x1p-106 0x0.000001p-1022 missing-errno
fma 0x1.4p-967 -0x1p-106 -0x0.000001p-1022 missing-errno
fma -0x1.19cab66d73e17p-959 0x1.c7108a8c5ff51p-107 -0x0.80b0ad65d9b64p-1022 missing-errno
fma -0x1.d2eaed6e8e9d3p-979 -0x1.4e066c62ac9ddp-63 -0x0.9245e6b003454p-1022 missing-errno
fma 0x1.153d650bb9f06p-907 0x1.2d01230d48407p-125 -0x0.b278d5acfc3cp-1022 missing-errno
fma -0x1.fffffffffffffp-711 0x1.fffffffffffffp-275 0x1.fffffe00007ffp-983 missing-errno
fma 0x1.4p-1022 0x1.0000000000002p-1 0x1p-1024 missing-errno
fma -0x1.4p-1022 0x1.0000000000002p-1 -0x1p-1024 missing-errno
fma 0x1.ffffffffffffcp-1022 0x1.0000000000001p-1 0x1p-1074 missing-errno
fma -0x1.ffffffffffffcp-1022 0x1.0000000000001p-1 -0x1p-1074 missing-errno
fma 0x1p-1074 0x1p-1 0x0.fffffffffffffp-1022 missing-errno
fma -0x1p-1074 0x1p-1 -0x0.fffffffffffffp-1022 missing-errno
fma 0x1p-1074 0x1.1p-1 0x0.fffffffffffffp-1022 missing-errno
fma -0x1p-1074 0x1.1p-1 -0x0.fffffffffffffp-1022 missing-errno
fma 0x1p-1074 0x1p-1074 0x1p1023 missing-errno
fma 0x1p-1074 -0x1p-1074 0x1p1023 missing-errno
fma 0x1p-1074 0x1p-1074 -0x1p1023 missing-errno
fma 0x1p-1074 -0x1p-1074 -0x1p1023 missing-errno
fma 0x1p-1074 0x1p-1074 0x1p-1022 missing-errno
fma 0x1p-1074 -0x1p-1074 0x1p-1022 missing-errno
fma 0x1p-1074 0x1p-1074 -0x1p-1022 missing-errno
fma 0x1p-1074 -0x1p-1074 -0x1p-1022 missing-errno
fma 0x1p-1074 0x1p-1074 0x0.fffffffffffffp-1022 missing-errno
fma 0x1p-1074 -0x1p-1074 0x0.fffffffffffffp-1022 missing-errno
fma 0x1p-1074 0x1p-1074 -0x0.fffffffffffffp-1022 missing-errno
fma 0x1p-1074 -0x1p-1074 -0x0.fffffffffffffp-1022 missing-errno
fma 0x1p-1074 0x1p-1074 0x1p-1074 missing-errno
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x1p-1074 -0x1p-1074 0x1p-1074 missing-errno
fma 0x1p-1074 0x1p-1074 -0x1p-1074 missing-errno
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1p-1074 -0x1p-1074 -0x1p-1074 missing-errno
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x0.fffffffffffff8p0 0x0.fffffffffffff8p0 -0x0.fffffffffffffp0
fma 0x0.fffffffffffff8p0 -0x0.fffffffffffff8p0 0x0.fffffffffffffp0
fma -0x0.fffffffffffff8p0 0x0.fffffffffffff8p0 0x0.fffffffffffffp0
fma -0x0.fffffffffffff8p0 -0x0.fffffffffffff8p0 -0x0.fffffffffffffp0
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1.0000000000001p-1022 0x1.0000000000001p-55 0x1p1023 missing-errno
fma 0x1.0000000000001p-1022 -0x1.0000000000001p-55 0x1p1023 missing-errno
fma 0x1.0000000000001p-1022 0x1.0000000000001p-55 -0x1p1023 missing-errno
fma 0x1.0000000000001p-1022 -0x1.0000000000001p-55 -0x1p1023 missing-errno
fma 0x1.0000000000001p-1022 0x1.0000000000001p-55 0x1p970 missing-errno
fma 0x1.0000000000001p-1022 -0x1.0000000000001p-55 0x1p970 missing-errno
fma 0x1.0000000000001p-1022 0x1.0000000000001p-55 -0x1p970 missing-errno
fma 0x1.0000000000001p-1022 -0x1.0000000000001p-55 -0x1p970 missing-errno
fma -0x8.03fcp+3696 0xf.fffffffffffffffp-6140 0x8.3ffffffffffffffp-2450 missing-errno
fma 0x9.fcp+2033 -0x8.000e1f000ff800fp-3613 -0xf.fffffffffffc0ffp-1579 missing-errno
fma 0xc.7fc000003ffffffp-1194 0x8.1e0003fffffffffp+15327 -0x8.fffep+14072 missing-errno
fma -0x8.0001fc000000003p+1798 0xcp-2230 0x8.f7e000000000007p-468 missing-errno
fma 0xc.0000000000007ffp+10130 -0x8.000000000000001p+4430 0xc.07000000001ffffp+14513 missing-errno
fma 0xb.ffffp-4777 0x8.000000fffffffffp-11612 -0x0.3800fff8p-16385 missing-errno
fma 0x1.4p-16382 0x1.0000000000000004p-1 0x1p-16384 missing-errno
fma -0x1.4p-16382 0x1.0000000000000004p-1 -0x1p-16384 missing-errno
fma 0x1.fffffffffffffff8p-16382 0x1.0000000000000002p-1 0x1p-16445 missing-errno
fma -0x1.fffffffffffffff8p-16382 0x1.0000000000000002p-1 -0x1p-16445 missing-errno
fma 0x1p-16445 0x1p-1 0x0.fffffffffffffffep-16382 missing-errno
fma -0x1p-16445 0x1p-1 -0x0.fffffffffffffffep-16382 missing-errno
fma 0x1p-16445 0x1.1p-1 0x0.fffffffffffffffep-16382 missing-errno
fma -0x1p-16445 0x1.1p-1 -0x0.fffffffffffffffep-16382 missing-errno
fma 0x1p-16445 0x1p-16445 0x1p16383 missing-errno
fma 0x1p-16445 -0x1p-16445 0x1p16383 missing-errno
fma 0x1p-16445 0x1p-16445 -0x1p16383 missing-errno
fma 0x1p-16445 -0x1p-16445 -0x1p16383 missing-errno
fma 0x1p-16445 0x1p-16445 0x1p-16382 missing-errno
fma 0x1p-16445 -0x1p-16445 0x1p-16382 missing-errno
fma 0x1p-16445 0x1p-16445 -0x1p-16382 missing-errno
fma 0x1p-16445 -0x1p-16445 -0x1p-16382 missing-errno
fma 0x1p-16445 0x1p-16445 0x0.fffffffffffffffep-16382 missing-errno
fma 0x1p-16445 -0x1p-16445 0x0.fffffffffffffffep-16382 missing-errno
fma 0x1p-16445 0x1p-16445 -0x0.fffffffffffffffep-16382 missing-errno
fma 0x1p-16445 -0x1p-16445 -0x0.fffffffffffffffep-16382 missing-errno
fma 0x1p-16445 0x1p-16445 0x1p-16445 missing-errno
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x1p-16445 -0x1p-16445 0x1p-16445 missing-errno
fma 0x1p-16445 0x1p-16445 -0x1p-16445 missing-errno
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1p-16445 -0x1p-16445 -0x1p-16445 missing-errno
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x0.ffffffffffffffffp0 0x0.ffffffffffffffffp0 -0x0.fffffffffffffffep0
fma 0x0.ffffffffffffffffp0 -0x0.ffffffffffffffffp0 0x0.fffffffffffffffep0
fma -0x0.ffffffffffffffffp0 0x0.ffffffffffffffffp0 0x0.fffffffffffffffep0
fma -0x0.ffffffffffffffffp0 -0x0.ffffffffffffffffp0 -0x0.fffffffffffffffep0
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1.0000000000000002p-16382 0x1.0000000000000002p-66 0x1p16383 missing-errno
fma 0x1.0000000000000002p-16382 -0x1.0000000000000002p-66 0x1p16383 missing-errno
fma 0x1.0000000000000002p-16382 0x1.0000000000000002p-66 -0x1p16383 missing-errno
fma 0x1.0000000000000002p-16382 -0x1.0000000000000002p-66 -0x1p16383 missing-errno
fma 0x1.0000000000000002p-16382 0x1.0000000000000002p-66 0x1p16319 missing-errno
fma 0x1.0000000000000002p-16382 -0x1.0000000000000002p-66 0x1p16319 missing-errno
fma 0x1.0000000000000002p-16382 0x1.0000000000000002p-66 -0x1p16319 missing-errno
fma 0x1.0000000000000002p-16382 -0x1.0000000000000002p-66 -0x1p16319 missing-errno
fma 0x1.bb2de33e02ccbbfa6e245a7c1f71p-2584 -0x1.6b500daf0580d987f1bc0cadfcddp-13777 0x1.613cd91d9fed34b33820e5ab9d8dp-16378 missing-errno
fma -0x1.f949b880cacb0f0c61540105321dp-5954 -0x1.3876cec84b4140f3bd6198731b7ep-10525 -0x0.a5dc1c6cfbc498c54fb0b504bf19p-16382 missing-errno
fma -0x1.0000fffffffffp-16221 0x1.0000001fffff8007fep-239 0x0.ff87ffffffffffffe000003fffffp-16382 missing-errno
fma -0x1.ac79c9376ef447f3827c9e9de008p-2228 -0x1.5ba830022b6139e21fbe7270cad8p-6314 0x1.e8282b6a26bb6a9daf5c8e73e9f9p-8616 missing-errno
fma -0x1.c69749ec574caaa2ab8e97ddb9f3p+2652 0x1.f34235ff9d095449c29b4831b62dp+3311 0x1.fbe4302df23354dbd0c4d3cfe606p+5879 missing-errno
fma -0x1.ca8835fc6ecfb5398625fc891be5p-1686 0x1.621e1972bbe2180e5be9dd7d8df5p-7671 -0x1.7d2d21b73b52cf20dec2a83902a4p-9395 missing-errno
fma -0x1.55cff679ec49c2541fab41fc843ep-11819 0x1.e60e9f464f9e8df0509647c7c971p+12325 0x1.eaa2a7649d765c2f564f7a5beca7p+454 missing-errno
fma 0x1.f0e7b1454908576f2537d863cf9bp+11432 0x1.cdce52f09d4ca76e68706f34b5d5p-1417 -0x1.2e986187c70f146235ea2066e486p+9979 missing-errno
fma 0x1.f102f7da4a57a3a4aab620e29452p-3098 -0x1.cc06a4ff40248f9e2dcc4b6afd84p-11727 0x1.d512a11126b5ac8ed8973b8580c8p-14849 missing-errno
fma -0x1.fc47ac7434b993cd8dcb2b431f25p-3816 0x1.fbc9750da8468852d84558e1db6dp-5773 -0x1.00a98abf783f75c40fe5b7a37d86p-9607 missing-errno
fma 0x1.00000000000007ffffffffffffffp-9045 -0x1.ffffffffffff80000001ffffffffp+4773 -0x1.f8p-4316 missing-errno
fma 0x1.4e922764c90701d4a2f21d01893dp-8683 -0x1.955a12e2d7c9447c27fa022fc865p+212 -0x1.e9634462eaef96528b90b6944578p-8521 missing-errno
fma 0x1.801181509c03bdbef10d6165588cp-15131 0x1.ad86f8e57d3d40bfa8007780af63p-368 -0x1.6e9df0dab1c9f1d7a6043c390741p-15507 missing-errno
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x1.ffffffffffffffp0 0x1.000000000000008p0 -0x1p-1000
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1.4p-16382 0x1.0000000000000000000000000002p-1 0x1p-16384 missing-errno
fma -0x1.4p-16382 0x1.0000000000000000000000000002p-1 -0x1p-16384 missing-errno
fma 0x1.fffffffffffffffffffffffffffcp-16382 0x1.0000000000000000000000000001p-1 0x1p-16494 missing-errno
fma -0x1.fffffffffffffffffffffffffffcp-16382 0x1.0000000000000000000000000001p-1 -0x1p-16494 missing-errno
fma 0x1p-16494 0x1p-1 0x0.ffffffffffffffffffffffffffffp-16382 missing-errno
fma -0x1p-16494 0x1p-1 -0x0.ffffffffffffffffffffffffffffp-16382 missing-errno
fma 0x1p-16494 0x1.1p-1 0x0.ffffffffffffffffffffffffffffp-16382 missing-errno
fma -0x1p-16494 0x1.1p-1 -0x0.ffffffffffffffffffffffffffffp-16382 missing-errno
fma 0x1p-16494 0x1p-16494 0x1p16383 missing-errno
fma 0x1p-16494 -0x1p-16494 0x1p16383 missing-errno
fma 0x1p-16494 0x1p-16494 -0x1p16383 missing-errno
fma 0x1p-16494 -0x1p-16494 -0x1p16383 missing-errno
fma 0x1p-16494 0x1p-16494 0x1p-16382 missing-errno
fma 0x1p-16494 -0x1p-16494 0x1p-16382 missing-errno
fma 0x1p-16494 0x1p-16494 -0x1p-16382 missing-errno
fma 0x1p-16494 -0x1p-16494 -0x1p-16382 missing-errno
fma 0x1p-16494 0x1p-16494 0x0.ffffffffffffffffffffffffffffp-16382 missing-errno
fma 0x1p-16494 -0x1p-16494 0x0.ffffffffffffffffffffffffffffp-16382 missing-errno
fma 0x1p-16494 0x1p-16494 -0x0.ffffffffffffffffffffffffffffp-16382 missing-errno
fma 0x1p-16494 -0x1p-16494 -0x0.ffffffffffffffffffffffffffffp-16382 missing-errno
fma 0x1p-16494 0x1p-16494 0x1p-16494 missing-errno
Move tests of fma from libm-test.inc to auto-libm-test-in. This patch moves tests of fma to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Because fma can have exact zero results depending on the rounding mode, results of fma cannot always be determined from a single value computed in higher precision with a sticky bit. Thus, this patch adds support for recomputing results with the original MPFR/MPC function in the case where an exact zero is involved. (This also affects some results for cpow; when we start testing cpow in all rounding modes, I think it will be most appropriate to make those tests use IGNORE_ZERO_INF_SIGN, since ISO C does not attempt to determine signs of zero results, or special caes in general, for cpow, and I think signs of zero for cpow are beyond the scope of glibc's accuracy goals.) Simply treating the existing test inputs for fma like those for other functions (i.e., as representing the given value rounded up or down to any of the supported floating-point formats) increases the size of auto-libm-test-out by about 16MB (i.e., about half the file is fma test data). While rounded versions of tests are perfectly reasonable test inputs for fma, in this case having them seems excessive, so this patch allows functions to specify in gen-auto-libm-tests that the given test inputs are only to be interpreted exactly, not as corresponding to values rounded up and down. This reduces the size of the generated test data for fma to a more reasonable 2MB. A consequence of this patch is that fma is now tested for correct presence or absence of "inexact" exceptions, where previously this wasn't tested because I didn't want to try to add that test coverage manually to all the existing tests. As far as I know, the existing fma implementations are already correct in this regard. This patch provides the first cases where the gen-auto-libm-tests support for distinguishing before-rounding/after-rounding underflow actually produces separate entries in auto-libm-test-out (for functions without exactly determined results, the affected cases are all considered underflow-optional, so this only affects functions like fma with exactly determined results). I didn't see any signs of problems with this logic in the output. Tested x86_64 and x86. * math/auto-libm-test-in: Add tests of fma. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fma_test_data): Use AUTO_TESTS_fff_f. (fma_towardzero_test_data): Likewise. (fma_downward_test_data): Likewise. (fma_upward_test_data): Likewise. * math/gen-auto-libm-tests.c (rounding_mode_desc): Add field mpc_mode. (rounding_modes): Add values for new field. (func_calc_method): Add value mpfr_fff_f. (func_calc_desc): Add mpfr_fff_f union field. (test_function): Add field exact_args. (FUNC): Add macro argument EXACT_ARGS. (FUNC_mpfr_f_f): Update call to FUNC. (FUNC_mpfr_f_f): Likewise. (FUNC_mpfr_ff_f): Likewise. (FUNC_mpfr_if_f): Likewise. (FUNC_mpc_c_f): Likewise. (FUNC_mpc_c_c): Likewise. (test_functions): Add fma. Update calls to FUNC. (handle_input_arg): Add argument exact_args. (add_test): Update call to handle_input_arg. (calc_generic_results): Add argument mode. Handle mpfr_fff_f. (output_for_one_input_case): Update call to calc_generic_results. Recalculate exact zero results in each rounding mode.
2014-02-18 21:48:51 +00:00
fma 0x1p-16494 -0x1p-16494 0x1p-16494 missing-errno
fma 0x1p-16494 0x1p-16494 -0x1p-16494 missing-errno
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1p-16494 -0x1p-16494 -0x1p-16494 missing-errno
fma 0x0.ffffffffffffffffffffffffffff8p0 0x0.ffffffffffffffffffffffffffff8p0 -0x0.ffffffffffffffffffffffffffffp0 missing-errno
fma 0x0.ffffffffffffffffffffffffffff8p0 -0x0.ffffffffffffffffffffffffffff8p0 0x0.ffffffffffffffffffffffffffffp0 missing-errno
fma -0x0.ffffffffffffffffffffffffffff8p0 0x0.ffffffffffffffffffffffffffff8p0 0x0.ffffffffffffffffffffffffffffp0 missing-errno
fma -0x0.ffffffffffffffffffffffffffff8p0 -0x0.ffffffffffffffffffffffffffff8p0 -0x0.ffffffffffffffffffffffffffffp0 missing-errno
fma 0x1.0000000000000000000000000001p-16382 0x1.0000000000000000000000000001p-66 0x1p16383 missing-errno
fma 0x1.0000000000000000000000000001p-16382 -0x1.0000000000000000000000000001p-66 0x1p16383 missing-errno
fma 0x1.0000000000000000000000000001p-16382 0x1.0000000000000000000000000001p-66 -0x1p16383 missing-errno
fma 0x1.0000000000000000000000000001p-16382 -0x1.0000000000000000000000000001p-66 -0x1p16383 missing-errno
fma 0x1.0000000000000000000000000001p-16382 0x1.0000000000000000000000000001p-66 0x1p16319 missing-errno
fma 0x1.0000000000000000000000000001p-16382 -0x1.0000000000000000000000000001p-66 0x1p16319 missing-errno
fma 0x1.0000000000000000000000000001p-16382 0x1.0000000000000000000000000001p-66 -0x1p16319 missing-errno
fma 0x1.0000000000000000000000000001p-16382 -0x1.0000000000000000000000000001p-66 -0x1p16319 missing-errno
fma 0x1.fffffep-126 0x1.fffffep25 0x1.fffffep127 missing-errno xfail-rounding:ibm128-libgcc
Fix fma spurious underflows (bug 18824). Various fma implementations have logic that, when computing fma (x, y, z) where z is large (so care needs taking to avoid internal overflow) but x * y is small, scale x * y up instead of down to avoid internal underflows resulting from scaling down. (In these cases, x * y is small enough that only its sign actually matters rather than the exact value.) The threshold for scaling up instead of down was correct for "if the unscaled values were multiplied, the low part of the multiplication could underflow", and the scaling was sufficient to ensure that the low part of the multiplication did not underflow (given that cases of very small x * y - less than half the least subnormal - were previously dealt with). However, the choice in the functions wasn't between scaling up or no scaling, but between scaling up and scaling down (scaling down actually being needed when x * y isn't so small compared to z and so the exact value does matter). Thus a larger threshold is needed to ensure that scaling down doesn't produce values the multiplication of whose low parts underflows. This patch increases the thresholds accordingly. Tested for x86_64, x86 and mips64 (with the MIPS version of s_fmal.c removed so that the ldbl-128 version gets tested instead of the soft-fp one). [BZ #18824] * sysdeps/ieee754/dbl-64/s_fma.c (__fma): Increase threshold for scaling x * y up instead of down. * sysdeps/ieee754/ldbl-128/s_fmal.c (__fmal): Likewise. * sysdeps/ieee754/ldbl-96/s_fmal.c (__fmal): Likewise. * math/auto-libm-test-in: Add more tests of fma. * math/auto-libm-test-out: Regenerated.
2015-08-14 17:15:06 +00:00
fma 0x1.fffffep-126 -0x1.fffffep25 0x1.fffffep127
fma 0x1.fffffep-126 0x1.fffffep25 -0x1.fffffep127
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1.fffffep-126 -0x1.fffffep25 -0x1.fffffep127 missing-errno xfail-rounding:ibm128-libgcc
fma 0x1.fffffffffffffp-1022 0x1.fffffffffffffp54 0x1.fffffffffffffp1023 missing-errno xfail-rounding:ibm128-libgcc
fma 0x1.fffffffffffffp-1022 -0x1.fffffffffffffp54 0x1.fffffffffffffp1023 missing-errno
fma 0x1.fffffffffffffp-1022 0x1.fffffffffffffp54 -0x1.fffffffffffffp1023 missing-errno
fma 0x1.fffffffffffffp-1022 -0x1.fffffffffffffp54 -0x1.fffffffffffffp1023 missing-errno xfail-rounding:ibm128-libgcc
Fix fma spurious underflows (bug 18824). Various fma implementations have logic that, when computing fma (x, y, z) where z is large (so care needs taking to avoid internal overflow) but x * y is small, scale x * y up instead of down to avoid internal underflows resulting from scaling down. (In these cases, x * y is small enough that only its sign actually matters rather than the exact value.) The threshold for scaling up instead of down was correct for "if the unscaled values were multiplied, the low part of the multiplication could underflow", and the scaling was sufficient to ensure that the low part of the multiplication did not underflow (given that cases of very small x * y - less than half the least subnormal - were previously dealt with). However, the choice in the functions wasn't between scaling up or no scaling, but between scaling up and scaling down (scaling down actually being needed when x * y isn't so small compared to z and so the exact value does matter). Thus a larger threshold is needed to ensure that scaling down doesn't produce values the multiplication of whose low parts underflows. This patch increases the thresholds accordingly. Tested for x86_64, x86 and mips64 (with the MIPS version of s_fmal.c removed so that the ldbl-128 version gets tested instead of the soft-fp one). [BZ #18824] * sysdeps/ieee754/dbl-64/s_fma.c (__fma): Increase threshold for scaling x * y up instead of down. * sysdeps/ieee754/ldbl-128/s_fmal.c (__fmal): Likewise. * sysdeps/ieee754/ldbl-96/s_fmal.c (__fmal): Likewise. * math/auto-libm-test-in: Add more tests of fma. * math/auto-libm-test-out: Regenerated.
2015-08-14 17:15:06 +00:00
fma 0x1.fffffffffffffffep-16382 0x1.fffffffffffffffep65 0x1.fffffffffffffffep16383 missing-errno
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1.fffffffffffffffep-16382 -0x1.fffffffffffffffep65 0x1.fffffffffffffffep16383 missing-errno
fma 0x1.fffffffffffffffep-16382 0x1.fffffffffffffffep65 -0x1.fffffffffffffffep16383 missing-errno
Fix fma spurious underflows (bug 18824). Various fma implementations have logic that, when computing fma (x, y, z) where z is large (so care needs taking to avoid internal overflow) but x * y is small, scale x * y up instead of down to avoid internal underflows resulting from scaling down. (In these cases, x * y is small enough that only its sign actually matters rather than the exact value.) The threshold for scaling up instead of down was correct for "if the unscaled values were multiplied, the low part of the multiplication could underflow", and the scaling was sufficient to ensure that the low part of the multiplication did not underflow (given that cases of very small x * y - less than half the least subnormal - were previously dealt with). However, the choice in the functions wasn't between scaling up or no scaling, but between scaling up and scaling down (scaling down actually being needed when x * y isn't so small compared to z and so the exact value does matter). Thus a larger threshold is needed to ensure that scaling down doesn't produce values the multiplication of whose low parts underflows. This patch increases the thresholds accordingly. Tested for x86_64, x86 and mips64 (with the MIPS version of s_fmal.c removed so that the ldbl-128 version gets tested instead of the soft-fp one). [BZ #18824] * sysdeps/ieee754/dbl-64/s_fma.c (__fma): Increase threshold for scaling x * y up instead of down. * sysdeps/ieee754/ldbl-128/s_fmal.c (__fmal): Likewise. * sysdeps/ieee754/ldbl-96/s_fmal.c (__fmal): Likewise. * math/auto-libm-test-in: Add more tests of fma. * math/auto-libm-test-out: Regenerated.
2015-08-14 17:15:06 +00:00
fma 0x1.fffffffffffffffep-16382 -0x1.fffffffffffffffep65 -0x1.fffffffffffffffep16383 missing-errno
fma 0x1.ffffffffffffffffffffffffffffp-16382 0x1.ffffffffffffffffffffffffffffp114 0x1.ffffffffffffffffffffffffffffp16383 missing-errno
Add narrowing fma functions This patch adds the narrowing fused multiply-add functions from TS 18661-1 / TS 18661-3 / C2X to glibc's libm: ffma, ffmal, dfmal, f32fmaf64, f32fmaf32x, f32xfmaf64 for all configurations; f32fmaf64x, f32fmaf128, f64fmaf64x, f64fmaf128, f32xfmaf64x, f32xfmaf128, f64xfmaf128 for configurations with _Float64x and _Float128; __f32fmaieee128 and __f64fmaieee128 aliases in the powerpc64le case (for calls to ffmal and dfmal when long double is IEEE binary128). Corresponding tgmath.h macro support is also added. The changes are mostly similar to those for the other narrowing functions previously added, especially that for sqrt, so the description of those generally applies to this patch as well. As with sqrt, I reused the same test inputs in auto-libm-test-in as for non-narrowing fma rather than adding extra or separate inputs for narrowing fma. The tests in libm-test-narrow-fma.inc also follow those for non-narrowing fma. The non-narrowing fma has a known bug (bug 6801) that it does not set errno on errors (overflow, underflow, Inf * 0, Inf - Inf). Rather than fixing this or having narrowing fma check for errors when non-narrowing does not (complicating the cases when narrowing fma can otherwise be an alias for a non-narrowing function), this patch does not attempt to check for errors from narrowing fma and set errno; the CHECK_NARROW_FMA macro is still present, but as a placeholder that does nothing, and this missing errno setting is considered to be covered by the existing bug rather than needing a separate open bug. missing-errno annotations are duly added to many of the auto-libm-test-in test inputs for fma. This completes adding all the new functions from TS 18661-1 to glibc, so will be followed by corresponding stdc-predef.h changes to define __STDC_IEC_60559_BFP__ and __STDC_IEC_60559_COMPLEX__, as the support for TS 18661-1 will be at a similar level to that for C standard floating-point facilities up to C11 (pragmas not implemented, but library functions done). (There are still further changes to be done to implement changes to the types of fromfp functions from N2548.) Tested as followed: natively with the full glibc testsuite for x86_64 (GCC 11, 7, 6) and x86 (GCC 11); with build-many-glibcs.py with GCC 11, 7 and 6; cross testing of math/ tests for powerpc64le, powerpc32 hard float, mips64 (all three ABIs, both hard and soft float). The different GCC versions are to cover the different cases in tgmath.h and tgmath.h tests properly (GCC 6 has _Float* only as typedefs in glibc headers, GCC 7 has proper _Float* support, GCC 8 adds __builtin_tgmath).
2021-09-22 21:25:31 +00:00
fma 0x1.ffffffffffffffffffffffffffffp-16382 -0x1.ffffffffffffffffffffffffffffp114 0x1.ffffffffffffffffffffffffffffp16383 missing-errno
fma 0x1.ffffffffffffffffffffffffffffp-16382 0x1.ffffffffffffffffffffffffffffp114 -0x1.ffffffffffffffffffffffffffffp16383 missing-errno
Fix fma spurious underflows (bug 18824). Various fma implementations have logic that, when computing fma (x, y, z) where z is large (so care needs taking to avoid internal overflow) but x * y is small, scale x * y up instead of down to avoid internal underflows resulting from scaling down. (In these cases, x * y is small enough that only its sign actually matters rather than the exact value.) The threshold for scaling up instead of down was correct for "if the unscaled values were multiplied, the low part of the multiplication could underflow", and the scaling was sufficient to ensure that the low part of the multiplication did not underflow (given that cases of very small x * y - less than half the least subnormal - were previously dealt with). However, the choice in the functions wasn't between scaling up or no scaling, but between scaling up and scaling down (scaling down actually being needed when x * y isn't so small compared to z and so the exact value does matter). Thus a larger threshold is needed to ensure that scaling down doesn't produce values the multiplication of whose low parts underflows. This patch increases the thresholds accordingly. Tested for x86_64, x86 and mips64 (with the MIPS version of s_fmal.c removed so that the ldbl-128 version gets tested instead of the soft-fp one). [BZ #18824] * sysdeps/ieee754/dbl-64/s_fma.c (__fma): Increase threshold for scaling x * y up instead of down. * sysdeps/ieee754/ldbl-128/s_fmal.c (__fmal): Likewise. * sysdeps/ieee754/ldbl-96/s_fmal.c (__fmal): Likewise. * math/auto-libm-test-in: Add more tests of fma. * math/auto-libm-test-out: Regenerated.
2015-08-14 17:15:06 +00:00
fma 0x1.ffffffffffffffffffffffffffffp-16382 -0x1.ffffffffffffffffffffffffffffp114 -0x1.ffffffffffffffffffffffffffffp16383 missing-errno
Fix hypot handling of subnormals (bug 16316, bug 16330).
2013-12-17 13:42:13 +00:00
hypot 0 0
hypot 0 -0
hypot -0 0
hypot -0 -0
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
# hypot (x,y) == hypot (+-x, +-y).
hypot 0.7 12.4
hypot -0.7 12.4
hypot 0.7 -12.4
hypot -0.7 -12.4
hypot 12.4 0.7
hypot -12.4 0.7
hypot 12.4 -0.7
hypot -12.4 -0.7
# hypot (x,0) == fabs (x).
hypot 0.75 0
hypot -0.75 0
hypot -5.7e7 0
hypot 0.75 1.25
hypot 1.0 0x1p-61
hypot 0x1p+0 0x1.fp-129
Fix dbl-64 hypot spurious underflows (bug 16314).
2013-12-17 13:43:40 +00:00
hypot 0x1.23456789abcdef0123456789ab8p-500 0x1.23456789abcdef0123456789ab8p-500
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
hypot 0x3p125 0x4p125
hypot 0x1.234566p-126 0x1.234566p-126
hypot 0x3p1021 0x4p1021
hypot 0x1p+0 0x0.3ep-1022
hypot 0x3p16381 0x4p16381
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
hypot 0x1p-149 0x1p-149
hypot 0x1p-1074 0x1p-1074
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
hypot 0x1p-16445 0x1p-16445
hypot 0x1p-16494 0x1p-16494
Fix hypot handling of subnormals (bug 16316, bug 16330).
2013-12-17 13:42:13 +00:00
hypot 0x0.fffffep-126 0x0.fp-127
hypot 0x0.fffffep-126 0x0.fp-130
hypot 0x0.fffffffffffffp-1022 0x0.fp-1023
hypot 0x0.fffffffffffffp-1022 0x0.fp-1026
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
hypot 0x0.ffffffp-16382 0x0.fp-16383
hypot 0x0.ffffffp-16382 0x0.fp-16386
hypot 0 min
hypot 0 min_subnorm
hypot 0 -min
hypot 0 -min_subnorm
hypot min 0
hypot min_subnorm 0
hypot -min 0
hypot -min_subnorm 0
hypot min min
hypot min_subnorm min_subnorm
hypot min min_subnorm
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
hypot 0x1.fp127 0x1.fp127
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
hypot 0x1.fp1023 0x1.fp1023
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
hypot 0x1.fp16383 0x1.fp16383 xfail-rounding:ibm128-libgcc
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
Fix hypot missing underflows (bug 18803). Similar to various other bugs in this area, hypot functions can fail to raise the underflow exception when the result is tiny and inexact but one or more low bits of the intermediate result that is scaled down (or, in the i386 case, converted from a wider evaluation format) are zero. This patch forces the exception in a similar way to previous fixes. Note that this issue cannot arise for implementations of hypotf using double (or wider) for intermediate evaluation (if hypotf should underflow, that means the double square root is being computed of some number of the form N*2^-298, for 0 < N < 2^46, which is exactly represented as a double, and whatever the rounding mode such a square root cannot have a mantissa with all zeroes after the initial 23 bits). Thus no changes are made to hypotf implementations in this patch, only to hypot and hypotl. Tested for x86_64, x86, mips64 and powerpc. [BZ #18803] * sysdeps/i386/fpu/e_hypot.S: Use DEFINE_DBL_MIN. (MO): New macro. (__ieee754_hypot) [PIC]: Load PIC register. (__ieee754_hypot): Use DBL_NARROW_EVAL_UFLOW_NONNEG instead of DBL_NARROW_EVAL. * sysdeps/ieee754/dbl-64/e_hypot.c (__ieee754_hypot): Use math_check_force_underflow_nonneg in case where result might be tiny. * sysdeps/ieee754/ldbl-128/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/ieee754/ldbl-96/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/powerpc/fpu/e_hypot.c (__ieee754_hypot): Likewise. * math/auto-libm-test-in: Add more tests of hypot. * math/auto-libm-test-out: Regenerated.
2015-09-24 23:43:57 +00:00
hypot 0x1p-127 0x1p-149
hypot 0x1p-1023 0x1p-1074
hypot 0x1p-970 0x1p-1074
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
hypot 0x1p-16383 0x1p-16445
hypot 0x1p-16384 0x1p-16446
hypot 0x1p-16383 0x1p-16494
Fix hypot missing underflows (bug 18803). Similar to various other bugs in this area, hypot functions can fail to raise the underflow exception when the result is tiny and inexact but one or more low bits of the intermediate result that is scaled down (or, in the i386 case, converted from a wider evaluation format) are zero. This patch forces the exception in a similar way to previous fixes. Note that this issue cannot arise for implementations of hypotf using double (or wider) for intermediate evaluation (if hypotf should underflow, that means the double square root is being computed of some number of the form N*2^-298, for 0 < N < 2^46, which is exactly represented as a double, and whatever the rounding mode such a square root cannot have a mantissa with all zeroes after the initial 23 bits). Thus no changes are made to hypotf implementations in this patch, only to hypot and hypotl. Tested for x86_64, x86, mips64 and powerpc. [BZ #18803] * sysdeps/i386/fpu/e_hypot.S: Use DEFINE_DBL_MIN. (MO): New macro. (__ieee754_hypot) [PIC]: Load PIC register. (__ieee754_hypot): Use DBL_NARROW_EVAL_UFLOW_NONNEG instead of DBL_NARROW_EVAL. * sysdeps/ieee754/dbl-64/e_hypot.c (__ieee754_hypot): Use math_check_force_underflow_nonneg in case where result might be tiny. * sysdeps/ieee754/ldbl-128/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/ieee754/ldbl-96/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/powerpc/fpu/e_hypot.c (__ieee754_hypot): Likewise. * math/auto-libm-test-in: Add more tests of hypot. * math/auto-libm-test-out: Regenerated.
2015-09-24 23:43:57 +00:00
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
hypot -0x1.fa7deap+0 0x1.a761bab383ac8p+0
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
j0 -1.0
j0 0.0
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
j0 -0
j0 min
j0 -min
j0 min_subnorm
j0 -min_subnorm
j0 0x1p-5
j0 0x1p-10
j0 0x1p-15
j0 0x1p-20
j0 0x1p-25
j0 0x1p-30
j0 0x1p-35
j0 0x1p-40
j0 0x1p-45
j0 0x1p-50
j0 0x1p-55
j0 0x1p-60
j0 0x1p-100
j0 0x1p-600
j0 0x1p-10000
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
j0 0.125
j0 0.75
j0 1.0
j0 1.5
j0 2.0
j0 8.0
j0 10.0
j0 4.0
j0 -4.0
j0 0x1.d7ce3ap+107
j0 -0x1.001000001p+593
j0 0x1p1023
j0 0x1p16382
j0 0x1p16383
Fix the inaccuracy of j0f/j1f/y0f/y1f [BZ #14469, #14470, #14471, #14472] For j0f/j1f/y0f/y1f, the largest error for all binary32 inputs is reduced to at most 9 ulps for all rounding modes. The new code is enabled only when there is a cancellation at the very end of the j0f/j1f/y0f/y1f computation, or for very large inputs, thus should not give any visible slowdown on average. Two different algorithms are used: * around the first 64 zeros of j0/j1/y0/y1, approximation polynomials of degree 3 are used, computed using the Sollya tool (https://www.sollya.org/) * for large inputs, an asymptotic formula from [1] is used [1] Fast and Accurate Bessel Function Computation, John Harrison, Proceedings of Arith 19, 2009. Inputs yielding the new largest errors are added to auto-libm-test-in, and ulps are regenerated for various targets (thanks Adhemerval Zanella). Tested on x86_64 with --disable-multi-arch and on powerpc64le-linux-gnu. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-01 06:14:10 +00:00
# the next values yield large errors for binary32
# (cf BZ #27670 for the xfail entry)
powerpc: Update ULPs and xfail more ibm128 outputs There are 2 new input values that require to be marked as xfail-rounding:ibm128-libgcc as they're known to fail because of libgcc issues with different rounding modes. Otherwise, the other tests just need an increase in ULP.
2020-04-07 14:41:29 +00:00
j0 0x2.602774p+0 xfail-rounding:ibm128-libgcc
Fix the inaccuracy of j0f/j1f/y0f/y1f [BZ #14469, #14470, #14471, #14472] For j0f/j1f/y0f/y1f, the largest error for all binary32 inputs is reduced to at most 9 ulps for all rounding modes. The new code is enabled only when there is a cancellation at the very end of the j0f/j1f/y0f/y1f computation, or for very large inputs, thus should not give any visible slowdown on average. Two different algorithms are used: * around the first 64 zeros of j0/j1/y0/y1, approximation polynomials of degree 3 are used, computed using the Sollya tool (https://www.sollya.org/) * for large inputs, an asymptotic formula from [1] is used [1] Fast and Accurate Bessel Function Computation, John Harrison, Proceedings of Arith 19, 2009. Inputs yielding the new largest errors are added to auto-libm-test-in, and ulps are regenerated for various targets (thanks Adhemerval Zanella). Tested on x86_64 with --disable-multi-arch and on powerpc64le-linux-gnu. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-01 06:14:10 +00:00
j0 0x1.04c39cp+6
j0 0x1.4b7066p+7
math: Fix inaccuracy of j0f for x >= 2^127 when sin(x)+cos(x) is tiny Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-08-07 19:14:53 +00:00
# the next value exercises the flt-32 code path for x >= 2^127
j0 0x8.2f4ecp+124
Fixed inaccuracy of j0f (BZ #28185) The largest errors over the full binary32 range are after this patch (on x86_64): RNDN: libm wrong by up to 9.00e+00 ulp(s) [9] for x=0x1.04c39cp+6 RNDZ: libm wrong by up to 9.00e+00 ulp(s) [9] for x=0x1.04c39cp+6 RNDU: libm wrong by up to 9.00e+00 ulp(s) [9] for x=0x1.04c39cp+6 RNDD: libm wrong by up to 8.98e+00 ulp(s) [9] for x=0x1.4b7066p+7 Inputs that were yielding huge errors have been added to "make check". Reviewed-by: Adhemeral Zanella <adhemerval.zanella@linaro.org>
2021-10-05 08:32:36 +00:00
# the next values gave an error > 9 ulps with glibc-2.34 (flt-32)
math: Also xfail the new j0f tests for ibm128-libgcc From commit 6bbf7298323bf31b. Checked on powerpc64-linux-gnu.
2021-10-06 13:47:52 +00:00
j0 0x1.31ec02p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.31e716p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.31f51ep+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.320226p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.31ed76p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32083cp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32614ep+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32dc02p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32e304p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32e00cp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32e504p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32e12p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32dc22p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32dc24p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32eb18p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.32fa0cp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.335806p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.335a0cp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.336p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.336c1cp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.336336p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33712ep+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.339506p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33950cp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.339d06p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33b402p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33b50ap+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33c4p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33c202p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33c702p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33c50ap+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33c312p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33c80ap+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33cb04p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33cfp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d1p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d102p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d10cp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d116p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d12cp+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d138p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d146p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d15p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
j0 0x1.33d152p+1 xfail:binary64 xfail:intel96 xfail:binary128 xfail:ibm128-libgcc
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
j1 -1.0
j1 0.0
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
j1 -0
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
j1 0.125
j1 0.75
j1 1.0
j1 1.5
j1 2.0
j1 8.0
j1 10.0
j1 0x1.3ffp+74
j1 0x1.ff00000000002p+840
j1 0x1p1023
j1 0x1p16382
j1 0x1p16383
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
j1 0x1p-5
j1 0x1p-10
j1 0x1p-15
j1 0x1p-20
j1 0x1p-25
j1 0x1p-30
j1 0x1p-35
j1 0x1p-40
j1 0x1p-45
j1 0x1p-50
j1 0x1p-55
j1 0x1p-60
Fix ldbl-128 j1l spurious underflows (bug 18612). The ldbl-128 implementation of j1l produces spurious underflow exceptions for some small arguments, as a result of squaring the argument. This patch fixes it just to use a linear approximation for sufficiently small arguments, and then to force an underflow exception only in the cases where it is required. Tested for mips64. [BZ #18612] * sysdeps/ieee754/ldbl-128/e_j1l.c (__ieee754_j1l): For small arguments, just return 0.5 times the argument, with underflow forced as needed. * math/auto-libm-test-in: Add more tests of j1. * math/auto-libm-test-out: Regenerated.
2015-06-29 17:51:32 +00:00
j1 0x1p-100
j1 0x1p-600
j1 0x1p-10000
Fix the inaccuracy of j0f/j1f/y0f/y1f [BZ #14469, #14470, #14471, #14472] For j0f/j1f/y0f/y1f, the largest error for all binary32 inputs is reduced to at most 9 ulps for all rounding modes. The new code is enabled only when there is a cancellation at the very end of the j0f/j1f/y0f/y1f computation, or for very large inputs, thus should not give any visible slowdown on average. Two different algorithms are used: * around the first 64 zeros of j0/j1/y0/y1, approximation polynomials of degree 3 are used, computed using the Sollya tool (https://www.sollya.org/) * for large inputs, an asymptotic formula from [1] is used [1] Fast and Accurate Bessel Function Computation, John Harrison, Proceedings of Arith 19, 2009. Inputs yielding the new largest errors are added to auto-libm-test-in, and ulps are regenerated for various targets (thanks Adhemerval Zanella). Tested on x86_64 with --disable-multi-arch and on powerpc64le-linux-gnu. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-01 06:14:10 +00:00
# the next values yield large errors in the binary32 format
# (cf BZ #27670 for the xfail entries)
powerpc: Update ULPs and xfail more ibm128 outputs There are 2 new input values that require to be marked as xfail-rounding:ibm128-libgcc as they're known to fail because of libgcc issues with different rounding modes. Otherwise, the other tests just need an increase in ULP.
2020-04-07 14:41:29 +00:00
j1 0x3.ae4b2p+0 xfail-rounding:ibm128-libgcc
Fix the inaccuracy of j0f/j1f/y0f/y1f [BZ #14469, #14470, #14471, #14472] For j0f/j1f/y0f/y1f, the largest error for all binary32 inputs is reduced to at most 9 ulps for all rounding modes. The new code is enabled only when there is a cancellation at the very end of the j0f/j1f/y0f/y1f computation, or for very large inputs, thus should not give any visible slowdown on average. Two different algorithms are used: * around the first 64 zeros of j0/j1/y0/y1, approximation polynomials of degree 3 are used, computed using the Sollya tool (https://www.sollya.org/) * for large inputs, an asymptotic formula from [1] is used [1] Fast and Accurate Bessel Function Computation, John Harrison, Proceedings of Arith 19, 2009. Inputs yielding the new largest errors are added to auto-libm-test-in, and ulps are regenerated for various targets (thanks Adhemerval Zanella). Tested on x86_64 with --disable-multi-arch and on powerpc64le-linux-gnu. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-01 06:14:10 +00:00
j1 0x1.2f28eap+7 xfail-rounding:binary64 xfail-rounding:binary128 xfail-rounding:intel96 xfail-rounding:ibm128-libgcc
j1 0x1.a1d20ap+6 xfail-rounding:binary128 xfail-rounding:intel96 xfail-rounding:ibm128-libgcc
Fix j1, jn missing errno setting on underflow (bug 18611). j1 and jn can underflow for small arguments, but fail to set errno when underflowing to 0. This patch fixes them to set errno in that case. Tested for x86_64, x86, mips64 and powerpc. [BZ #18611] * sysdeps/ieee754/dbl-64/e_j1.c (__ieee754_j1): Set errno and avoid excess range and precision on underflow. * sysdeps/ieee754/dbl-64/e_jn.c (__ieee754_jn): Likewise. * sysdeps/ieee754/flt-32/e_j1f.c (__ieee754_j1f): Likewise. * sysdeps/ieee754/flt-32/e_jnf.c (__ieee754_jnf): Likewise. * sysdeps/ieee754/ldbl-128/e_j1l.c (__ieee754_j1l): Set errno on underflow. * sysdeps/ieee754/ldbl-128/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-96/e_j1l.c (__ieee754_j1l): Likewise. * sysdeps/ieee754/ldbl-96/e_jnl.c (__ieee754_jnl): Likewise. * math/auto-libm-test-in: Do not allow missing errno setting for tests of j1 and jn. * math/auto-libm-test-out: Regenerated.
2015-10-23 21:37:33 +00:00
j1 min
j1 -min
j1 min_subnorm
j1 -min_subnorm
Move TEST_f_f tests for [e-j]* functions from libm-test.inc to auto-libm-test-in.
2013-12-03 16:22:49 +00:00
Move tests of jn and yn from libm-test.inc to auto-libm-test-in. This patch moves tests of jn and yn to auto-libm-test-in, adding the required support for gen-auto-libm-tests (and adding a missing assertion there and fixing logic that was broken for functions with integer arguments). Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of jn and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (jn_test_data): Use AUTO_TESTS_if_f. (yn_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_if_f. (func_calc_desc): Add mpfr_if_f union field. (FUNC_mpfr_if_f): New macro. (test_functions): Add jn and yn. (calc_generic_results): Assert type of second input for mpfr_ff_f. Handle mpfr_if_f. (output_for_one_input_case): Disable all checking for arguments fitting floating-point types in case of an integer argument. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-18 17:59:29 +00:00
# jn (0, x) == j0 (x).
jn 0 -1.0
jn 0 0.0
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
jn 0 -0
jn 0 min
jn 0 -min
jn 0 min_subnorm
jn 0 -min_subnorm
Move tests of jn and yn from libm-test.inc to auto-libm-test-in. This patch moves tests of jn and yn to auto-libm-test-in, adding the required support for gen-auto-libm-tests (and adding a missing assertion there and fixing logic that was broken for functions with integer arguments). Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of jn and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (jn_test_data): Use AUTO_TESTS_if_f. (yn_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_if_f. (func_calc_desc): Add mpfr_if_f union field. (FUNC_mpfr_if_f): New macro. (test_functions): Add jn and yn. (calc_generic_results): Assert type of second input for mpfr_ff_f. Handle mpfr_if_f. (output_for_one_input_case): Disable all checking for arguments fitting floating-point types in case of an integer argument. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-18 17:59:29 +00:00
jn 0 0.125
jn 0 0.75
jn 0 1.0
jn 0 1.5
jn 0 2.0
jn 0 8.0
jn 0 10.0
jn 0 4.0
jn 0 -4.0
# jn (1, x) == j1 (x).
jn 1 -1.0
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
jn 1 -0
Move tests of jn and yn from libm-test.inc to auto-libm-test-in. This patch moves tests of jn and yn to auto-libm-test-in, adding the required support for gen-auto-libm-tests (and adding a missing assertion there and fixing logic that was broken for functions with integer arguments). Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of jn and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (jn_test_data): Use AUTO_TESTS_if_f. (yn_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_if_f. (func_calc_desc): Add mpfr_if_f union field. (FUNC_mpfr_if_f): New macro. (test_functions): Add jn and yn. (calc_generic_results): Assert type of second input for mpfr_ff_f. Handle mpfr_if_f. (output_for_one_input_case): Disable all checking for arguments fitting floating-point types in case of an integer argument. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-18 17:59:29 +00:00
jn 1 0.0
jn 1 0.125
jn 1 0.75
jn 1 1.0
jn 1 1.5
jn 1 2.0
jn 1 8.0
jn 1 10.0
Fix j1, jn missing errno setting on underflow (bug 18611). j1 and jn can underflow for small arguments, but fail to set errno when underflowing to 0. This patch fixes them to set errno in that case. Tested for x86_64, x86, mips64 and powerpc. [BZ #18611] * sysdeps/ieee754/dbl-64/e_j1.c (__ieee754_j1): Set errno and avoid excess range and precision on underflow. * sysdeps/ieee754/dbl-64/e_jn.c (__ieee754_jn): Likewise. * sysdeps/ieee754/flt-32/e_j1f.c (__ieee754_j1f): Likewise. * sysdeps/ieee754/flt-32/e_jnf.c (__ieee754_jnf): Likewise. * sysdeps/ieee754/ldbl-128/e_j1l.c (__ieee754_j1l): Set errno on underflow. * sysdeps/ieee754/ldbl-128/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-96/e_j1l.c (__ieee754_j1l): Likewise. * sysdeps/ieee754/ldbl-96/e_jnl.c (__ieee754_jnl): Likewise. * math/auto-libm-test-in: Do not allow missing errno setting for tests of j1 and jn. * math/auto-libm-test-out: Regenerated.
2015-10-23 21:37:33 +00:00
jn 1 min
jn 1 -min
jn 1 min_subnorm
jn 1 -min_subnorm
Move tests of jn and yn from libm-test.inc to auto-libm-test-in. This patch moves tests of jn and yn to auto-libm-test-in, adding the required support for gen-auto-libm-tests (and adding a missing assertion there and fixing logic that was broken for functions with integer arguments). Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of jn and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (jn_test_data): Use AUTO_TESTS_if_f. (yn_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_if_f. (func_calc_desc): Add mpfr_if_f union field. (FUNC_mpfr_if_f): New macro. (test_functions): Add jn and yn. (calc_generic_results): Assert type of second input for mpfr_ff_f. Handle mpfr_if_f. (output_for_one_input_case): Disable all checking for arguments fitting floating-point types in case of an integer argument. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-18 17:59:29 +00:00
jn 3 -1.0
jn 3 0.0
jn 3 0.125
jn 3 0.75
jn 3 1.0
jn 3 2.0
jn 3 10.0
jn 10 -1.0
jn 10 0.0
jn 10 0.125
jn 10 0.75
jn 10 1.0
jn 10 2.0
jn 10 10.0
jn 2 2.4048255576957729
jn 3 2.4048255576957729
jn 4 2.4048255576957729
jn 5 2.4048255576957729
jn 6 2.4048255576957729
jn 7 2.4048255576957729
jn 8 2.4048255576957729
jn 9 2.4048255576957729
jn 2 0x1.ffff62p+99
jn 2 0x1p127
jn 2 0x1p1023
jn 2 0x1p16383
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
jn -1 1
jn -2 1
jn -3 1
jn -4 1
jn -1 -1
jn -2 -1
jn -3 -1
jn -4 -1
Fix j1, jn missing errno setting on underflow (bug 18611). j1 and jn can underflow for small arguments, but fail to set errno when underflowing to 0. This patch fixes them to set errno in that case. Tested for x86_64, x86, mips64 and powerpc. [BZ #18611] * sysdeps/ieee754/dbl-64/e_j1.c (__ieee754_j1): Set errno and avoid excess range and precision on underflow. * sysdeps/ieee754/dbl-64/e_jn.c (__ieee754_jn): Likewise. * sysdeps/ieee754/flt-32/e_j1f.c (__ieee754_j1f): Likewise. * sysdeps/ieee754/flt-32/e_jnf.c (__ieee754_jnf): Likewise. * sysdeps/ieee754/ldbl-128/e_j1l.c (__ieee754_j1l): Set errno on underflow. * sysdeps/ieee754/ldbl-128/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-96/e_j1l.c (__ieee754_j1l): Likewise. * sysdeps/ieee754/ldbl-96/e_jnl.c (__ieee754_jnl): Likewise. * math/auto-libm-test-in: Do not allow missing errno setting for tests of j1 and jn. * math/auto-libm-test-out: Regenerated.
2015-10-23 21:37:33 +00:00
jn 10 min
jn 10 -min
jn 10 min_subnorm
jn 10 -min_subnorm
Fix j1, jn missing underflows (bug 16559). Similar to various other bugs in this area, j1 and jn implementations can fail to raise the underflow exception when the internal computation is exact although the actual function is inexact. This patch forces the exception in a similar way to other such fixes. (The ldbl-128 / ldbl-128ibm j1l implementation is different and doesn't need a change for this until spurious underflows in it are fixed.) Tested for x86_64, x86, mips64 and powerpc. [BZ #16559] * sysdeps/ieee754/dbl-64/e_j1.c: Include <float.h>. (__ieee754_j1): Force underflow exception for small results. * sysdeps/ieee754/dbl-64/e_jn.c (__ieee754_jn): Likewise. * sysdeps/ieee754/flt-32/e_j1f.c: Include <float.h>. (__ieee754_j1f): Force underflow exception for small results. * sysdeps/ieee754/flt-32/e_jnf.c (__ieee754_jnf): Likewise. * sysdeps/ieee754/ldbl-128/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-96/e_j1l.c: Include <float.h>. (__ieee754_j1l): Force underflow exception for small results. * sysdeps/ieee754/ldbl-96/e_jnl.c (__ieee754_jnl): Likewise. * math/auto-libm-test-in: Add more tests of j1 and jn. * math/auto-libm-test-out: Regenerated.
2015-06-29 16:52:16 +00:00
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
lgamma max
Move tests of lgamma from libm-test.inc to auto-libm-test-in.
2013-12-08 23:56:28 +00:00
lgamma 1
lgamma 3
lgamma 0.5
lgamma 0.7
lgamma 1.2
Fix lgammaf spurious underflows (bug 18220). The flt-32 implementation of lgammaf produces spurious underflow exceptions for some large arguments, because of calculations involving x^-2 multiplied by small constants. This patch fixes this by adjusting the threshold for a simpler computation to 2**26 (the error in the simpler computation is on the order of 0.5 * log (x), for a result on the order of x * log (x)). Tested for x86_64 and x86. [BZ #18220] * sysdeps/ieee754/flt-32/e_lgammaf_r.c (__ieee754_lgammaf_r): Use 2**26 not 2**58 as threshold for returning x * (log (x) - 1). * math/auto-libm-test-in: Add another test of lgamma. * math/auto-libm-test-out: Regenerated.
2015-05-15 17:21:08 +00:00
lgamma 0x3.8p56
Move tests of lgamma from libm-test.inc to auto-libm-test-in.
2013-12-08 23:56:28 +00:00
lgamma 0x1p-5
lgamma -0x1p-5
lgamma 0x1p-10
lgamma -0x1p-10
lgamma 0x1p-15
lgamma -0x1p-15
lgamma 0x1p-20
lgamma -0x1p-20
lgamma 0x1p-25
lgamma -0x1p-25
lgamma 0x1p-30
lgamma -0x1p-30
lgamma 0x1p-40
lgamma -0x1p-40
lgamma 0x1p-50
lgamma -0x1p-50
lgamma 0x1p-60
lgamma -0x1p-60
lgamma 0x1p-64
lgamma -0x1p-64
lgamma 0x1p-70
lgamma -0x1p-70
lgamma 0x1p-100
lgamma -0x1p-100
lgamma 0x1p-126
lgamma -0x1p-126
lgamma 0x1p-149
lgamma -0x1p-149
lgamma 0x1p-200
lgamma -0x1p-200
lgamma 0x1p-500
lgamma -0x1p-500
lgamma 0x1p-1000
lgamma -0x1p-1000
lgamma 0x1p-1022
lgamma -0x1p-1022
lgamma 0x1p-1074
lgamma -0x1p-1074
lgamma 0x1p-5000
lgamma -0x1p-5000
lgamma 0x1p-10000
lgamma -0x1p-10000
lgamma 0x1p-16382
lgamma -0x1p-16382
lgamma 0x1p-16445
lgamma -0x1p-16445
lgamma 0x1p-16494
lgamma -0x1p-16494
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
lgamma -0x3.ec4298p+0
Move tests of lgamma from libm-test.inc to auto-libm-test-in.
2013-12-08 23:56:28 +00:00
Fix ldbl-128 / ldbl-128ibm lgamma overflow handling (bug 16347, bug 19046). The ldbl-128 / ldbl-128ibm implementation of lgamma has problems with its handling of large arguments. It has an overflow threshold that is correct only for ldbl-128, despite being used for both types - with diagnostic control macros as a temporary measure to disable warnings about that constant overflowing for ldbl-128ibm - and it has a calculation that's roughly x * log(x) - x, resulting in overflows for arguments that are roughly at most a factor 1/log(threshold) below the overflow threshold. This patch fixes both issues, using an overflow threshold appropriate for the type in question and adding another case for large arguments that avoids the possible intermediate overflow. Tested for x86_64, x86, mips64 and powerpc. [BZ #16347] [BZ #19046] * sysdeps/ieee754/ldbl-128/e_lgammal_r.c: Do not include <libc-internal.h>. (MAXLGM): Do not use diagnostic control macros. [LDBL_MANT_DIG == 106] (MAXLGM): Change value to overflow threshold for ldbl-128ibm. (__ieee754_lgammal_r): For large arguments, multiply by log - 1 instead of multiplying by log then subtracting. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated.
2015-10-01 17:15:54 +00:00
# Values +/- 10ulp from overflow threshold. (Values very close to
# overflow threshold produce results very close of that threshold,
# where a result inaccurate by a few ulp could differ from the ideal
# result in whether it overflows; +/- 10ulp is sufficient for overflow
# or its absence to be unambiguous under glibc's accuracy standards).
Remove type specific information from auto-libm-test-in Apply the following sed regexes to auto-libm-test-in in order: s/flt-32/binary32/ s/dbl-64/binary64/ s/ldbl-96-intel/intel96/ s/ldbl-96-m68k/m68k96/ s/ldbl-128ibm/ibm128/ s/ldbl-128/binary128/ and fixup ldbl-96 comment manually.
2016-06-08 19:41:05 +00:00
# This also means the ibm128 inputs are XFAILed for binary64 and
# the binary128 inputs for intel96 and m68k96, as too close to the
# threshold.
Fix ldbl-128 / ldbl-128ibm lgamma overflow handling (bug 16347, bug 19046). The ldbl-128 / ldbl-128ibm implementation of lgamma has problems with its handling of large arguments. It has an overflow threshold that is correct only for ldbl-128, despite being used for both types - with diagnostic control macros as a temporary measure to disable warnings about that constant overflowing for ldbl-128ibm - and it has a calculation that's roughly x * log(x) - x, resulting in overflows for arguments that are roughly at most a factor 1/log(threshold) below the overflow threshold. This patch fixes both issues, using an overflow threshold appropriate for the type in question and adding another case for large arguments that avoids the possible intermediate overflow. Tested for x86_64, x86, mips64 and powerpc. [BZ #16347] [BZ #19046] * sysdeps/ieee754/ldbl-128/e_lgammal_r.c: Do not include <libc-internal.h>. (MAXLGM): Do not use diagnostic control macros. [LDBL_MANT_DIG == 106] (MAXLGM): Change value to overflow threshold for ldbl-128ibm. (__ieee754_lgammal_r): For large arguments, multiply by log - 1 instead of multiplying by log then subtracting. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated.
2015-10-01 17:15:54 +00:00
lgamma 0x3.12be0cp+120
lgamma 0x3.12be6p+120
lgamma 0x5.d53649e2d4674p+1012
Update libm-test XFAILs for ibm128 format. This patch cleans up and updates the libm-test XFAILs for the ibm128 format. More of them are changed to use a new ibm128-libgcc conditional, to reflect that they are not in fact needed if you've patched libgcc to fix the known issues (at substantial performance cost). Many additional XFAILs are added for tests that fail with unpatched libgcc (most but not all of them xfail-rounding). Note that further such fixes will be needed for test-ldouble actually to pass with default libgcc (in particular, XFAILs for pow tests and for various affected tests directly embedded in libm-test.inc). With patched libgcc, there may be a few XFAILs needed but the results are already substantially clean apart from a few ulps differences. Tested for powerpc. * math/libm-test.inc (TEST_COND_ibm128_libgcc): New macro. (init_max_error) [TEST_COND_ibm128]: Increase maximum error allowed to 16 ulps. * math/auto-libm-test-in: Change most XFAILs for ibm128 to use ibm128-libgcc. XFAIL more tests for ibm128-libgcc. * math/auto-libm-test-out: Regenerated.
2017-01-04 23:29:42 +00:00
lgamma 0x5.d53649e2d46c8p+1012 xfail-rounding:ibm128-libgcc
lgamma 0x5.d53649e2d469dbc1f01e99fd52p+1012 xfail:binary64 xfail-rounding:ibm128-libgcc
lgamma 0x5.d53649e2d469dbc1f01e99fd7cp+1012 xfail:binary64 xfail-rounding:ibm128-libgcc
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
lgamma 0x5.c6aa645fffef5f5p+16368
lgamma 0x5.c6aa645fffef5ff8p+16368
lgamma 0x5.c6aa645fffef5fa912b9b480f7acp+16368 xfail:intel96 xfail:m68k96
lgamma 0x5.c6aa645fffef5fa912b9b480f8p+16368 xfail:intel96 xfail:m68k96
Fix ldbl-128 / ldbl-128ibm lgamma overflow handling (bug 16347, bug 19046). The ldbl-128 / ldbl-128ibm implementation of lgamma has problems with its handling of large arguments. It has an overflow threshold that is correct only for ldbl-128, despite being used for both types - with diagnostic control macros as a temporary measure to disable warnings about that constant overflowing for ldbl-128ibm - and it has a calculation that's roughly x * log(x) - x, resulting in overflows for arguments that are roughly at most a factor 1/log(threshold) below the overflow threshold. This patch fixes both issues, using an overflow threshold appropriate for the type in question and adding another case for large arguments that avoids the possible intermediate overflow. Tested for x86_64, x86, mips64 and powerpc. [BZ #16347] [BZ #19046] * sysdeps/ieee754/ldbl-128/e_lgammal_r.c: Do not include <libc-internal.h>. (MAXLGM): Do not use diagnostic control macros. [LDBL_MANT_DIG == 106] (MAXLGM): Change value to overflow threshold for ldbl-128ibm. (__ieee754_lgammal_r): For large arguments, multiply by log - 1 instead of multiplying by log then subtracting. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated.
2015-10-01 17:15:54 +00:00
Fix lgamma (negative) inaccuracy (bug 2542, bug 2543, bug 2558). The existing implementations of lgamma functions (except for the ia64 versions) use the reflection formula for negative arguments. This suffers large inaccuracy from cancellation near zeros of lgamma (near where the gamma function is +/- 1). This patch fixes this inaccuracy. For arguments above -2, there are no zeros and no large cancellation, while for sufficiently large negative arguments the zeros are so close to integers that even for integers +/- 1ulp the log(gamma(1-x)) term dominates and cancellation is not significant. Thus, it is only necessary to take special care about cancellation for arguments around a limited number of zeros. Accordingly, this patch uses precomputed tables of relevant zeros, expressed as the sum of two floating-point values. The log of the ratio of two sines can be computed accurately using log1p in cases where log would lose accuracy. The log of the ratio of two gamma(1-x) values can be computed using Stirling's approximation (the difference between two values of that approximation to lgamma being computable without computing the two values and then subtracting), with appropriate adjustments (which don't reduce accuracy too much) in cases where 1-x is too small to use Stirling's approximation directly. In the interval from -3 to -2, using the ratios of sines and of gamma(1-x) can still produce too much cancellation between those two parts of the computation (and that interval is also the worst interval for computing the ratio between gamma(1-x) values, which computation becomes more accurate, while being less critical for the final result, for larger 1-x). Because this can result in errors slightly above those accepted in glibc, this interval is instead dealt with by polynomial approximations. Separate polynomial approximations to (|gamma(x)|-1)(x-n)/(x-x0) are used for each interval of length 1/8 from -3 to -2, where n (-3 or -2) is the nearest integer to the 1/8-interval and x0 is the zero of lgamma in the relevant half-integer interval (-3 to -2.5 or -2.5 to -2). Together, the two approaches are intended to give sufficient accuracy for all negative arguments in the problem range. Outside that range, the previous implementation continues to be used. Tested for x86_64, x86, mips64 and powerpc. The mips64 and powerpc testing shows up pre-existing problems for ldbl-128 and ldbl-128ibm with large negative arguments giving spurious "invalid" exceptions (exposed by newly added tests for cases this patch doesn't affect the logic for); I'll address those problems separately. [BZ #2542] [BZ #2543] [BZ #2558] * sysdeps/ieee754/dbl-64/e_lgamma_r.c (__ieee754_lgamma_r): Call __lgamma_neg for arguments from -28.0 to -2.0. * sysdeps/ieee754/flt-32/e_lgammaf_r.c (__ieee754_lgammaf_r): Call __lgamma_negf for arguments from -15.0 to -2.0. * sysdeps/ieee754/ldbl-128/e_lgammal_r.c (__ieee754_lgammal_r): Call __lgamma_negl for arguments from -48.0 or -50.0 to -2.0. * sysdeps/ieee754/ldbl-96/e_lgammal_r.c (__ieee754_lgammal_r): Call __lgamma_negl for arguments from -33.0 to -2.0. * sysdeps/ieee754/dbl-64/lgamma_neg.c: New file. * sysdeps/ieee754/dbl-64/lgamma_product.c: Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c: Likewise. * sysdeps/ieee754/flt-32/lgamma_productf.c: Likewise. * sysdeps/ieee754/ldbl-128/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-128/lgamma_productl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/lgamma_productl.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_product.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_productl.c: Likewise. * sysdeps/generic/math_private.h (__lgamma_negf): New prototype. (__lgamma_neg): Likewise. (__lgamma_negl): Likewise. (__lgamma_product): Likewise. (__lgamma_productl): Likewise. * math/Makefile (libm-calls): Add lgamma_neg and lgamma_product. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-10 22:27:58 +00:00
lgamma -0x1.fa471547c2fe5p+1
lgamma -0x1.9260dcp+1
lgamma -0xffffffp-1
lgamma -0x1fffffffffffffp-1
lgamma -0xffffffffffffffffp-1
lgamma -0x3ffffffffffffffffffffffffffp-1
lgamma -0x1ffffffffffffffffffffffffffffp-1
Fix ldbl-128/ldbl-128ibm lgamma spurious "invalid", incorrect signgam (bug 18952). The ldbl-128 / ldbl-128ibm implementation of lgammal converts (the floor of minus) non-integer negative arguments to int to determine the value of signgam. When those values are outside the range of int, this produces spurious "invalid" exceptions and incorrect values of signgam. This patch fixes this by instead determining signgam through comparing half the integer in question to floor of half the integer. Tested for mips64, x86_64 and x86. [BZ #18952] * sysdeps/ieee754/ldbl-128/e_lgammal_r.c (__ieee754_lgammal_r): Do not convert non-integer negative arguments to int to determine the value of signgam. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated.
2015-09-11 15:34:25 +00:00
lgamma -0x100000000.8p0
lgamma -0x100000001.8p0
Fix lgamma (negative) inaccuracy (bug 2542, bug 2543, bug 2558). The existing implementations of lgamma functions (except for the ia64 versions) use the reflection formula for negative arguments. This suffers large inaccuracy from cancellation near zeros of lgamma (near where the gamma function is +/- 1). This patch fixes this inaccuracy. For arguments above -2, there are no zeros and no large cancellation, while for sufficiently large negative arguments the zeros are so close to integers that even for integers +/- 1ulp the log(gamma(1-x)) term dominates and cancellation is not significant. Thus, it is only necessary to take special care about cancellation for arguments around a limited number of zeros. Accordingly, this patch uses precomputed tables of relevant zeros, expressed as the sum of two floating-point values. The log of the ratio of two sines can be computed accurately using log1p in cases where log would lose accuracy. The log of the ratio of two gamma(1-x) values can be computed using Stirling's approximation (the difference between two values of that approximation to lgamma being computable without computing the two values and then subtracting), with appropriate adjustments (which don't reduce accuracy too much) in cases where 1-x is too small to use Stirling's approximation directly. In the interval from -3 to -2, using the ratios of sines and of gamma(1-x) can still produce too much cancellation between those two parts of the computation (and that interval is also the worst interval for computing the ratio between gamma(1-x) values, which computation becomes more accurate, while being less critical for the final result, for larger 1-x). Because this can result in errors slightly above those accepted in glibc, this interval is instead dealt with by polynomial approximations. Separate polynomial approximations to (|gamma(x)|-1)(x-n)/(x-x0) are used for each interval of length 1/8 from -3 to -2, where n (-3 or -2) is the nearest integer to the 1/8-interval and x0 is the zero of lgamma in the relevant half-integer interval (-3 to -2.5 or -2.5 to -2). Together, the two approaches are intended to give sufficient accuracy for all negative arguments in the problem range. Outside that range, the previous implementation continues to be used. Tested for x86_64, x86, mips64 and powerpc. The mips64 and powerpc testing shows up pre-existing problems for ldbl-128 and ldbl-128ibm with large negative arguments giving spurious "invalid" exceptions (exposed by newly added tests for cases this patch doesn't affect the logic for); I'll address those problems separately. [BZ #2542] [BZ #2543] [BZ #2558] * sysdeps/ieee754/dbl-64/e_lgamma_r.c (__ieee754_lgamma_r): Call __lgamma_neg for arguments from -28.0 to -2.0. * sysdeps/ieee754/flt-32/e_lgammaf_r.c (__ieee754_lgammaf_r): Call __lgamma_negf for arguments from -15.0 to -2.0. * sysdeps/ieee754/ldbl-128/e_lgammal_r.c (__ieee754_lgammal_r): Call __lgamma_negl for arguments from -48.0 or -50.0 to -2.0. * sysdeps/ieee754/ldbl-96/e_lgammal_r.c (__ieee754_lgammal_r): Call __lgamma_negl for arguments from -33.0 to -2.0. * sysdeps/ieee754/dbl-64/lgamma_neg.c: New file. * sysdeps/ieee754/dbl-64/lgamma_product.c: Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c: Likewise. * sysdeps/ieee754/flt-32/lgamma_productf.c: Likewise. * sysdeps/ieee754/ldbl-128/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-128/lgamma_productl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/lgamma_productl.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_product.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_productl.c: Likewise. * sysdeps/generic/math_private.h (__lgamma_negf): New prototype. (__lgamma_neg): Likewise. (__lgamma_negl): Likewise. (__lgamma_product): Likewise. (__lgamma_productl): Likewise. * math/Makefile (libm-calls): Add lgamma_neg and lgamma_product. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-10 22:27:58 +00:00
lgamma -0.25
lgamma -0.5
lgamma -0.75
lgamma -1.25
lgamma -1.5
lgamma -1.75
lgamma -0x2.08p0
lgamma -0x2.1p0
lgamma -0x2.18p0
lgamma -0x2.2p0
lgamma -0x2.28p0
lgamma -0x2.3p0
lgamma -0x2.38p0
lgamma -0x2.4p0
lgamma -0x2.48p0
lgamma -0x2.5p0
lgamma -0x2.58p0
lgamma -0x2.6p0
lgamma -0x2.68p0
lgamma -0x2.7p0
lgamma -0x2.78p0
lgamma -0x2.8p0
lgamma -0x2.88p0
lgamma -0x2.9p0
lgamma -0x2.98p0
lgamma -0x2.ap0
lgamma -0x2.a8p0
lgamma -0x2.bp0
lgamma -0x2.b8p0
lgamma -0x2.cp0
lgamma -0x2.c8p0
lgamma -0x2.dp0
lgamma -0x2.d8p0
lgamma -0x2.ep0
lgamma -0x2.e8p0
lgamma -0x2.fp0
lgamma -0x2.f8p0
lgamma -0x3.08p0
lgamma -0x3.1p0
lgamma -0x3.18p0
lgamma -0x3.2p0
lgamma -0x3.28p0
lgamma -0x3.3p0
lgamma -0x3.38p0
lgamma -0x3.4p0
lgamma -0x3.48p0
lgamma -0x3.5p0
lgamma -0x3.58p0
lgamma -0x3.6p0
lgamma -0x3.68p0
lgamma -0x3.7p0
lgamma -0x3.78p0
lgamma -0x3.8p0
lgamma -0x3.88p0
lgamma -0x3.9p0
lgamma -0x3.98p0
lgamma -0x3.ap0
lgamma -0x3.a8p0
lgamma -0x3.bp0
lgamma -0x3.b8p0
lgamma -0x3.cp0
lgamma -0x3.c8p0
lgamma -0x3.dp0
lgamma -0x3.d8p0
lgamma -0x3.ep0
lgamma -0x3.e8p0
lgamma -0x3.fp0
lgamma -0x3.f8p0
lgamma -4.25
lgamma -4.5
lgamma -4.75
lgamma -5.25
lgamma -5.5
lgamma -5.75
lgamma -6.25
lgamma -6.5
lgamma -6.75
lgamma -7.25
lgamma -7.5
lgamma -7.75
lgamma -8.25
lgamma -8.5
lgamma -8.75
lgamma -9.25
lgamma -9.5
lgamma -9.75
lgamma -10.25
lgamma -10.5
lgamma -10.75
lgamma -11.25
lgamma -11.5
lgamma -11.75
lgamma -12.25
lgamma -12.5
lgamma -12.75
lgamma -13.25
lgamma -13.5
lgamma -13.75
lgamma -14.25
lgamma -14.5
lgamma -14.75
lgamma -15.25
lgamma -15.5
lgamma -15.75
lgamma -16.25
lgamma -16.5
lgamma -16.75
lgamma -17.25
lgamma -17.5
lgamma -17.75
lgamma -18.25
lgamma -18.5
lgamma -18.75
lgamma -19.25
lgamma -19.5
lgamma -19.75
lgamma -20.25
lgamma -20.5
lgamma -20.75
lgamma -21.25
lgamma -21.5
lgamma -21.75
lgamma -22.25
lgamma -22.5
lgamma -22.75
lgamma -23.25
lgamma -23.5
lgamma -23.75
lgamma -24.25
lgamma -24.5
lgamma -24.75
lgamma -25.25
lgamma -25.5
lgamma -25.75
lgamma -26.25
lgamma -26.5
lgamma -26.75
lgamma -27.25
lgamma -27.5
lgamma -27.75
lgamma -28.25
lgamma -28.5
lgamma -28.75
lgamma -29.25
lgamma -29.5
lgamma -29.75
lgamma -30.25
lgamma -30.5
lgamma -30.75
lgamma -31.25
lgamma -31.5
lgamma -31.75
lgamma -32.25
lgamma -32.5
lgamma -32.75
lgamma -33.25
lgamma -33.5
lgamma -33.75
lgamma -34.25
lgamma -34.5
lgamma -34.75
lgamma -35.25
lgamma -35.5
lgamma -35.75
lgamma -36.25
lgamma -36.5
lgamma -36.75
lgamma -37.25
lgamma -37.5
lgamma -37.75
lgamma -38.25
lgamma -38.5
lgamma -38.75
lgamma -39.25
lgamma -39.5
lgamma -39.75
lgamma -40.25
lgamma -40.5
lgamma -40.75
lgamma -41.25
lgamma -41.5
lgamma -41.75
lgamma -42.25
lgamma -42.5
lgamma -42.75
lgamma -43.25
lgamma -43.5
lgamma -43.75
lgamma -44.25
lgamma -44.5
lgamma -44.75
lgamma -45.25
lgamma -45.5
lgamma -45.75
lgamma -46.25
lgamma -46.5
lgamma -46.75
lgamma -47.25
lgamma -47.5
lgamma -47.75
lgamma -48.25
lgamma -48.5
lgamma -48.75
lgamma -49.25
lgamma -49.5
lgamma -49.75
lgamma -50.25
lgamma -50.5
lgamma -50.75
lgamma -51.25
lgamma -51.5
lgamma -51.75
lgamma -52.25
lgamma -52.5
lgamma -52.75
lgamma -53.25
lgamma -53.5
lgamma -53.75
lgamma -54.25
lgamma -54.5
lgamma -54.75
lgamma -55.25
lgamma -55.5
lgamma -55.75
lgamma -56.25
lgamma -56.5
lgamma -56.75
lgamma -57.25
lgamma -57.5
lgamma -57.75
lgamma -58.25
lgamma -58.5
lgamma -58.75
lgamma -59.25
lgamma -59.5
lgamma -59.75
lgamma -60.25
lgamma -60.5
lgamma -60.75
Remove type specific information from auto-libm-test-in Apply the following sed regexes to auto-libm-test-in in order: s/flt-32/binary32/ s/dbl-64/binary64/ s/ldbl-96-intel/intel96/ s/ldbl-96-m68k/m68k96/ s/ldbl-128ibm/ibm128/ s/ldbl-128/binary128/ and fixup ldbl-96 comment manually.
2016-06-08 19:41:05 +00:00
# Integers +/- 1ulp for binary128 (gen-auto-libm-tests will round these
Fix lgamma (negative) inaccuracy (bug 2542, bug 2543, bug 2558). The existing implementations of lgamma functions (except for the ia64 versions) use the reflection formula for negative arguments. This suffers large inaccuracy from cancellation near zeros of lgamma (near where the gamma function is +/- 1). This patch fixes this inaccuracy. For arguments above -2, there are no zeros and no large cancellation, while for sufficiently large negative arguments the zeros are so close to integers that even for integers +/- 1ulp the log(gamma(1-x)) term dominates and cancellation is not significant. Thus, it is only necessary to take special care about cancellation for arguments around a limited number of zeros. Accordingly, this patch uses precomputed tables of relevant zeros, expressed as the sum of two floating-point values. The log of the ratio of two sines can be computed accurately using log1p in cases where log would lose accuracy. The log of the ratio of two gamma(1-x) values can be computed using Stirling's approximation (the difference between two values of that approximation to lgamma being computable without computing the two values and then subtracting), with appropriate adjustments (which don't reduce accuracy too much) in cases where 1-x is too small to use Stirling's approximation directly. In the interval from -3 to -2, using the ratios of sines and of gamma(1-x) can still produce too much cancellation between those two parts of the computation (and that interval is also the worst interval for computing the ratio between gamma(1-x) values, which computation becomes more accurate, while being less critical for the final result, for larger 1-x). Because this can result in errors slightly above those accepted in glibc, this interval is instead dealt with by polynomial approximations. Separate polynomial approximations to (|gamma(x)|-1)(x-n)/(x-x0) are used for each interval of length 1/8 from -3 to -2, where n (-3 or -2) is the nearest integer to the 1/8-interval and x0 is the zero of lgamma in the relevant half-integer interval (-3 to -2.5 or -2.5 to -2). Together, the two approaches are intended to give sufficient accuracy for all negative arguments in the problem range. Outside that range, the previous implementation continues to be used. Tested for x86_64, x86, mips64 and powerpc. The mips64 and powerpc testing shows up pre-existing problems for ldbl-128 and ldbl-128ibm with large negative arguments giving spurious "invalid" exceptions (exposed by newly added tests for cases this patch doesn't affect the logic for); I'll address those problems separately. [BZ #2542] [BZ #2543] [BZ #2558] * sysdeps/ieee754/dbl-64/e_lgamma_r.c (__ieee754_lgamma_r): Call __lgamma_neg for arguments from -28.0 to -2.0. * sysdeps/ieee754/flt-32/e_lgammaf_r.c (__ieee754_lgammaf_r): Call __lgamma_negf for arguments from -15.0 to -2.0. * sysdeps/ieee754/ldbl-128/e_lgammal_r.c (__ieee754_lgammal_r): Call __lgamma_negl for arguments from -48.0 or -50.0 to -2.0. * sysdeps/ieee754/ldbl-96/e_lgammal_r.c (__ieee754_lgammal_r): Call __lgamma_negl for arguments from -33.0 to -2.0. * sysdeps/ieee754/dbl-64/lgamma_neg.c: New file. * sysdeps/ieee754/dbl-64/lgamma_product.c: Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c: Likewise. * sysdeps/ieee754/flt-32/lgamma_productf.c: Likewise. * sysdeps/ieee754/ldbl-128/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-128/lgamma_productl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/lgamma_productl.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_product.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_productl.c: Likewise. * sysdeps/generic/math_private.h (__lgamma_negf): New prototype. (__lgamma_neg): Likewise. (__lgamma_negl): Likewise. (__lgamma_product): Likewise. (__lgamma_productl): Likewise. * math/Makefile (libm-calls): Add lgamma_neg and lgamma_product. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-10 22:27:58 +00:00
# to produce integers +/- 1ulp for other formats).
lgamma -0xf.fffffffffffffffffffffffffff8p-4
lgamma -0x1.0000000000000000000000000001p+0
lgamma -0x1.ffffffffffffffffffffffffffffp+0
lgamma -0x2.0000000000000000000000000002p+0
lgamma -0x2.fffffffffffffffffffffffffffep+0
lgamma -0x3.0000000000000000000000000002p+0
lgamma -0x3.fffffffffffffffffffffffffffep+0
lgamma -0x4.0000000000000000000000000004p+0
lgamma -0x4.fffffffffffffffffffffffffffcp+0
lgamma -0x5.0000000000000000000000000004p+0
lgamma -0x5.fffffffffffffffffffffffffffcp+0
lgamma -0x6.0000000000000000000000000004p+0
lgamma -0x6.fffffffffffffffffffffffffffcp+0
lgamma -0x7.0000000000000000000000000004p+0
lgamma -0x7.fffffffffffffffffffffffffffcp+0
lgamma -0x8.0000000000000000000000000008p+0
lgamma -0x8.fffffffffffffffffffffffffff8p+0
lgamma -0x9.0000000000000000000000000008p+0
lgamma -0x9.fffffffffffffffffffffffffff8p+0
lgamma -0xa.0000000000000000000000000008p+0
lgamma -0xa.fffffffffffffffffffffffffff8p+0
lgamma -0xb.0000000000000000000000000008p+0
lgamma -0xb.fffffffffffffffffffffffffff8p+0
lgamma -0xc.0000000000000000000000000008p+0
lgamma -0xc.fffffffffffffffffffffffffff8p+0
lgamma -0xd.0000000000000000000000000008p+0
lgamma -0xd.fffffffffffffffffffffffffff8p+0
lgamma -0xe.0000000000000000000000000008p+0
lgamma -0xe.fffffffffffffffffffffffffff8p+0
lgamma -0xf.0000000000000000000000000008p+0
lgamma -0xf.fffffffffffffffffffffffffff8p+0
lgamma -0x1.0000000000000000000000000001p+4
lgamma -0x1.0fffffffffffffffffffffffffffp+4
lgamma -0x1.1000000000000000000000000001p+4
lgamma -0x1.1fffffffffffffffffffffffffffp+4
lgamma -0x1.2000000000000000000000000001p+4
lgamma -0x1.2fffffffffffffffffffffffffffp+4
lgamma -0x1.3000000000000000000000000001p+4
lgamma -0x1.3fffffffffffffffffffffffffffp+4
lgamma -0x1.4000000000000000000000000001p+4
lgamma -0x1.4fffffffffffffffffffffffffffp+4
lgamma -0x1.5000000000000000000000000001p+4
lgamma -0x1.5fffffffffffffffffffffffffffp+4
lgamma -0x1.6000000000000000000000000001p+4
lgamma -0x1.6fffffffffffffffffffffffffffp+4
lgamma -0x1.7000000000000000000000000001p+4
lgamma -0x1.7fffffffffffffffffffffffffffp+4
lgamma -0x1.8000000000000000000000000001p+4
lgamma -0x1.8fffffffffffffffffffffffffffp+4
lgamma -0x1.9000000000000000000000000001p+4
lgamma -0x1.9fffffffffffffffffffffffffffp+4
lgamma -0x1.a000000000000000000000000001p+4
lgamma -0x1.afffffffffffffffffffffffffffp+4
lgamma -0x1.b000000000000000000000000001p+4
lgamma -0x1.bfffffffffffffffffffffffffffp+4
lgamma -0x1.c000000000000000000000000001p+4
lgamma -0x1.cfffffffffffffffffffffffffffp+4
lgamma -0x1.d000000000000000000000000001p+4
lgamma -0x1.dfffffffffffffffffffffffffffp+4
lgamma -0x1.e000000000000000000000000001p+4
lgamma -0x1.efffffffffffffffffffffffffffp+4
lgamma -0x1.f000000000000000000000000001p+4
lgamma -0x1.ffffffffffffffffffffffffffffp+4
lgamma -0x2.0000000000000000000000000002p+4
lgamma -0x2.0ffffffffffffffffffffffffffep+4
lgamma -0x2.1000000000000000000000000002p+4
lgamma -0x2.1ffffffffffffffffffffffffffep+4
lgamma -0x2.2000000000000000000000000002p+4
lgamma -0x2.2ffffffffffffffffffffffffffep+4
lgamma -0x2.3000000000000000000000000002p+4
lgamma -0x2.3ffffffffffffffffffffffffffep+4
lgamma -0x2.4000000000000000000000000002p+4
lgamma -0x2.4ffffffffffffffffffffffffffep+4
lgamma -0x2.5000000000000000000000000002p+4
lgamma -0x2.5ffffffffffffffffffffffffffep+4
lgamma -0x2.6000000000000000000000000002p+4
lgamma -0x2.6ffffffffffffffffffffffffffep+4
lgamma -0x2.7000000000000000000000000002p+4
lgamma -0x2.7ffffffffffffffffffffffffffep+4
lgamma -0x2.8000000000000000000000000002p+4
lgamma -0x2.8ffffffffffffffffffffffffffep+4
lgamma -0x2.9000000000000000000000000002p+4
lgamma -0x2.9ffffffffffffffffffffffffffep+4
lgamma -0x2.a000000000000000000000000002p+4
lgamma -0x2.affffffffffffffffffffffffffep+4
lgamma -0x2.b000000000000000000000000002p+4
lgamma -0x2.bffffffffffffffffffffffffffep+4
lgamma -0x2.c000000000000000000000000002p+4
lgamma -0x2.cffffffffffffffffffffffffffep+4
lgamma -0x2.d000000000000000000000000002p+4
lgamma -0x2.dffffffffffffffffffffffffffep+4
lgamma -0x2.e000000000000000000000000002p+4
lgamma -0x2.effffffffffffffffffffffffffep+4
lgamma -0x2.f000000000000000000000000002p+4
lgamma -0x2.fffffffffffffffffffffffffffep+4
lgamma -0x3.0000000000000000000000000002p+4
lgamma -0x3.0ffffffffffffffffffffffffffep+4
lgamma -0x3.1000000000000000000000000002p+4
lgamma -0x3.1ffffffffffffffffffffffffffep+4
lgamma -0x3.2000000000000000000000000002p+4
lgamma -0x3.2ffffffffffffffffffffffffffep+4
lgamma -0x3.3000000000000000000000000002p+4
lgamma -0x3.3ffffffffffffffffffffffffffep+4
lgamma -0x3.4000000000000000000000000002p+4
lgamma -0x3.4ffffffffffffffffffffffffffep+4
lgamma -0x3.5000000000000000000000000002p+4
lgamma -0x3.5ffffffffffffffffffffffffffep+4
lgamma -0x3.6000000000000000000000000002p+4
lgamma -0x3.6ffffffffffffffffffffffffffep+4
lgamma -0x3.7000000000000000000000000002p+4
lgamma -0x3.7ffffffffffffffffffffffffffep+4
lgamma -0x3.8000000000000000000000000002p+4
lgamma -0x3.8ffffffffffffffffffffffffffep+4
lgamma -0x3.9000000000000000000000000002p+4
lgamma -0x3.9ffffffffffffffffffffffffffep+4
lgamma -0x3.a000000000000000000000000002p+4
lgamma -0x3.affffffffffffffffffffffffffep+4
lgamma -0x3.b000000000000000000000000002p+4
lgamma -0x3.bffffffffffffffffffffffffffep+4
lgamma -0x3.c000000000000000000000000002p+4
# Zeroes of lgamma, until the point where they just duplicate integers
# +/- 1ulp.
lgamma -0x2.74ff92c01f0d82abec9f315f1a0712c334804d9cp+0
lgamma -0x2.bf6821437b20197995a4b4641eaebf4b00b482ap+0
lgamma -0x3.24c1b793cb35efb8be699ad3d9ba65454cb7fac8p+0
lgamma -0x3.f48e2a8f85fca170d4561291236cc320a4887d1cp+0
lgamma -0x4.0a139e16656030c39f0b0de18112ac17bfd6be9p+0
lgamma -0x4.fdd5de9bbabf3510d0aa4076988501d7d7812528p+0
lgamma -0x5.021a95fc2db6432a4c56e595394decc6af0430d8p+0
lgamma -0x5.ffa4bd647d0357dd4ed62cbd31edf8e3f8e5deb8p+0
lgamma -0x6.005ac9625f233b607c2d96d16385cb86ac56934p+0
lgamma -0x6.fff2fddae1bbff3d626b65c23fd21f40300a3ba8p+0
lgamma -0x7.000cff7b7f87adf4482dcdb98782ab2661ca58bp+0
lgamma -0x7.fffe5fe05673c3ca9e82b522b0ca9d2e8837cd2p+0
lgamma -0x8.0001a01459fc9f60cb3cec1cec8576677ca538ep+0
lgamma -0x8.ffffd1c425e80ffc864e95749259e7e20210e8p+0
lgamma -0x9.00002e3bb47d86d6d843fedc351deb7ad09ec5fp+0
lgamma -0x9.fffffb606bdfdcd062ae77a50547c69d2eb6f34p+0
lgamma -0xa.0000049f93bb9927b45d95e15441e03086db914p+0
lgamma -0xa.ffffff9466e9f1b36dacd2adbd18d05a4e45806p+0
lgamma -0xb.0000006b9915315d965a6ffea40e4bea39000ddp+0
lgamma -0xb.fffffff7089387387de41acc3d3c978bd839c8cp+0
lgamma -0xc.00000008f76c7731567c0f0250f387920df5676p+0
lgamma -0xc.ffffffff4f6dcf617f97a5ffc757d548d2890cdp+0
lgamma -0xd.00000000b092309c06683dd1b903e3700857a16p+0
lgamma -0xd.fffffffff36345ab9e184a3e09d1176dc48e47fp+0
lgamma -0xe.000000000c9cba545e94e75ec5718f753e2501ep+0
lgamma -0xe.ffffffffff28c060c6604ef30371f89d37357cap+0
lgamma -0xf.0000000000d73f9f399bd0e420f85e9ee31b0b9p+0
lgamma -0xf.fffffffffff28c060c6621f512e72e4d113626ap+0
lgamma -0x1.000000000000d73f9f399da1424bf93b91f177dp+4
lgamma -0x1.0ffffffffffff3569c47e7a93e1c46a08a2e008ap+4
lgamma -0x1.1000000000000ca963b8185688876ca5a3a64ec2p+4
lgamma -0x1.1fffffffffffff4bec3ce234132d08b2b726187cp+4
lgamma -0x1.20000000000000b413c31dcbeca4c3b2ffacbb4ap+4
lgamma -0x1.2ffffffffffffff685b25cbf5f545ced932e3848p+4
lgamma -0x1.30000000000000097a4da340a0ab81b7b1f1f002p+4
lgamma -0x1.3fffffffffffffff86af516ff7f76bd67e720d58p+4
lgamma -0x1.40000000000000007950ae9008089413ccc8a354p+4
lgamma -0x1.4ffffffffffffffffa391c4248c2a39cfdd49d4ap+4
lgamma -0x1.500000000000000005c6e3bdb73d5c62f55ed532p+4
lgamma -0x1.5fffffffffffffffffbcc71a49201eb5aeb96c74p+4
lgamma -0x1.6000000000000000004338e5b6dfe14a513fb4dp+4
lgamma -0x1.6ffffffffffffffffffd13c97d9d38fcc4d08d7p+4
lgamma -0x1.70000000000000000002ec368262c7033b2f6f32p+4
lgamma -0x1.7fffffffffffffffffffe0d30fe68d0a88335b4cp+4
lgamma -0x1.800000000000000000001f2cf01972f577cca4b4p+4
lgamma -0x1.8ffffffffffffffffffffec0c3322e9a0572b1bcp+4
lgamma -0x1.90000000000000000000013f3ccdd165fa8d4e44p+4
lgamma -0x1.9ffffffffffffffffffffff3b8bd01cad8d32e38p+4
lgamma -0x1.a0000000000000000000000c4742fe35272cd1c8p+4
lgamma -0x1.afffffffffffffffffffffff8b9538f48cc5737ep+4
lgamma -0x1.b00000000000000000000000746ac70b733a8c82p+4
lgamma -0x1.bffffffffffffffffffffffffbd79d7672bde8b2p+4
lgamma -0x1.c00000000000000000000000042862898d42174ep+4
lgamma -0x1.cfffffffffffffffffffffffffdb4c0ce9794ea6p+4
lgamma -0x1.d000000000000000000000000024b3f31686b15ap+4
lgamma -0x1.dffffffffffffffffffffffffffec6cd3afb82ap+4
lgamma -0x1.e0000000000000000000000000013932c5047d6p+4
Add more tests of lgamma. This patch adds some randomly-generated tests of lgamma that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 22:17:19 +00:00
lgamma 0x8.8d2d5p+0
lgamma 0x1.6a324ap+52
lgamma 0x9.62f59p+0
lgamma 0xa.d55d6b4d78e28p+0
lgamma 0x8.d6315p+0
lgamma 0xb.2e679p+0
lgamma 0xb.01191p+0
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
lgamma 0xb.26fdap+0
Add more tests of csqrt, lgamma, log10, sinh. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of csqrt, lgamma, log10 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-08 17:55:11 +00:00
lgamma 0xb.4ad0ap+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
lgamma 0xe.7a678p+20
Fix lgamma (negative) inaccuracy (bug 2542, bug 2543, bug 2558). The existing implementations of lgamma functions (except for the ia64 versions) use the reflection formula for negative arguments. This suffers large inaccuracy from cancellation near zeros of lgamma (near where the gamma function is +/- 1). This patch fixes this inaccuracy. For arguments above -2, there are no zeros and no large cancellation, while for sufficiently large negative arguments the zeros are so close to integers that even for integers +/- 1ulp the log(gamma(1-x)) term dominates and cancellation is not significant. Thus, it is only necessary to take special care about cancellation for arguments around a limited number of zeros. Accordingly, this patch uses precomputed tables of relevant zeros, expressed as the sum of two floating-point values. The log of the ratio of two sines can be computed accurately using log1p in cases where log would lose accuracy. The log of the ratio of two gamma(1-x) values can be computed using Stirling's approximation (the difference between two values of that approximation to lgamma being computable without computing the two values and then subtracting), with appropriate adjustments (which don't reduce accuracy too much) in cases where 1-x is too small to use Stirling's approximation directly. In the interval from -3 to -2, using the ratios of sines and of gamma(1-x) can still produce too much cancellation between those two parts of the computation (and that interval is also the worst interval for computing the ratio between gamma(1-x) values, which computation becomes more accurate, while being less critical for the final result, for larger 1-x). Because this can result in errors slightly above those accepted in glibc, this interval is instead dealt with by polynomial approximations. Separate polynomial approximations to (|gamma(x)|-1)(x-n)/(x-x0) are used for each interval of length 1/8 from -3 to -2, where n (-3 or -2) is the nearest integer to the 1/8-interval and x0 is the zero of lgamma in the relevant half-integer interval (-3 to -2.5 or -2.5 to -2). Together, the two approaches are intended to give sufficient accuracy for all negative arguments in the problem range. Outside that range, the previous implementation continues to be used. Tested for x86_64, x86, mips64 and powerpc. The mips64 and powerpc testing shows up pre-existing problems for ldbl-128 and ldbl-128ibm with large negative arguments giving spurious "invalid" exceptions (exposed by newly added tests for cases this patch doesn't affect the logic for); I'll address those problems separately. [BZ #2542] [BZ #2543] [BZ #2558] * sysdeps/ieee754/dbl-64/e_lgamma_r.c (__ieee754_lgamma_r): Call __lgamma_neg for arguments from -28.0 to -2.0. * sysdeps/ieee754/flt-32/e_lgammaf_r.c (__ieee754_lgammaf_r): Call __lgamma_negf for arguments from -15.0 to -2.0. * sysdeps/ieee754/ldbl-128/e_lgammal_r.c (__ieee754_lgammal_r): Call __lgamma_negl for arguments from -48.0 or -50.0 to -2.0. * sysdeps/ieee754/ldbl-96/e_lgammal_r.c (__ieee754_lgammal_r): Call __lgamma_negl for arguments from -33.0 to -2.0. * sysdeps/ieee754/dbl-64/lgamma_neg.c: New file. * sysdeps/ieee754/dbl-64/lgamma_product.c: Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c: Likewise. * sysdeps/ieee754/flt-32/lgamma_productf.c: Likewise. * sysdeps/ieee754/ldbl-128/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-128/lgamma_productl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/lgamma_productl.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_negl.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_product.c: Likewise. * sysdeps/ieee754/ldbl-96/lgamma_productl.c: Likewise. * sysdeps/generic/math_private.h (__lgamma_negf): New prototype. (__lgamma_neg): Likewise. (__lgamma_negl): Likewise. (__lgamma_product): Likewise. (__lgamma_productl): Likewise. * math/Makefile (libm-calls): Add lgamma_neg and lgamma_product. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-10 22:27:58 +00:00
lgamma -0x2.dea4ccp-4
lgamma -0x2.dd306p-4
lgamma -0x1.bdc8bp+0
lgamma -0x4.0a82e8p-4
Add more randomly-generated libm tests. This patch adds more libm test inputs found through random test generation to increase observed ulps on x86_64. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, atanh, cbrt, cosh, csqrt, erfc, expm1 and lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-11 15:03:10 +00:00
lgamma -0x1.bca67ap+0
lgamma -0x3.46446bb6a23aap+0
lgamma -0x3.f3d2c40911814p+0
Add more tests of lgamma. This patch adds some randomly-generated tests of lgamma that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of lgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 22:17:19 +00:00
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
log 1
log e
log 1/e
log 2
log 10
log 0.75
Add more libm tests (ilogb, is*, j0, j1, jn, lgamma, log*). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of log, log10, log1p and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (MAX_EXP): New macro. (ilogb_test_data): Add more tests. (isfinite_test_data): Likewise. (isgreater_test_data): Likewise. (isgreaterequal_test_data): Likewise. (isinf_test_data): Likewise. (isless_test_data): Likewise. (islessequal_test_data): Likewise. (islessgreater_test_data): Likewise. (isnan_test_data): Likewise. (isnormal_test_data): Likewise. (issignaling_test_data): Likewise. (isunordered_test_data): Likewise. (j0_test_data): Likewise. (j1_test_data): Likewise. (jn_test_data): Likewise. (lgamma_test_data): Likewise. (log_test_data): Likewise. (log10_test_data): Likewise. (log1p_test_data): Likewise. (log2_test_data): Likewise. (logb_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-10-23 22:46:05 +00:00
log 0x1.000002p0
log 0x1.0000000000001p0
log 0x1.0000000000000002p0
log 0x1.000000000000000000000000008p0
log 0x1.0000000000000000000000000001p0
log 0x0.ffffffp0
log 0x0.fffffffffffff8p0
log 0x0.ffffffffffffffffp0
log 0x0.ffffffffffffffffffffffffffcp0
log 0x0.ffffffffffffffffffffffffffff8p0
Fix ldbl-128 logl for subnormals (bug 16338). This patch fixes bug 16338, ldbl-128 logl not handling subnormals (with consequent inaccuracy for lgammal as well). The fix is simply to use __frexpl when determining the exponent, as done already in log2l and log10l. Given the lack of testing of small arguments to any of the log* functions, appropriate tests are added for all of them. Tested x86_64 and x86 and ulps updated accordingly, and spot tests also run for mips64 to confirm the ldbl-128 fix. Note that while this fixes lgammal inaccuracy for small positive arguments, I suspect that there will still be problems with spurious underflows in that case. * sysdeps/ieee754/ldbl-128/e_logl.c (__ieee754_logl): Use __frexpl to determine exponent and adjust argument to have exponent of -1. * math/auto-libm-test-in: Add more tests of log, log10, log1p and log2. * math/auto-libm-test-out: Regenerated. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2013-12-18 11:38:27 +00:00
log min
log min_subnorm
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
log max
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
Add more tests of log, log10, log1p, log2. This patch adds some randomly-generated tests of log, log10, log1p and log2 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of log, log10, log2 and log1p. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 21:08:37 +00:00
log 0xb.0d5dfp-4
log 0x1.6c3f6p+0
log 0xa.ae688p-4
log 0x1.017f8ap+44
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
log 0x1.0b5c1ep+36
log 0x2.1b17c2887e938p+928
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
log 0x1.929d9cp+0
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
log 0x1.770072p+0
add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)
2020-11-25 18:03:33 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
log 0x1.d405118796bd8p-1
Add more tests of log, log10, log1p, log2. This patch adds some randomly-generated tests of log, log10, log1p and log2 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of log, log10, log2 and log1p. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 21:08:37 +00:00
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
log10 1
log10 0.1
log10 10.0
log10 100.0
log10 10000.0
log10 e
log10 0.75
Add more libm tests (ilogb, is*, j0, j1, jn, lgamma, log*). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of log, log10, log1p and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (MAX_EXP): New macro. (ilogb_test_data): Add more tests. (isfinite_test_data): Likewise. (isgreater_test_data): Likewise. (isgreaterequal_test_data): Likewise. (isinf_test_data): Likewise. (isless_test_data): Likewise. (islessequal_test_data): Likewise. (islessgreater_test_data): Likewise. (isnan_test_data): Likewise. (isnormal_test_data): Likewise. (issignaling_test_data): Likewise. (isunordered_test_data): Likewise. (j0_test_data): Likewise. (j1_test_data): Likewise. (jn_test_data): Likewise. (lgamma_test_data): Likewise. (log_test_data): Likewise. (log10_test_data): Likewise. (log1p_test_data): Likewise. (log2_test_data): Likewise. (logb_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-10-23 22:46:05 +00:00
log10 0x1.000002p0
log10 0x1.0000000000001p0
log10 0x1.0000000000000002p0
log10 0x1.000000000000000000000000008p0
log10 0x1.0000000000000000000000000001p0
log10 0x0.ffffffp0
log10 0x0.fffffffffffff8p0
log10 0x0.ffffffffffffffffp0
log10 0x0.ffffffffffffffffffffffffffcp0
log10 0x0.ffffffffffffffffffffffffffff8p0
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary128)
log10 0x1.6a291ea0aa11fb374f1df8b3ac6bp+0
Fix ldbl-128 logl for subnormals (bug 16338). This patch fixes bug 16338, ldbl-128 logl not handling subnormals (with consequent inaccuracy for lgammal as well). The fix is simply to use __frexpl when determining the exponent, as done already in log2l and log10l. Given the lack of testing of small arguments to any of the log* functions, appropriate tests are added for all of them. Tested x86_64 and x86 and ulps updated accordingly, and spot tests also run for mips64 to confirm the ldbl-128 fix. Note that while this fixes lgammal inaccuracy for small positive arguments, I suspect that there will still be problems with spurious underflows in that case. * sysdeps/ieee754/ldbl-128/e_logl.c (__ieee754_logl): Use __frexpl to determine exponent and adjust argument to have exponent of -1. * math/auto-libm-test-in: Add more tests of log, log10, log1p and log2. * math/auto-libm-test-out: Regenerated. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2013-12-18 11:38:27 +00:00
log10 min
log10 min_subnorm
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
log10 max
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
Add more tests of log, log10, log1p, log2. This patch adds some randomly-generated tests of log, log10, log1p and log2 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of log, log10, log2 and log1p. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 21:08:37 +00:00
log10 0x9.ad6e3p-4
log10 0x1.7163aep+0
log10 0xa.9d0d4p-4
log10 0x1.251ec6p+0
log10 0x1.022e82p+0
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
log10 0x9.b3727e3feb538p-4
Add more tests of csqrt, lgamma, log10, sinh. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of csqrt, lgamma, log10 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-08 17:55:11 +00:00
log10 0xf.bf1b2p-4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
log10 0x1.6b5f7ap+96
Add more tests of log, log10, log1p, log2. This patch adds some randomly-generated tests of log, log10, log1p and log2 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of log, log10, log2 and log1p. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 21:08:37 +00:00
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
log1p 0
log1p -0
log1p e-1
log1p -0.25
log1p -0.875
Add more libm tests (ilogb, is*, j0, j1, jn, lgamma, log*). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of log, log10, log1p and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (MAX_EXP): New macro. (ilogb_test_data): Add more tests. (isfinite_test_data): Likewise. (isgreater_test_data): Likewise. (isgreaterequal_test_data): Likewise. (isinf_test_data): Likewise. (isless_test_data): Likewise. (islessequal_test_data): Likewise. (islessgreater_test_data): Likewise. (isnan_test_data): Likewise. (isnormal_test_data): Likewise. (issignaling_test_data): Likewise. (isunordered_test_data): Likewise. (j0_test_data): Likewise. (j1_test_data): Likewise. (jn_test_data): Likewise. (lgamma_test_data): Likewise. (log_test_data): Likewise. (log10_test_data): Likewise. (log1p_test_data): Likewise. (log2_test_data): Likewise. (logb_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-10-23 22:46:05 +00:00
log1p 0x1p-5
log1p 0x1p-10
log1p 0x1p-15
log1p 0x1p-20
log1p 0x1p-25
log1p 0x1p-30
log1p 0x1p-35
log1p 0x1p-40
log1p 0x1p-45
log1p 0x1p-50
log1p 0x1p-55
log1p 0x1p-60
log1p 0x1p-100
log1p 0x1p-600
log1p 0x1p-10000
Fix log1p missing underflows (bug 16339). Similar to various other bugs in this area, some log1p implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. (The ldbl-128ibm implementation doesn't currently need any change as it already generates this exception, albeit through code that would generate spurious exceptions in other cases; special code for this issue will only be needed there when fixing the spurious exceptions.) Tested for x86_64, x86, powerpc and mips64. [BZ #16339] * sysdeps/i386/fpu/s_log1p.S (dbl_min): New object. (__log1p): Force underflow exception for results with small absolute value. * sysdeps/i386/fpu/s_log1pf.S (flt_min): New object. (__log1pf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/dbl-64/s_log1p.c: Include <float.h>. (__log1p): Force underflow exception for results with small absolute value. * sysdeps/ieee754/flt-32/s_log1pf.c: Include <float.h>. (__log1pf): Force underflow exception for results with small absolute value. * sysdeps/ieee754/ldbl-128/s_log1pl.c: Include <float.h>. (__log1pl): Force underflow exception for results with small absolute value. * math/auto-libm-test-in: Do not allow missing underflow exceptions from log1p. * math/auto-libm-test-out: Regenerated.
2015-05-14 23:38:07 +00:00
log1p min
log1p min_subnorm
log1p -min
log1p -min_subnorm
Fix log1pl (LDBL_MAX) in FE_UPWARD mode (bug 16564). Bug 16564 is spurious overflow of log1pl (LDBL_MAX) in FE_UPWARD mode, resulting from log1pl adding 1 to its argument (for arguments not close to 0), which overflows in that mode. This patch fixes this by avoiding adding 1 to large arguments (precisely what counts as large depends on the floating-point format). Tested x86_64 and x86, and spot-checked log1pl tests on mips64 and powerpc64. [BZ #16564] * sysdeps/i386/fpu/s_log1pl.S (__log1pl): Do not add 1 to positive arguments with exponent 65 or above. * sysdeps/ieee754/ldbl-128/s_log1pl.c (__log1pl): Do not add 1 to arguments 0x1p113L or above. * sysdeps/ieee754/ldbl-128ibm/s_log1pl.c (__log1pl): Do not add 1 to arguments 0x1p107L or above. * sysdeps/x86_64/fpu/s_log1pl.S (__log1pl): Do not add 1 to positive arguments with exponent 65 or above. * math/auto-libm-test-in: Add more tests of log1p. * math/auto-libm-test-out: Regenerated.
2014-05-14 12:38:56 +00:00
log1p 0x1p10
log1p 0x1p20
log1p 0x1p30
log1p 0x1p50
log1p 0x1p60
log1p 0x1p100
log1p 0x1p1000
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary128)
log1p 0x6.a0cf42befce9ed4085ef59254b48p-4
Fix log1pl (LDBL_MAX) in FE_UPWARD mode (bug 16564). Bug 16564 is spurious overflow of log1pl (LDBL_MAX) in FE_UPWARD mode, resulting from log1pl adding 1 to its argument (for arguments not close to 0), which overflows in that mode. This patch fixes this by avoiding adding 1 to large arguments (precisely what counts as large depends on the floating-point format). Tested x86_64 and x86, and spot-checked log1pl tests on mips64 and powerpc64. [BZ #16564] * sysdeps/i386/fpu/s_log1pl.S (__log1pl): Do not add 1 to positive arguments with exponent 65 or above. * sysdeps/ieee754/ldbl-128/s_log1pl.c (__log1pl): Do not add 1 to arguments 0x1p113L or above. * sysdeps/ieee754/ldbl-128ibm/s_log1pl.c (__log1pl): Do not add 1 to arguments 0x1p107L or above. * sysdeps/x86_64/fpu/s_log1pl.S (__log1pl): Do not add 1 to positive arguments with exponent 65 or above. * math/auto-libm-test-in: Add more tests of log1p. * math/auto-libm-test-out: Regenerated.
2014-05-14 12:38:56 +00:00
log1p max
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
Add more tests of log, log10, log1p, log2. This patch adds some randomly-generated tests of log, log10, log1p and log2 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of log, log10, log2 and log1p. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 21:08:37 +00:00
log1p 0x7.2a4368p-4
log1p 0x6.d3a118p-4
log1p 0x5.03f228p+0
log1p 0x7.264963888ac9p-4
log1p 0x8.786bdp-4
log1p 0x7.89dc17790eeb4p-4
log1p 0x9.81ccf8887c24a7bp-4
log1p 0xa.5028608bd65f38dp-4
log1p 0x5.bf78873e20a2d468p-4
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
log1p 0x7.aa5198p-4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
log1p 0x2.564fap+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
log1p 0x7.fc242a2235222ef8p-4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
log1p -0x4.f37d3c9ce0b14bdd86eb157df5d4p-4
log1p 0x7.2eca50c4d93196362b4f37f6e8dcp-4
log1p -0x6.3fef3067427e43dfcde9e48f74bcp-4
log1p 0x6.af53d00fd2845d4772260ef5adc4p-4
Add more tests of log, log10, log1p, log2. This patch adds some randomly-generated tests of log, log10, log1p and log2 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of log, log10, log2 and log1p. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 21:08:37 +00:00
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
log2 1
log2 e
log2 2.0
log2 16.0
log2 256.0
log2 0.75
Add more libm tests (ilogb, is*, j0, j1, jn, lgamma, log*). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of log, log10, log1p and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (MAX_EXP): New macro. (ilogb_test_data): Add more tests. (isfinite_test_data): Likewise. (isgreater_test_data): Likewise. (isgreaterequal_test_data): Likewise. (isinf_test_data): Likewise. (isless_test_data): Likewise. (islessequal_test_data): Likewise. (islessgreater_test_data): Likewise. (isnan_test_data): Likewise. (isnormal_test_data): Likewise. (issignaling_test_data): Likewise. (isunordered_test_data): Likewise. (j0_test_data): Likewise. (j1_test_data): Likewise. (jn_test_data): Likewise. (lgamma_test_data): Likewise. (log_test_data): Likewise. (log10_test_data): Likewise. (log1p_test_data): Likewise. (log2_test_data): Likewise. (logb_test_data): Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-10-23 22:46:05 +00:00
log2 0x1.000002p0
log2 0x1.0000000000001p0
log2 0x1.0000000000000002p0
log2 0x1.000000000000000000000000008p0
log2 0x1.0000000000000000000000000001p0
log2 0x0.ffffffp0
log2 0x0.fffffffffffff8p0
log2 0x0.ffffffffffffffffp0
log2 0x0.ffffffffffffffffffffffffffcp0
log2 0x0.ffffffffffffffffffffffffffff8p0
Add more tests of log2. In testing for x86_64 on an AMD processor, I observed libm test failures of the form: testing long double (without inline functions) Failure: Test: log2_downward (0x2.b7e151628aed4p+0) Result: is: 1.44269504088896356633e+00 0xb.8aa3b295c17f67600000p-3 should be: 1.44269504088896356622e+00 0xb.8aa3b295c17f67500000p-3 difference: 1.08420217248550443400e-19 0x8.00000000000000000000p-66 ulp : 1.0000 max.ulp : 0.0000 Maximal error of `log2_downward' is : 1 ulp accepted: 0 ulp These issues arise because the maximum ulps when regenerating on one processor are not the same as on another processor, so regeneration on several processors may be needed when updating libm-test-ulps to avoid failures for some users testing glibc - but such regeneration on multiple processors is inconvenient. Causes can be: on x86 and, for x86_64, for long double, variation in results of x87 instructions for transcendental operations between processors; on x86, variation in compiler excess precision between compiler versions and configurations; on any processor where the compiler may contract expressions using fused multiply-add, variation in what contraction occurs. Although it's hard to be sure libm-test-ulps covers all ulps that may be seen in any configuration for the given architecture, in practice it helps simply to add wider test coverage to make it more likely that, when testing on one processor, the ulps seen are the biggest that can be seen for that function on that processor, and hopefully they are also the biggest that can be seen for that function in other configurations for that architecture. Thus, this patch adds some tests of log2 that increase the ulps I see on x86_64 on an Intel processor, so that hopefully future from-scratch regenerations on that processor will produce ulps big enough not to have errors from testing on AMD processors. These tests were found by randomly generating inputs and seeing what produced ulps larger than those currently in libm-test-ulps. Of course such increases also improve the accuracy of the empirical table of known ulps generated from libm-test-ulps files that goes in the manual. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of log2. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-24 23:06:28 +00:00
log2 0x1.28d3b4p+0
log2 0xe.d99dap-4
log2 0x1.63d202d04392cp+0
log2 0xf.d9ce0b1a50e08p-4
log2 0x1.07465bdc7e41b52ep+0
log2 0xf.4dfb4p-48
Add more tests of log, log10, log1p, log2. This patch adds some randomly-generated tests of log, log10, log1p and log2 that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of log, log10, log2 and log1p. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 21:08:37 +00:00
log2 0x1.0a588ep+0
log2 0xb.e77c6p-4
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
log2 0x1.4fe37ep+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
log2 0x3.9b0754p+8
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
log2 0xb.e132ap-4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
log2 0xb.5bf82dc51f02035p-4
log2 0xb.7704dc9beb05p-4
log2 0xb.56f63c18e93eecdp-4
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (binary128)
log2 0xb.54170d5cfa8fd72a47d6bda19068p-4
Fix ldbl-128 logl for subnormals (bug 16338). This patch fixes bug 16338, ldbl-128 logl not handling subnormals (with consequent inaccuracy for lgammal as well). The fix is simply to use __frexpl when determining the exponent, as done already in log2l and log10l. Given the lack of testing of small arguments to any of the log* functions, appropriate tests are added for all of them. Tested x86_64 and x86 and ulps updated accordingly, and spot tests also run for mips64 to confirm the ldbl-128 fix. Note that while this fixes lgammal inaccuracy for small positive arguments, I suspect that there will still be problems with spurious underflows in that case. * sysdeps/ieee754/ldbl-128/e_logl.c (__ieee754_logl): Use __frexpl to determine exponent and adjust argument to have exponent of -1. * math/auto-libm-test-in: Add more tests of log, log10, log1p and log2. * math/auto-libm-test-out: Regenerated. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2013-12-18 11:38:27 +00:00
log2 min
log2 min_subnorm
Add more libm tests (fmod, fpclassify, frexp, hypot, ilogb, j0, j1, jn, log, log10, log2). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. 2015-10-21 Joseph Myers <joseph@codesourcery.com> * math/auto-libm-test-in: Add more tests of hypot, j0, j1, jn, log, log10 and log2. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (fmod_test_data): Add more tests. (fpclassify_test_data): Likewise. (frexp_test_data): Likewise. (hypot_test_data): Likewise. (ilogb_test_data): Likewise.
2015-10-21 01:05:01 +00:00
log2 max
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
Add narrowing multiply functions. This patch adds the narrowing multiply functions from TS 18661-1 to glibc's libm: fmul, fmull, dmull, f32mulf64, f32mulf32x, f32xmulf64 for all configurations; f32mulf64x, f32mulf128, f64mulf64x, f64mulf128, f32xmulf64x, f32xmulf128, f64xmulf128 for configurations with _Float64x and _Float128; __nldbl_dmull for ldbl-opt. The changes are mostly essentially the same as for the narrowing add functions, so the description of those generally applies to this patch as well. f32xmulf64 for i386 cannot use precision control as used for add and subtract, because that would result in double rounding for subnormal results, so that uses round-to-odd with long double intermediate result instead. The soft-fp support involves adding a new FP_TRUNC_COOKED since soft-fp multiplication uses cooked inputs and outputs. Tested for x86_64, x86, mips64 (all three ABIs, both hard and soft float) and powerpc, and with build-many-glibcs.py. * math/Makefile (libm-narrow-fns): Add mul. (libm-test-funcs-narrow): Likewise. * math/Versions (GLIBC_2.28): Add narrowing multiply functions. * math/bits/mathcalls-narrow.h (mul): Use __MATHCALL_NARROW. * math/gen-auto-libm-tests.c (test_functions): Add mul. * math/math-narrow.h (CHECK_NARROW_MUL): New macro. (NARROW_MUL_ROUND_TO_ODD): Likewise. (NARROW_MUL_TRIVIAL): Likewise. * soft-fp/op-common.h (FP_TRUNC_COOKED): Likewise. * sysdeps/ieee754/float128/float128_private.h (__fmull): New macro. (__dmull): Likewise. * sysdeps/ieee754/ldbl-opt/Makefile (libnldbl-calls): Add fmul and dmul. (CFLAGS-nldbl-dmul.c): New variable. (CFLAGS-nldbl-fmul.c): Likewise. * sysdeps/ieee754/ldbl-opt/Versions (GLIBC_2.28): Add __nldbl_dmull. * sysdeps/ieee754/ldbl-opt/nldbl-compat.h (__nldbl_dmull): New prototype. * manual/arith.texi (Misc FP Arithmetic): Document fmul, fmull, dmull, fMmulfN, fMmulfNx, fMxmulfN and fMxmulfNx. * math/auto-libm-test-in: Add tests of mul. * math/auto-libm-test-out-narrow-mul: New generated file. * math/libm-test-narrow-mul.inc: New file. * sysdeps/i386/fpu/s_f32xmulf64.c: Likewise. * sysdeps/ieee754/dbl-64/s_f32xmulf64.c: Likewise. * sysdeps/ieee754/dbl-64/s_fmul.c: Likewise. * sysdeps/ieee754/float128/s_f32mulf128.c: Likewise. * sysdeps/ieee754/float128/s_f64mulf128.c: Likewise. * sysdeps/ieee754/float128/s_f64xmulf128.c: Likewise. * sysdeps/ieee754/ldbl-128/s_dmull.c: Likewise. * sysdeps/ieee754/ldbl-128/s_f64xmulf128.c: Likewise. * sysdeps/ieee754/ldbl-128/s_fmull.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_dmull.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_fmull.c: Likewise. * sysdeps/ieee754/ldbl-96/s_dmull.c: Likewise. * sysdeps/ieee754/ldbl-96/s_fmull.c: Likewise. * sysdeps/ieee754/ldbl-opt/nldbl-dmul.c: Likewise. * sysdeps/ieee754/ldbl-opt/nldbl-fmul.c: Likewise. * sysdeps/ieee754/soft-fp/s_dmull.c: Likewise. * sysdeps/ieee754/soft-fp/s_fmul.c: Likewise. * sysdeps/ieee754/soft-fp/s_fmull.c: Likewise. * sysdeps/powerpc/fpu/libm-test-ulps: Update. * sysdeps/mach/hurd/i386/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/aarch64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/alpha/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/arm/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/hppa/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/i386/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/m68k/coldfire/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/m68k/m680x0/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/microblaze/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/mips/mips32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/mips/mips64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/nios2/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/fpu/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/nofpu/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/libm-le.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/riscv/rv64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/s390/s390-32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/s390/s390-64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sh/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/x86_64/64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/x86_64/x32/libm.abilist: Likewise.
2018-05-16 00:05:28 +00:00
mul 0 0
mul 0 -0
mul -0 0
mul -0 -0
mul max max
mul max -max
mul -max max
mul -max -max
mul min min
mul min -min
mul -min min
mul -min -min
mul min_subnorm min_subnorm
mul min_subnorm -min_subnorm
mul -min_subnorm min_subnorm
mul -min_subnorm -min_subnorm
mul 1 2
mul 1 -2
mul -1 2
mul -1 -2
mul 100.5 0.75
mul 100.5 -0.75
mul -100.5 0.75
mul -100.5 -0.75
# Cases where the product of two values in a wider format is very
# close to half way between two representable values in a narrower
# format, so that double rounding would sometimes yield the wrong
# result. These examples have products of the form
# (2^a + 1)(2^b + 1)/2^(a+b), where a is the width of the narrower
# format, b is greater than the width of the wider format and factors
# have been rearranged between the terms so that both terms do fit
# within the wider format.
# a = 24, b = 54
mul 0x5000005p-24 0xccccccccccccdp-54
# a = 24, b = 65
mul 0x3000003p-24 0xaaaaaaaaaaaaaaabp-65
# a = 24, b = 114
mul 0x5000005p-24 0xcccccccccccccccccccccccccccdp-114
# a = 53, b = 65
mul 0x60000000000003p-53 0xaaaaaaaaaaaaaaabp-65
# a = 53, b = 114
mul 0xa0000000000005p-53 0xcccccccccccccccccccccccccccdp-114
# a = 64, b = 114
mul 0x50000000000000005p-64 0xcccccccccccccccccccccccccccdp-114
# Similar, for double rounding to 64 bit of a product of 53-bit values.
# This product equals 2^64 + 2^11 + 1.
mul 97689974585 188829449
Fix f64xdivf128, f64xmulf128 spurious underflows (bug 28358) As described in bug 28358, the round-to-odd computations used in the libm functions that round their results to a narrower format can yield spurious underflow exceptions in the following circumstances: the narrowing only narrows the precision of the type and not the exponent range (i.e., it's narrowing _Float128 to _Float64x on x86_64, x86 or ia64), the architecture does after-rounding tininess detection (which applies to all those architectures), the result is inexact, tiny before rounding but not tiny after rounding (with the chosen rounding mode) for _Float64x (which is possible for narrowing mul, div and fma, not for narrowing add, sub or sqrt), so the underflow exception resulting from the toward-zero computation in _Float128 is spurious for _Float64x. Fixed by making ROUND_TO_ODD call feclearexcept (FE_UNDERFLOW) in the problem cases (as indicated by an extra argument to the macro); there is never any need to preserve underflow exceptions from this part of the computation, because the conversion of the round-to-odd value to the narrower type will underflow in exactly the cases in which the function should raise that exception, but it may be more efficient to avoid the extra manipulation of the floating-point environment when not needed. Tested for x86_64 and x86, and with build-many-glibcs.py.
2021-09-21 21:54:37 +00:00
# Cases where there is underflow before rounding (for some format) but
# might not be after rounding, depending on the rounding mode.
mul 0x0.ffffff8p-126 0x1.0000001p0
mul 0x0.ffffff8p-126 -0x1.0000001p0
mul -0x0.ffffff8p-126 0x1.0000001p0
mul -0x0.ffffff8p-126 -0x1.0000001p0
Adjust new narrowing div/mul tests for IBM long double, update powerpc ULPs Testing for powerpc shows some of the new narrowing div/mul tests need XFAILing for IBM long double and some ULPs updates are needed for those tests.
2021-09-22 12:35:44 +00:00
mul 0x0.fffffffffffffcp-1022 0x1.00000000000001p0 missing-underflow:arg-ibm128
mul 0x0.fffffffffffffcp-1022 -0x1.00000000000001p0 missing-underflow:arg-ibm128
mul -0x0.fffffffffffffcp-1022 0x1.00000000000001p0 missing-underflow:arg-ibm128
mul -0x0.fffffffffffffcp-1022 -0x1.00000000000001p0 missing-underflow:arg-ibm128
mul 0x0.ffffffffffffffff8p-16382 0x1.00000000000000001p0 missing-underflow:arg-ibm128
mul 0x0.ffffffffffffffff8p-16382 -0x1.00000000000000001p0 missing-underflow:arg-ibm128
mul -0x0.ffffffffffffffff8p-16382 0x1.00000000000000001p0 missing-underflow:arg-ibm128
mul -0x0.ffffffffffffffff8p-16382 -0x1.00000000000000001p0 missing-underflow:arg-ibm128
Fix f64xdivf128, f64xmulf128 spurious underflows (bug 28358) As described in bug 28358, the round-to-odd computations used in the libm functions that round their results to a narrower format can yield spurious underflow exceptions in the following circumstances: the narrowing only narrows the precision of the type and not the exponent range (i.e., it's narrowing _Float128 to _Float64x on x86_64, x86 or ia64), the architecture does after-rounding tininess detection (which applies to all those architectures), the result is inexact, tiny before rounding but not tiny after rounding (with the chosen rounding mode) for _Float64x (which is possible for narrowing mul, div and fma, not for narrowing add, sub or sqrt), so the underflow exception resulting from the toward-zero computation in _Float128 is spurious for _Float64x. Fixed by making ROUND_TO_ODD call feclearexcept (FE_UNDERFLOW) in the problem cases (as indicated by an extra argument to the macro); there is never any need to preserve underflow exceptions from this part of the computation, because the conversion of the round-to-odd value to the narrower type will underflow in exactly the cases in which the function should raise that exception, but it may be more efficient to avoid the extra manipulation of the floating-point environment when not needed. Tested for x86_64 and x86, and with build-many-glibcs.py.
2021-09-21 21:54:37 +00:00
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
pow 0 0
pow 0 -0
pow -0 0
pow -0 -0
pow 10 0
pow 10 -0
pow -10 0
pow -10 -0
pow 1 1
pow 1 -1
pow 1 1.25
pow 1 -1.25
pow 1 0x1p62
pow 1 0x1p63
pow 1 0x1p64
pow 1 0x1p72
pow 1 min_subnorm
pow 1 -min_subnorm
# pow (x, +-0) == 1.
pow 32.75 0
pow 32.75 -0
pow -32.75 0
pow -32.75 -0
pow 0x1p72 0
pow 0x1p72 -0
pow 0x1p-72 0
pow 0x1p-72 -0
pow 0x1p72 0x1p72
pow 10 -0x1p72
pow max max
pow 10 -max
pow 0 1
pow 0 11
pow -0 1
pow -0 11
pow 0 2
pow 0 11.1
pow -0 2
pow -0 11.1
# pow (+0, y) == +0 for y an odd integer > 0.
pow 0.0 27
pow 0.0 0xffffff
pow 0.0 0x1.fffffffffffffp+52
pow 0.0 0x1.fffffffffffffffep+63
pow 0.0 0x1.ffffffffffffffffffffffffff8p+105
pow 0.0 0x1.ffffffffffffffffffffffffffffp+112
# pow (-0, y) == -0 for y an odd integer > 0.
pow -0 27
pow -0 0xffffff
pow -0 0x1fffffe
pow -0 0x1.fffffffffffffp+52
pow -0 0x1.fffffffffffffp+53
pow -0 0x1.fffffffffffffffep+63
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
pow -0 0x1.fffffffffffffffep+64 xfail-rounding:ibm128-libgcc
pow -0 0x1.ffffffffffffffffffffffffff8p+105 xfail-rounding:ibm128-libgcc
pow -0 0x1.ffffffffffffffffffffffffff8p+106 xfail-rounding:ibm128-libgcc
pow -0 0x1.ffffffffffffffffffffffffffffp+112 xfail-rounding:ibm128-libgcc
pow -0 0x1.ffffffffffffffffffffffffffffp+113 xfail-rounding:ibm128-libgcc
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
# pow (+0, y) == +0 for y > 0 and not an odd integer.
pow 0.0 4
pow 0.0 0x1p24
pow 0.0 0x1p127
pow 0.0 max
pow 0.0 min_subnorm
# pow (-0, y) == +0 for y > 0 and not an odd integer.
Add more tests of pow. Prompted by a gcc-patches discussion, this patch adds tests of pow for the cases where pow (x, 0.5) is required to return a different result from sqrt (x), as those cases were previously missing from the tests (although they worked correctly). Tested for x86_64 and x86. * math/auto-libm-test-in: Add another test of pow. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (pow_test_data): Add another test.
2015-11-10 17:34:13 +00:00
pow -0 0.5
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
pow -0 4
pow -0 0x1p24
pow -0 0x1p127
pow -0 max
pow -0 min_subnorm
pow 16 0.25
pow 0x1p64 0.125
pow 2 4
pow 256 8
pow 0.75 1.25
pow -7.49321e+133 -9.80818e+16
pow -1.0 -0xffffff
pow -1.0 -0x1fffffe
pow -1.0 -0x1.fffffffffffffp+52
pow -1.0 -0x1.fffffffffffffp+53
pow -1.0 -0x1.fffffffffffffffep+63
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
pow -1.0 -0x1.fffffffffffffffep+64 xfail-rounding:ibm128-libgcc
pow -1.0 -0x1.ffffffffffffffffffffffffff8p+105 xfail-rounding:ibm128-libgcc
pow -1.0 -0x1.ffffffffffffffffffffffffff8p+106 xfail-rounding:ibm128-libgcc
pow -1.0 -0x1.ffffffffffffffffffffffffffffp+112 xfail-rounding:ibm128-libgcc
pow -1.0 -0x1.ffffffffffffffffffffffffffffp+113 xfail-rounding:ibm128-libgcc
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
pow -1.0 -max
pow -1.0 0xffffff
pow -1.0 0x1fffffe
pow -1.0 0x1.fffffffffffffp+52
pow -1.0 0x1.fffffffffffffp+53
pow -1.0 0x1.fffffffffffffffep+63
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
pow -1.0 0x1.fffffffffffffffep+64 xfail-rounding:ibm128-libgcc
pow -1.0 0x1.ffffffffffffffffffffffffff8p+105 xfail-rounding:ibm128-libgcc
pow -1.0 0x1.ffffffffffffffffffffffffff8p+106 xfail-rounding:ibm128-libgcc
pow -1.0 0x1.ffffffffffffffffffffffffffffp+112 xfail-rounding:ibm128-libgcc
pow -1.0 0x1.ffffffffffffffffffffffffffffp+113 xfail-rounding:ibm128-libgcc
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
pow -1.0 max
pow -2.0 126
pow -2.0 127
pow -2.0 -126
pow -2.0 -127
pow -2.0 -0xffffff
pow -2.0 -0x1fffffe
pow -2.0 -0x1.fffffffffffffp+52
pow -2.0 -0x1.fffffffffffffp+53
pow -2.0 -0x1.fffffffffffffffep+63
pow -2.0 -0x1.fffffffffffffffep+64
pow -2.0 -0x1.ffffffffffffffffffffffffff8p+105
pow -2.0 -0x1.ffffffffffffffffffffffffff8p+106
pow -2.0 -0x1.ffffffffffffffffffffffffffffp+112
pow -2.0 -0x1.ffffffffffffffffffffffffffffp+113
pow -2.0 -max
pow -2.0 0xffffff
pow -2.0 0x1fffffe
pow -2.0 0x1.fffffffffffffp+52
pow -2.0 0x1.fffffffffffffp+53
pow -2.0 0x1.fffffffffffffffep+63
pow -2.0 0x1.fffffffffffffffep+64
pow -2.0 0x1.ffffffffffffffffffffffffff8p+105
pow -2.0 0x1.ffffffffffffffffffffffffff8p+106
pow -2.0 0x1.ffffffffffffffffffffffffffffp+112
pow -2.0 0x1.ffffffffffffffffffffffffffffp+113
pow -2.0 max
pow -max -2
pow -max -3
pow -max 2
pow -max 3
pow -max -0xffffff
pow -max -0x1fffffe
pow -max -0x1.fffffffffffffp+52
pow -max -0x1.fffffffffffffp+53
pow -max -0x1.fffffffffffffffep+63
pow -max -0x1.fffffffffffffffep+64
pow -max -0x1.ffffffffffffffffffffffffff8p+105
pow -max -0x1.ffffffffffffffffffffffffff8p+106
pow -max -0x1.ffffffffffffffffffffffffffffp+112
pow -max -0x1.ffffffffffffffffffffffffffffp+113
pow -max -max
pow -max 0xffffff
pow -max 0x1fffffe
pow -max 0x1.fffffffffffffp+52
pow -max 0x1.fffffffffffffp+53
pow -max 0x1.fffffffffffffffep+63
pow -max 0x1.fffffffffffffffep+64
pow -max 0x1.ffffffffffffffffffffffffff8p+105
pow -max 0x1.ffffffffffffffffffffffffff8p+106
pow -max 0x1.ffffffffffffffffffffffffffffp+112
pow -max 0x1.ffffffffffffffffffffffffffffp+113
pow -max max
Fix pow overflow in non-default rounding modes (bug 16315). This patch fixes bug 16315, bad pow handling of overflow/underflow in non-default rounding modes. Tests of pow are duly converted to ALL_RM_TEST to run all tests in all rounding modes. There are two main issues here. First, various implementations compute a negative result by negating a positive result, but this yields inappropriate overflow / underflow values for directed rounding, so either overflow / underflow results need recomputing in the correct sign, or the relevant overflowing / underflowing operation needs to be made to have a result of the correct sign. Second, the dbl-64 implementation sets FE_TONEAREST internally; in the overflow / underflow case, the result needs recomputing in the original rounding mode. Tested x86_64 and x86 and ulps updated accordingly. [BZ #16315] * sysdeps/i386/fpu/e_pow.S (__ieee754_pow): Ensure possibly overflowing or underflowing operations take place with sign of result. * sysdeps/i386/fpu/e_powf.S (__ieee754_powf): Likewise. * sysdeps/i386/fpu/e_powl.S (__ieee754_powl): Likewise. * sysdeps/ieee754/dbl-64/e_pow.c: Include <math.h>. (__ieee754_pow): Recompute overflowing and underflowing results in original rounding mode. * sysdeps/x86/fpu/powl_helper.c: Include <stdbool.h>. (__powl_helper): Allow negative argument X and scale negated value as needed. Avoid passing value outside [-1, 1] to f2xm1. * sysdeps/x86_64/fpu/e_powl.S (__ieee754_powl): Ensure possibly overflowing or underflowing operations take place with sign of result. * sysdeps/x86_64/fpu/multiarch/e_pow.c [HAVE_FMA4_SUPPORT]: Include <math.h>. * math/auto-libm-test-in: Add more tests of pow. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (pow_test): Use ALL_RM_TEST. (pow_tonearest_test_data): Remove. (pow_test_tonearest): Likewise. (pow_towardzero_test_data): Likewise. (pow_test_towardzero): Likewise. (pow_downward_test_data): Likewise. (pow_test_downward): Likewise. (pow_upward_test_data): Likewise. (pow_test_upward): Likewise. (main): Don't call removed functions. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2014-06-23 20:12:33 +00:00
pow -0x1p65 2
pow -0x1p65 3
pow -0x1p65 4
pow -0x1p65 5
pow -0x1p43 3
pow -0x1p43 4
pow -0x1p43 5
pow -0x1p33 4
pow -0x1p33 5
pow -0x1p26 5
pow -0x1p-65 -2
pow -0x1p-65 -3
pow -0x1p-65 -4
pow -0x1p-65 -5
pow -0x1p-43 -3
pow -0x1p-43 -4
pow -0x1p-43 -5
pow -0x1p-33 -4
pow -0x1p-33 -5
pow -0x1p-26 -5
pow -0x1p513 2
pow -0x1p513 3
pow -0x1p513 4
pow -0x1p513 5
pow -0x1p342 3
pow -0x1p342 4
pow -0x1p342 5
pow -0x1p257 4
pow -0x1p257 5
pow -0x1p205 5
pow -0x1p-513 -2
pow -0x1p-513 -3
pow -0x1p-513 -4
pow -0x1p-513 -5
pow -0x1p-342 -3
pow -0x1p-342 -4
pow -0x1p-342 -5
pow -0x1p-257 -4
pow -0x1p-257 -5
pow -0x1p-205 -5
pow -0x1p8192 2
pow -0x1p8192 3
pow -0x1p8192 4
pow -0x1p8192 5
pow -0x1p5462 3
pow -0x1p5462 4
pow -0x1p5462 5
pow -0x1p4097 4
pow -0x1p4097 5
pow -0x1p3277 5
pow -0x1p64 257
pow -0x1p-8192 -2
pow -0x1p-8192 -3
pow -0x1p-8192 -4
pow -0x1p-8192 -5
pow -0x1p-5462 -3
pow -0x1p-5462 -4
pow -0x1p-5462 -5
pow -0x1p-4097 -4
pow -0x1p-4097 -5
pow -0x1p-3277 -5
pow -0x1p-64 -257
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
pow -0.5 126
pow -0.5 127
pow -0.5 -126
pow -0.5 -127
pow -0.5 -0xffffff
pow -0.5 -0x1fffffe
pow -0.5 -0x1.fffffffffffffp+52
pow -0.5 -0x1.fffffffffffffp+53
pow -0.5 -0x1.fffffffffffffffep+63
pow -0.5 -0x1.fffffffffffffffep+64
pow -0.5 -0x1.ffffffffffffffffffffffffff8p+105
pow -0.5 -0x1.ffffffffffffffffffffffffff8p+106
pow -0.5 -0x1.ffffffffffffffffffffffffffffp+112
pow -0.5 -0x1.ffffffffffffffffffffffffffffp+113
pow -0.5 -max
pow -0.5 0xffffff
pow -0.5 0x1fffffe
pow -0.5 0x1.fffffffffffffp+52
pow -0.5 0x1.fffffffffffffp+53
pow -0.5 0x1.fffffffffffffffep+63
pow -0.5 0x1.fffffffffffffffep+64
pow -0.5 0x1.ffffffffffffffffffffffffff8p+105
pow -0.5 0x1.ffffffffffffffffffffffffff8p+106
pow -0.5 0x1.ffffffffffffffffffffffffffffp+112
pow -0.5 0x1.ffffffffffffffffffffffffffffp+113
pow -0.5 max
pow -min -2
pow -min -3
pow -min 1
pow -min 2
pow -min 3
pow -min -0xffffff
pow -min -0x1fffffe
pow -min -0x1.fffffffffffffp+52
pow -min -0x1.fffffffffffffp+53
pow -min -0x1.fffffffffffffffep+63
pow -min -0x1.fffffffffffffffep+64
pow -min -0x1.ffffffffffffffffffffffffff8p+105
pow -min -0x1.ffffffffffffffffffffffffff8p+106
pow -min -0x1.ffffffffffffffffffffffffffffp+112
pow -min -0x1.ffffffffffffffffffffffffffffp+113
pow -min -max
pow -min 0xffffff
pow -min 0x1fffffe
pow -min 0x1.fffffffffffffp+52
pow -min 0x1.fffffffffffffp+53
pow -min 0x1.fffffffffffffffep+63
pow -min 0x1.fffffffffffffffep+64
pow -min 0x1.ffffffffffffffffffffffffff8p+105
pow -min 0x1.ffffffffffffffffffffffffff8p+106
pow -min 0x1.ffffffffffffffffffffffffffffp+112
pow -min 0x1.ffffffffffffffffffffffffffffp+113
pow -min max
pow 0x0.ffffffp0 10
pow 0x0.ffffffp0 100
pow 0x0.ffffffp0 1000
pow 0x0.ffffffp0 0x1p24
pow 0x0.ffffffp0 0x1p30
pow 0x0.ffffffp0 0x1.234566p30
pow 0x0.ffffffp0 -10
pow 0x0.ffffffp0 -100
pow 0x0.ffffffp0 -1000
pow 0x0.ffffffp0 -0x1p24
pow 0x0.ffffffp0 -0x1p30
pow 0x0.ffffffp0 -0x1.234566p30
pow 0x1.000002p0 0x1p24
pow 0x1.000002p0 0x1.234566p29
pow 0x1.000002p0 -0x1.234566p29
pow 0x0.fffffffffffff8p0 0x1.23456789abcdfp62
pow 0x0.fffffffffffff8p0 -0x1.23456789abcdfp62
pow 0x1.0000000000001p0 0x1.23456789abcdfp61
pow 0x1.0000000000001p0 -0x1.23456789abcdfp61
pow 0x0.ffffffffffffffffp0 0x1.23456789abcdef0ep77
pow 0x0.ffffffffffffffffp0 -0x1.23456789abcdef0ep77
pow 0x1.0000000000000002p0 0x1.23456789abcdef0ep76
pow 0x1.0000000000000002p0 -0x1.23456789abcdef0ep76
pow 0x0.ffffffffffffffffffffffffffff8p0 0x1.23456789abcdef0123456789abcdp126
pow 0x0.ffffffffffffffffffffffffffff8p0 -0x1.23456789abcdef0123456789abcdp126
pow 0x1.0000000000000000000000000001p0 0x1.23456789abcdef0123456789abcdp125
pow 0x1.0000000000000000000000000001p0 -0x1.23456789abcdef0123456789abcdp125
Fix powf (close to -1, large) (bug 18647). The flt-32 implementation of powf wrongly uses x-1 instead of |x|-1 when computing log (x) for the case where |x| is close to 1 and y is large. This patch fixes the logic accordingly. Relevant tests existed for x close to 1, and corresponding tests are added for x close to -1, as well as for some new variant cases. Tested for x86_64 and x86. [BZ #18647] * sysdeps/ieee754/flt-32/e_powf.c (__ieee754_powf): For large y and |x| close to 1, use absolute value of x when computing log. * math/auto-libm-test-in: Add more tests of pow. * math/auto-libm-test-out: Regenerated.
2015-08-05 15:01:58 +00:00
pow -0x0.ffffffp0 10
pow -0x0.ffffffp0 100
pow -0x0.ffffffp0 1000
pow -0x0.ffffffp0 0x1p24
pow -0x0.ffffffp0 0x1p30
pow -0x0.ffffffp0 0x1.234566p30
pow -0x0.ffffffp0 -10
pow -0x0.ffffffp0 -100
pow -0x0.ffffffp0 -1000
pow -0x0.ffffffp0 -0x1p24
pow -0x0.ffffffp0 -0x1p30
pow -0x0.ffffffp0 -0x1.234566p30
pow -0x1.000002p0 0x1p24
pow -0x1.000002p0 0x1.234566p29
pow -0x1.000002p0 -0x1.234566p29
pow -0x0.fffffffffffff8p0 0x1.23456789abcdfp62
pow -0x0.fffffffffffff8p0 -0x1.23456789abcdfp62
pow -0x1.0000000000001p0 0x1.23456789abcdfp61
pow -0x1.0000000000001p0 -0x1.23456789abcdfp61
pow -0x0.ffffffffffffffffp0 0x1.23456789abcdef0ep77
pow -0x0.ffffffffffffffffp0 -0x1.23456789abcdef0ep77
pow -0x1.0000000000000002p0 0x1.23456789abcdef0ep76
pow -0x1.0000000000000002p0 -0x1.23456789abcdef0ep76
pow -0x0.ffffffffffffffffffffffffffff8p0 0x1.23456789abcdef0123456789abcdp126
pow -0x0.ffffffffffffffffffffffffffff8p0 -0x1.23456789abcdef0123456789abcdp126
pow -0x1.0000000000000000000000000001p0 0x1.23456789abcdef0123456789abcdp125
pow -0x1.0000000000000000000000000001p0 -0x1.23456789abcdef0123456789abcdp125
pow 0x1.000002p0 0x1p30
pow -0x1.000002p0 0x1p30
pow 0x1.000002p0 max
pow -0x1.000002p0 max
pow 0x1.00000ep0 0x1p30
pow -0x1.00000ep0 0x1p30
pow 0x1.00000ep0 max
pow -0x1.00000ep0 max
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
pow 1e4932 0.75
pow 1e4928 0.75
pow 1e4924 0.75
pow 1e4920 0.75
pow 10.0 4932.0
pow 10.0 4931.0
pow 10.0 4930.0
pow 10.0 4929.0
pow 10.0 -4931.0
pow 10.0 -4930.0
pow 10.0 -4929.0
pow 1e27 182.0
pow 1e27 -182.0
pow min_subnorm min_subnorm
pow min_subnorm -min_subnorm
pow max min_subnorm
pow max -min_subnorm
pow 0.99 min_subnorm
pow 0.99 -min_subnorm
pow 1.01 min_subnorm
pow 1.01 -min_subnorm
pow 2.0 -100000.0
Fix pow missing underflows (bug 18825). Similar to various other bugs in this area, pow functions can fail to raise the underflow exception when the result is tiny and inexact but one or more low bits of the intermediate result that is scaled down (or, in the i386 case, converted from a wider evaluation format) are zero. This patch forces the exception in a similar way to previous fixes, thereby concluding the fixes for known bugs with missing underflow exceptions currently filed in Bugzilla. Tested for x86_64, x86, mips64 and powerpc. [BZ #18825] * sysdeps/i386/fpu/i386-math-asm.h (FLT_NARROW_EVAL_UFLOW_NONNAN): New macro. (DBL_NARROW_EVAL_UFLOW_NONNAN): Likewise. (LDBL_CHECK_FORCE_UFLOW_NONNAN): Likewise. * sysdeps/i386/fpu/e_pow.S: Use DEFINE_DBL_MIN. (__ieee754_pow): Use DBL_NARROW_EVAL_UFLOW_NONNAN instead of DBL_NARROW_EVAL, reloading the PIC register as needed. * sysdeps/i386/fpu/e_powf.S: Use DEFINE_FLT_MIN. (__ieee754_powf): Use FLT_NARROW_EVAL_UFLOW_NONNAN instead of FLT_NARROW_EVAL. Use separate return path for case when first argument is NaN. * sysdeps/i386/fpu/e_powl.S: Include <i386-math-asm.h>. Use DEFINE_LDBL_MIN. (__ieee754_powl): Use LDBL_CHECK_FORCE_UFLOW_NONNAN, reloading the PIC register. * sysdeps/ieee754/dbl-64/e_pow.c (__ieee754_pow): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/flt-32/e_powf.c (__ieee754_powf): Force underflow for subnormal result. * sysdeps/ieee754/ldbl-128/e_powl.c (__ieee754_powl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_powl.c (__ieee754_powl): Use math_check_force_underflow_nonneg. * sysdeps/x86/fpu/powl_helper.c (__powl_helper): Use math_check_force_underflow. * sysdeps/x86_64/fpu/x86_64-math-asm.h (LDBL_CHECK_FORCE_UFLOW_NONNAN): New macro. * sysdeps/x86_64/fpu/e_powl.S: Include <x86_64-math-asm.h>. Use DEFINE_LDBL_MIN. (__ieee754_powl): Use LDBL_CHECK_FORCE_UFLOW_NONNAN. * math/auto-libm-test-in: Add more tests of pow. * math/auto-libm-test-out: Regenerated.
2015-09-25 22:29:10 +00:00
pow 2 -126.125
pow 2 -126.25
pow 2 -126.375
pow 2 -126.5
pow 2 -126.625
pow 2 -126.75
pow 2 -126.875
pow 2 -969.125
pow 2 -969.25
pow 2 -969.375
pow 2 -969.5
pow 2 -969.625
pow 2 -969.75
pow 2 -969.875
pow 2 -1022.125
pow 2 -1022.25
pow 2 -1022.375
pow 2 -1022.5
pow 2 -1022.625
pow 2 -1022.75
pow 2 -1022.875
pow 2 -16382.125
pow 2 -16382.25
pow 2 -16382.375
pow 2 -16382.5
pow 2 -16382.625
pow 2 -16382.75
pow 2 -16382.875
pow 2 -16383.125
pow 2 -16383.25
pow 2 -16383.375
pow 2 -16383.5
pow 2 -16383.625
pow 2 -16383.75
pow 2 -16383.875
pow 0.5 126.125
pow 0.5 126.25
pow 0.5 126.375
pow 0.5 126.5
pow 0.5 126.625
pow 0.5 126.75
pow 0.5 126.875
pow 0.5 969.125
pow 0.5 969.25
pow 0.5 969.375
pow 0.5 969.5
pow 0.5 969.625
pow 0.5 969.75
pow 0.5 969.875
pow 0.5 1022.125
pow 0.5 1022.25
pow 0.5 1022.375
pow 0.5 1022.5
pow 0.5 1022.625
pow 0.5 1022.75
pow 0.5 1022.875
pow 0.5 16382.125
pow 0.5 16382.25
pow 0.5 16382.375
pow 0.5 16382.5
pow 0.5 16382.625
pow 0.5 16382.75
pow 0.5 16382.875
pow 0.5 16383.125
pow 0.5 16383.25
pow 0.5 16383.375
pow 0.5 16383.5
pow 0.5 16383.625
pow 0.5 16383.75
pow 0.5 16383.875
pow 0x1.00001p1 -126
pow -0x1.00002p1 -126
pow 0x1.00003p1 -126
pow -0x1.00004p1 -126
pow 0x1.00005p1 -126
pow -0x1.00006p1 -126
pow 0x1.00007p1 -126
pow 0x1.00001p1 -969
pow -0x1.00002p1 -969
pow 0x1.00003p1 -969
pow -0x1.00004p1 -969
pow 0x1.00005p1 -969
pow -0x1.00006p1 -969
pow 0x1.00007p1 -969
pow 0x1.00001p1 -1022
pow -0x1.00002p1 -1022
pow 0x1.00003p1 -1022
pow -0x1.00004p1 -1022
pow 0x1.00005p1 -1022
pow -0x1.00006p1 -1022
pow 0x1.00007p1 -1022
pow 0x1.00001p1 -16382
pow -0x1.00002p1 -16382
pow 0x1.00003p1 -16382
pow -0x1.00004p1 -16382
pow 0x1.00005p1 -16382
pow -0x1.00006p1 -16382
pow 0x1.00007p1 -16382
pow 0x1.00001p1 -16383
pow -0x1.00002p1 -16383
pow 0x1.00003p1 -16383
pow -0x1.00004p1 -16383
pow 0x1.00005p1 -16383
pow -0x1.00006p1 -16383
pow 0x1.00007p1 -16383
pow 0x0.ffff1p-1 126
pow -0x0.ffff2p-1 126
pow 0x0.ffff3p-1 126
pow -0x0.ffff4p-1 126
pow 0x0.ffff5p-1 126
pow -0x0.ffff6p-1 126
pow 0x0.ffff7p-1 126
pow 0x0.ffff1p-1 969
pow -0x0.ffff2p-1 969
pow 0x0.ffff3p-1 969
pow -0x0.ffff4p-1 969
pow 0x0.ffff5p-1 969
pow -0x0.ffff6p-1 969
pow 0x0.ffff7p-1 969
pow 0x0.ffff1p-1 1022
pow -0x0.ffff2p-1 1022
pow 0x0.ffff3p-1 1022
pow -0x0.ffff4p-1 1022
pow 0x0.ffff5p-1 1022
pow -0x0.ffff6p-1 1022
pow 0x0.ffff7p-1 1022
pow 0x0.ffff1p-1 16382
pow -0x0.ffff2p-1 16382
pow 0x0.ffff3p-1 16382
pow -0x0.ffff4p-1 16382
pow 0x0.ffff5p-1 16382
pow -0x0.ffff6p-1 16382
pow 0x0.ffff7p-1 16382
pow 0x0.ffff1p-1 16383
pow -0x0.ffff2p-1 16383
pow 0x0.ffff3p-1 16383
pow -0x0.ffff4p-1 16383
pow 0x0.ffff5p-1 16383
pow -0x0.ffff6p-1 16383
pow 0x0.ffff7p-1 16383
pow 0x2.000b3p0 -1022
pow 0x2.000582p0 -1022
pow 2 -0x3.fe513p+8
pow 2 -0x3.fe4e8p+8
Fix x86_64 / x86 powl inaccuracy for integer exponents (bug 19848). Bug 19848 reports cases where powl on x86 / x86_64 has error accumulation, for small integer exponents, larger than permitted by glibc's accuracy goals, at least in some rounding modes. This patch further restricts the exponent range for which the small-integer-exponent logic is used to limit the possible error accumulation. Tested for x86_64 and x86 and ulps updated accordingly. [BZ #19848] * sysdeps/i386/fpu/e_powl.S (p3): Rename to p2 and change value from 8 to 4. (__ieee754_powl): Compare integer exponent against 4 not 8. * sysdeps/x86_64/fpu/e_powl.S (p3): Rename to p2 and change value from 8 to 4. (__ieee754_powl): Compare integer exponent against 4 not 8. * math/auto-libm-test-in: Add more tests of pow. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2016-03-24 01:32:52 +00:00
pow 10 -1
pow 10 -2
pow 10 -3
pow 10 -4
pow 10 -5
pow 10 -6
pow 10 -7
pow 0x0.ffffffffffffffffp0 1
pow 0x0.ffffffffffffffffp0 2
pow 0x0.ffffffffffffffffp0 3
pow 0x0.ffffffffffffffffp0 4
pow 0x0.ffffffffffffffffp0 5
pow 0x0.ffffffffffffffffp0 6
pow 0x0.ffffffffffffffffp0 7
pow 0x0.ffffffffffffffffp0 -1
pow 0x0.ffffffffffffffffp0 -2
pow 0x0.ffffffffffffffffp0 -3
pow 0x0.ffffffffffffffffp0 -4
pow 0x0.ffffffffffffffffp0 -5
pow 0x0.ffffffffffffffffp0 -6
pow 0x0.ffffffffffffffffp0 -7
pow 0x1.0000000000000002p0 1
pow 0x1.0000000000000002p0 2
pow 0x1.0000000000000002p0 3
pow 0x1.0000000000000002p0 4
pow 0x1.0000000000000002p0 5
pow 0x1.0000000000000002p0 6
pow 0x1.0000000000000002p0 7
pow 0x1.0000000000000002p0 -1
pow 0x1.0000000000000002p0 -2
pow 0x1.0000000000000002p0 -3
pow 0x1.0000000000000002p0 -4
pow 0x1.0000000000000002p0 -5
pow 0x1.0000000000000002p0 -6
pow 0x1.0000000000000002p0 -7
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
pow 1.0625 1.125
pow 1.5 1.03125
Add test for bug 18104. [BZ #18104] * math/auto-libm-test-in: Add another test of pow. * math/auto-libm-test-out: Regenerated.
2015-03-10 17:53:40 +00:00
pow 0x1.7d1a0a6f2p+681 1.5
Add another test of pow. This patch adds a randomly-generated test of pow that is observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add another test of pow. * math/auto-libm-test-out: Regenerated. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-05-01 22:31:24 +00:00
pow 0x1.ce78f2p+0 -0x2.7f1f78p+4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
pow 0xf.fffffp+124 -0x5.b5b648p+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
pow 0x1.430d4cp+0 0x5.0e462p+4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
pow 0x9.8b82ap-4 -0x1.99907ap+12
Fix powf inaccuracy (bug 18956). The flt-32 version of powf can be inaccurate because of bugs in the extra-precision calculation of (x-1)/(x+1) or (x-1.5)/(x+1.5) as part of calculating log(x) with extra precision: a constant used (as part of adding 1 or 1.5 through integer arithmetic) is incorrect, and then the code fails to mask a computed high part before using it in arithmetic that relies on s_h*t_h being exactly representable. This patch fixes these bugs. Tested for x86_64 and x86. x86_64 ulps for powf removed and regenerated to reflect reduced ulps from the increased accuracy for existing tests. [BZ #18956] * sysdeps/ieee754/flt-32/e_powf.c (__ieee754_powf): Add 0x00400000 not 0x0040000 for high bit of mantissa. Mask with 0xfffff000 when extracting high part. * math/auto-libm-test-in: Add another test of pow. * math/auto-libm-test-out: Regenerated. * sysdeps/x86_64/fpu/libm-test-ulps: Update.
2015-09-26 00:27:06 +00:00
pow 0xd.73035p-4 -0x1.47bb8p+8
Fix powf inaccuracy (bug 21112). Bug 21112 reports a case where powf is substantially inaccurate. This results from a multiplication where cp_h*p_h is required to be exact, and p_h is masked to have only 12 leading nonzero bits in its mantissa, but the value of cp_h has the 13th bit nonzero, leading to inexact multiplication results in some cases that can result in large errors in the final result of powf. This patch fixes this by using a value of cp_h correctly rounded to nearest to 12 bits, with a corresponding updated value of cp_l. Tested for x86_64 and x86. [BZ #21112] * sysdeps/ieee754/flt-32/e_powf.c (cp_h): Use value with trailing 12 bits zero. (cp_l): Update for new value of cp_h. * math/auto-libm-test-in: Add another test of pow. * math/auto-libm-test-out-pow: Regenerated.
2017-02-07 17:15:47 +00:00
pow 0x1.059c76p+0 0x1.ff80bep+11
Fix powf overflow handling in non-nearest rounding mode [BZ #23961] The threshold value at which powf overflows depends on the rounding mode and the current check did not take this into account. So when the result was rounded away from zero it could become infinity without setting errno to ERANGE. Example: pow(0x1.7ac7cp+5, 23) is 0x1.fffffep+127 + 0.1633ulp If the result goes above 0x1.fffffep+127 + 0.5ulp then errno is set, which is fine in nearest rounding mode, but powf(0x1.7ac7cp+5, 23) is inf in upward rounding mode powf(-0x1.7ac7cp+5, 23) is -inf in downward rounding mode and the previous implementation did not set errno in these cases. The fix tries to avoid affecting the common code path or calling a function that may introduce a stack frame, so float arithmetics is used to check the rounding mode and the threshold is selected accordingly. [BZ #23961] * math/auto-libm-test-in: Add new test case. * math/auto-libm-test-out-pow: Regenerated. * sysdeps/ieee754/flt-32/e_powf.c (__powf): Fix overflow check.
2018-12-10 11:08:36 +00:00
pow 0x1.7ac7cp+5 23
pow -0x1.7ac7cp+5 23
Move tests of atan2, hypot and pow from libm-test.inc to auto-libm-test-in.
2013-12-16 21:18:07 +00:00
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
sin 0
sin -0
sin pi/6
sin -pi/6
sin pi/2
sin -pi/2
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
sin pi
sin -pi
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
sin 0.75
sin 0x1p65
sin -0x1p65
sin 0x1.7f4134p+103
sin 0.80190127184058835
sin 2.522464e-1
sin 1e22
sin 0x1p1023
sin 0x1p16383
sin 0x1p+120
sin 0x1p+127
sin 0x1.fffff8p+127
sin 0x1.fffffep+127
sin 0x1p+50
sin 0x1p+28
sin 0.93340582292648832662962377071381
sin 2.3328432680770916363144351635128
sin 3.7439477503636453548097051680088
sin 3.9225160069792437411706487182528
sin 4.0711651639931289992091478779912
sin 4.7858438478542097982426639646292
sin 5.9840767662578002727968851104379
sin 1
sin 2
sin 3
sin 4
sin 5
sin 6
sin 7
sin 8
sin 9
sin 10
Fix sign of input to bsloww1 (BZ #16623) In 84ba214c, I removed some redundant sign computations and in the process, I incorrectly got rid of a temporary variable, thus passing the absolute value of the input to bsloww1. This caused #16623. This fix undoes the incorrect change.
2014-02-27 15:42:09 +00:00
sin 0x1.2001469775ce6p32
Add more tests of cos, sin, sincos. This patch adds some randomly-generated tests of cos, sin and sincos that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cos, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 22:41:00 +00:00
sin -0x3.3de320f6be87ep+1020
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
sin 0xe.9f1e5bc3bb88p+112
Add more tests of acosh, atanh, cos, csqrt, erfc, sin, sincos. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, cos, csqrt, erfc, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-06 17:30:18 +00:00
sin 0x4.7857dp+68
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
sin 0x6.287cc8749212e72p+0
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
sin -0x1.02e34cp+0
sin 0xf.f0274p+4
sin 0x3.042d88p+0
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
sin 0x1.d12ed2p-12
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (ldbl-96)
sin -0x6.e2368c006c018228p+16
# the next value generates larger error bounds on x86_64 (binary128)
sin 0x5.6a5005df4363833413fa44f74ae8p+64
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
sin max
sin -max
Fix sin, sincos missing underflows (bug 16526, bug 16538). Similar to various other bugs in this area, some sin and sincos implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86, mips64 and powerpc. [BZ #16526] [BZ #16538] * sysdeps/ieee754/dbl-64/s_sin.c: Include <float.h>. (__sin): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/k_sinf.c: Include <float.h>. (__kernel_sinf): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/k_sincosl.c: Include <float.h>. (__kernel_sincosl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/k_sinl.c: Include <float.h>. (__kernel_sinl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/k_sincosl.c: Include <float.h>. (__kernel_sincosl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/k_sinl.c: Include <float.h>. (__kernel_sinl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-96/k_sinl.c: Include <float.h>. (__kernel_sinl): Force underflow exception for arguments with small absolute value. * sysdeps/powerpc/fpu/k_sinf.c: Include <float.h>. (__kernel_sinf): Force underflow exception for arguments with small absolute value. * math/auto-libm-test-in: Add more tests of sin and sincos. * math/auto-libm-test-out: Regenerated.
2015-06-23 22:24:20 +00:00
sin min
sin -min
sin min_subnorm
sin -min_subnorm
Fix cos computation for multiple precision fallback (bz #20357) During the sincos consolidation I made two mistakes, one was a logical error due to which cos(0x1.8475e5afd4481p+0) returned sin(0x1.8475e5afd4481p+0) instead. The second issue was an error in negating inputs for the correct quadrants for sine. I could not find a suitable test case for this despite running a program to search for such an input for a couple of hours. Following patch fixes both issues. Tested on x86_64. Thanks to Matt Clay for identifying the issue. [BZ #20357] * sysdeps/ieee754/dbl-64/s_sin.c (sloww): Fix up condition to call __mpsin/__mpcos and to negate values. * math/auto-libm-test-in: Add test. * math/auto-libm-test-out: Regenerate.
2016-07-18 17:03:09 +00:00
sin 0x1.8475e5afd4481p+0
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
Move tests of sincos from libm-test.inc to auto-libm-test-in. This patch moves tests of sincos to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86 and ulps updated accordingly. (auto-libm-test-out diffs omitted below.) * math/auto-libm-test-in: Add tests of sincos. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (sincos_test_data): Use AUTO_TESTS_fFF_11. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_f_11. (func_calc_desc): Add mpfr_f_11 union field. (test_functions): Add sincos. (calc_generic_results): Handle mpfr_f_11. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-19 17:21:01 +00:00
sincos 0
sincos -0
sincos pi/2
sincos pi/6
sincos pi/3
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
sincos pi
sincos -pi
Move tests of sincos from libm-test.inc to auto-libm-test-in. This patch moves tests of sincos to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86 and ulps updated accordingly. (auto-libm-test-out diffs omitted below.) * math/auto-libm-test-in: Add tests of sincos. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (sincos_test_data): Use AUTO_TESTS_fFF_11. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_f_11. (func_calc_desc): Add mpfr_f_11 union field. (test_functions): Add sincos. (calc_generic_results): Handle mpfr_f_11. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-19 17:21:01 +00:00
sincos 0.75
sincos 0x1p65
sincos -0x1p65
sincos 0.80190127184058835
sincos 1e22
sincos 0x1p1023
sincos 0x1p16383
sincos 0x1p+120
sincos 0x1p+127
sincos 0x1.fffff8p+127
sincos 0x1.fffffep+127
sincos 0x1p+50
sincos 0x1p+28
Add more tests of cos, sin, sincos. This patch adds some randomly-generated tests of cos, sin and sincos that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cos, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 22:41:00 +00:00
sincos -0x3.3de320f6be87ep+1020
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
sincos 0xe.9f1e5bc3bb88p+112
Add more tests of acosh, atanh, cos, csqrt, erfc, sin, sincos. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, cos, csqrt, erfc, sin and sincos. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-06 17:30:18 +00:00
sincos 0x4.7857dp+68
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
sincos 0x6.287cc8749212e72p+0
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
sincos -0x1.02e34cp+0
sincos 0xf.f0274p+4
sincos 0x3.042d88p+0
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
sincos max
sincos -max
Fix sin, sincos missing underflows (bug 16526, bug 16538). Similar to various other bugs in this area, some sin and sincos implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86, mips64 and powerpc. [BZ #16526] [BZ #16538] * sysdeps/ieee754/dbl-64/s_sin.c: Include <float.h>. (__sin): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/k_sinf.c: Include <float.h>. (__kernel_sinf): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/k_sincosl.c: Include <float.h>. (__kernel_sincosl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/k_sinl.c: Include <float.h>. (__kernel_sinl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/k_sincosl.c: Include <float.h>. (__kernel_sincosl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/k_sinl.c: Include <float.h>. (__kernel_sinl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-96/k_sinl.c: Include <float.h>. (__kernel_sinl): Force underflow exception for arguments with small absolute value. * sysdeps/powerpc/fpu/k_sinf.c: Include <float.h>. (__kernel_sinf): Force underflow exception for arguments with small absolute value. * math/auto-libm-test-in: Add more tests of sin and sincos. * math/auto-libm-test-out: Regenerated.
2015-06-23 22:24:20 +00:00
sincos min
sincos -min
sincos min_subnorm
sincos -min_subnorm
Fix cos computation for multiple precision fallback (bz #20357) During the sincos consolidation I made two mistakes, one was a logical error due to which cos(0x1.8475e5afd4481p+0) returned sin(0x1.8475e5afd4481p+0) instead. The second issue was an error in negating inputs for the correct quadrants for sine. I could not find a suitable test case for this despite running a program to search for such an input for a couple of hours. Following patch fixes both issues. Tested on x86_64. Thanks to Matt Clay for identifying the issue. [BZ #20357] * sysdeps/ieee754/dbl-64/s_sin.c (sloww): Fix up condition to call __mpsin/__mpcos and to negate values. * math/auto-libm-test-in: Add test. * math/auto-libm-test-out: Regenerate.
2016-07-18 17:03:09 +00:00
sincos 0x1.8475e5afd4481p+0
Move tests of sincos from libm-test.inc to auto-libm-test-in. This patch moves tests of sincos to auto-libm-test-in, adding the required support to gen-auto-libm-tests. Tested x86_64 and x86 and ulps updated accordingly. (auto-libm-test-out diffs omitted below.) * math/auto-libm-test-in: Add tests of sincos. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (sincos_test_data): Use AUTO_TESTS_fFF_11. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_f_11. (func_calc_desc): Add mpfr_f_11 union field. (test_functions): Add sincos. (calc_generic_results): Handle mpfr_f_11. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-19 17:21:01 +00:00
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
sinh 0
sinh -0
sinh 0.75
Fix x86/x86_64 expm1 inaccuracy near 0 in directed rounding modes (bug 16293). Bug 16293 is inaccuracy of x86/x86_64 versions of expm1, near 0 in directed rounding modes, that arises from frndint rounding the exponent to 1 or -1 instead of 0, resulting in large cancellation error. This inaccuracy in turn affects other functions such as sinh that use expm1. This patch fixes the problem by setting round-to-nearest mode temporarily around the affected calls to frndint. I don't think this is needed for other uses of frndint, such as in exp itself, as only for expm1 is the cancellation error significant. Tested x86_64 and x86 and ulps updated accordingly. * sysdeps/i386/fpu/e_expl.S (IEEE754_EXPL) [USE_AS_EXPM1L]: Set round-to-nearest mode when using frndint. * sysdeps/i386/fpu/s_expm1.S (__expm1): Likewise. * sysdeps/i386/fpu/s_expm1f.S (__expm1f): Likewise. * sysdeps/x86_64/fpu/e_expl.S (IEEE754_EXPL) [USE_AS_EXPM1L]: Likewise. * math/auto-libm-test-in: Add more tests of expm1. Do not expect sinh test to fail. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (TEST_COND_x86_64): Remove macro. (TEST_COND_x86): Likewise. (expm1_tonearest_test_data): New array. (expm1_test_tonearest): New function. (expm1_towardzero_test_data): New array. (expm1_test_towardzero): New function. (expm1_downward_test_data): New array. (expm1_test_downward): New function. (expm1_upward_test_data): New array. (expm1_test_upward): New function. (main): Run the new test functions. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-19 13:36:10 +00:00
sinh 0x8p-32
Fix ldbl-128ibm sinhl inaccuracy near 0 (bug 18789). ldbl-128ibm sinhl uses a too-big threshold to decide when to return the argument, resulting in large errors. This patch fixes it to use a more appropriate threshold. Tested for x86_64, x86 and powerpc. [BZ #18789] * sysdeps/ieee754/ldbl-128ibm/e_sinhl.c (__ieee754_sinhl): Use smaller threshold for returning the argument. * math/auto-libm-test-in: Add more tests of sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-08-10 15:25:10 +00:00
sinh 0x1p-5
sinh -0x1p-5
sinh 0x1p-10
sinh -0x1p-10
sinh 0x1p-20
sinh -0x1p-20
sinh 0x1p-30
sinh -0x1p-30
sinh 0x1p-40
sinh -0x1p-40
sinh 0x1p-50
sinh -0x1p-50
sinh 0x1p-60
sinh -0x1p-60
sinh 0x1p-70
sinh -0x1p-70
sinh 0x1p-100
sinh -0x1p-100
sinh 0x1p-1000
sinh -0x1p-1000
sinh 0x1p-10000
sinh -0x1p-10000
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
sinh 22
sinh 23
sinh 24
Add more tests of cosh, sinh. This patch adds some randomly-generated tests of cosh and sinh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cosh and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-03-24 23:48:04 +00:00
sinh -0x7.55d7f8p-4
sinh -0x3.f392f8p-4
sinh 0x1.c56446p+0
sinh 0x6.cac622d51eebcp-4
sinh -0x5.c4cb02389c094p+0
sinh -0x1.646850f515ef2p+0
sinh -0x7.a8c5f68c81fae5dp-4
sinh 0x3.4a037p-4
sinh -0x3.eba6dbcbeceb2p-4
sinh -0x2.55f63p+0
Add further tests of cosh and sinh. This patch adds some further randomly-generated tests of cosh and sinh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of cosh and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-04-30 22:32:08 +00:00
sinh -0x3.ca68c96337692p-4
sinh -0x3.92da05a85024b314p-4
sinh -0x3.3e6292ed442d450cp-4
sinh 0x7.6e259d2436fc4p-4
Add further tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. (This process must eventually converge, when my random test generation stops finding inputs that increase the listed ulps, except maybe for any cases uncovered where the errors exceed the maximum allowed 9ulp error and so indicate actual libm bugs needing fixing.) Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of acosh, atanh, clog, clog10, csqrt, erfc, exp2, expm1, log10, log2 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-05 22:59:41 +00:00
sinh 0x3.d6e088p-4
sinh -0x7.688eap-4
Add more tests of csqrt, lgamma, log10, sinh. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of csqrt, lgamma, log10 and sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-08 17:55:11 +00:00
sinh -0xd.dce79p-4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
sinh 0x8.a3127p+4
sinh 0x1.c0709p-12
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
sinh 0xb.7f67c3586c24p-4
sinh -0x1.3dda8ap+0
sinh -0x5.ee9218p-4
sinh -0x1.bcfc98p+0
sinh -0x6.9bbb6df7c5d08p-4
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (ldbl-96)
sinh 0x2.c5d376167f4052f4p+12
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
sinh max
sinh -max
Fix sinh missing underflows (bug 16519). Similar to various other bugs in this area, some sinh implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86, mips64 and powerpc. [BZ #16519] * sysdeps/ieee754/dbl-64/e_sinh.c: Include <float.h>. (__ieee754_sinh): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/e_sinhf.c: Include <float.h>. (__ieee754_sinhf): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/e_sinhl.c: Include <float.h>. (__ieee754_sinhl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/e_sinhl.c: Include <float.h>. (__ieee754_sinhl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-96/e_sinhl.c: Include <float.h>. (__ieee754_sinhl): Force underflow exception for arguments with small absolute value. * math/auto-libm-test-in: Add more tests of sinh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-08-06 23:01:09 +00:00
sinh min
sinh -min
sinh min_subnorm
sinh -min_subnorm
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
sinh 0x5.96a7ep+4
sinh 0x5.96a7e8p+4
sinh 0x2.c679d1f73f0fap+8
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
sinh 0x2.c679d1f73f0fcp+8
Update libm-test XFAILs for ibm128 format. This patch cleans up and updates the libm-test XFAILs for the ibm128 format. More of them are changed to use a new ibm128-libgcc conditional, to reflect that they are not in fact needed if you've patched libgcc to fix the known issues (at substantial performance cost). Many additional XFAILs are added for tests that fail with unpatched libgcc (most but not all of them xfail-rounding). Note that further such fixes will be needed for test-ldouble actually to pass with default libgcc (in particular, XFAILs for pow tests and for various affected tests directly embedded in libm-test.inc). With patched libgcc, there may be a few XFAILs needed but the results are already substantially clean apart from a few ulps differences. Tested for powerpc. * math/libm-test.inc (TEST_COND_ibm128_libgcc): New macro. (init_max_error) [TEST_COND_ibm128]: Increase maximum error allowed to 16 ulps. * math/auto-libm-test-in: Change most XFAILs for ibm128 to use ibm128-libgcc. XFAIL more tests for ibm128-libgcc. * math/auto-libm-test-out: Regenerated.
2017-01-04 23:29:42 +00:00
sinh 0x2.c679d1f73f0fb624d358b213a7p+8 xfail-rounding:ibm128-libgcc
sinh 0x2.c679d1f73f0fb624d358b213a8p+8 xfail-rounding:ibm128-libgcc
math: Remove inline math tests With mathinline removal there is no need to keep building and testing inline math tests. The gen-libm-tests.py support to generate ULP_I_* is removed and all libm-test-ulps files are updated to longer have the i{float,double,ldouble} entries. The support for no-test-inline is also removed from both gen-auto-libm-tests and the auto-libm-test-out-* were regenerated. Checked on x86_64-linux-gnu and i686-linux-gnu.
2020-03-17 18:46:29 +00:00
sinh 0x2.c5d37700c6bb03a4p+12
sinh 0x2.c5d37700c6bb03a8p+12
sinh 0x2.c5d37700c6bb03a6c24b6c9b494cp+12
sinh 0x2.c5d37700c6bb03a6c24b6c9b494ep+12
add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)
2020-11-25 18:03:33 +00:00
# the next value generates larger error bounds on x86_64 (binary64)
sinh -0x1.633c62890fa14p+9
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
Add narrowing subtract functions. This patch adds the narrowing subtract functions from TS 18661-1 to glibc's libm: fsub, fsubl, dsubl, f32subf64, f32subf32x, f32xsubf64 for all configurations; f32subf64x, f32subf128, f64subf64x, f64subf128, f32xsubf64x, f32xsubf128, f64xsubf128 for configurations with _Float64x and _Float128; __nldbl_dsubl for ldbl-opt. The changes are essentially the same as for the narrowing add functions, so the description of those generally applies to this patch as well. Tested for x86_64, x86, mips64 (all three ABIs, both hard and soft float) and powerpc, and with build-many-glibcs.py. * math/Makefile (libm-narrow-fns): Add sub. (libm-test-funcs-narrow): Likewise. * math/Versions (GLIBC_2.28): Add narrowing subtract functions. * math/bits/mathcalls-narrow.h (sub): Use __MATHCALL_NARROW. * math/gen-auto-libm-tests.c (test_functions): Add sub. * math/math-narrow.h (CHECK_NARROW_SUB): New macro. (NARROW_SUB_ROUND_TO_ODD): Likewise. (NARROW_SUB_TRIVIAL): Likewise. * sysdeps/ieee754/float128/float128_private.h (__fsubl): New macro. (__dsubl): Likewise. * sysdeps/ieee754/ldbl-opt/Makefile (libnldbl-calls): Add fsub and dsub. (CFLAGS-nldbl-dsub.c): New variable. (CFLAGS-nldbl-fsub.c): Likewise. * sysdeps/ieee754/ldbl-opt/Versions (GLIBC_2.28): Add __nldbl_dsubl. * sysdeps/ieee754/ldbl-opt/nldbl-compat.h (__nldbl_dsubl): New prototype. * manual/arith.texi (Misc FP Arithmetic): Document fsub, fsubl, dsubl, fMsubfN, fMsubfNx, fMxsubfN and fMxsubfNx. * math/auto-libm-test-in: Add tests of sub. * math/auto-libm-test-out-narrow-sub: New generated file. * math/libm-test-narrow-sub.inc: New file. * sysdeps/i386/fpu/s_f32xsubf64.c: Likewise. * sysdeps/ieee754/dbl-64/s_f32xsubf64.c: Likewise. * sysdeps/ieee754/dbl-64/s_fsub.c: Likewise. * sysdeps/ieee754/float128/s_f32subf128.c: Likewise. * sysdeps/ieee754/float128/s_f64subf128.c: Likewise. * sysdeps/ieee754/float128/s_f64xsubf128.c: Likewise. * sysdeps/ieee754/ldbl-128/s_dsubl.c: Likewise. * sysdeps/ieee754/ldbl-128/s_f64xsubf128.c: Likewise. * sysdeps/ieee754/ldbl-128/s_fsubl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_dsubl.c: Likewise. * sysdeps/ieee754/ldbl-128ibm/s_fsubl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_dsubl.c: Likewise. * sysdeps/ieee754/ldbl-96/s_fsubl.c: Likewise. * sysdeps/ieee754/ldbl-opt/nldbl-dsub.c: Likewise. * sysdeps/ieee754/ldbl-opt/nldbl-fsub.c: Likewise. * sysdeps/ieee754/soft-fp/s_dsubl.c: Likewise. * sysdeps/ieee754/soft-fp/s_fsub.c: Likewise. * sysdeps/ieee754/soft-fp/s_fsubl.c: Likewise. * sysdeps/powerpc/fpu/libm-test-ulps: Update. * sysdeps/mach/hurd/i386/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/aarch64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/alpha/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/arm/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/hppa/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/i386/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/m68k/coldfire/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/m68k/m680x0/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/microblaze/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/mips/mips32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/mips/mips64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/nios2/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/fpu/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/nofpu/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/libm-le.abilist: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/riscv/rv64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/s390/s390-32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/s390/s390-64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sh/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/tile/tilegx32/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/tile/tilegx64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/x86_64/64/libm.abilist: Likewise. * sysdeps/unix/sysv/linux/x86_64/x32/libm.abilist: Likewise.
2018-03-20 00:34:52 +00:00
sub 0 0
sub 0 -0
sub -0 0
sub -0 -0
sub max max
sub max -max
sub -max max
sub -max -max
sub min min missing-underflow:arg-ibm128
sub min -min missing-underflow:arg-ibm128
sub -min min missing-underflow:arg-ibm128
sub -min -min missing-underflow:arg-ibm128
sub min_subnorm min_subnorm missing-underflow:arg-ibm128
sub min_subnorm -min_subnorm missing-underflow:arg-ibm128
sub -min_subnorm min_subnorm missing-underflow:arg-ibm128
sub -min_subnorm -min_subnorm missing-underflow:arg-ibm128
sub 1 2
sub 1 -2
sub -1 2
sub -1 -2
sub 100.5 0.75
sub 100.5 -0.75
sub -100.5 0.75
sub -100.5 -0.75
sub 1 0x1p-23
sub 1 0x1.7fp-23
sub 1 0x1.8p-23
sub 1 0x1.81p-23
sub 1 0x1p-24
sub 1 0x1.1p-24
sub 1 0x0.fp-24
sub 1 min
sub 1 -min
sub 1 min_subnorm
sub 1 -min_subnorm
sub -1 min
sub -1 -min
sub -1 min_subnorm
sub -1 -min_subnorm
# Cases where larger argument is half way between two values of a
# floating-point format, so that double rounding would sometimes yield
# the wrong result.
sub 0x1.000001p0 min
sub 0x1.000001p0 -min
sub 0x1.000001p0 min_subnorm
sub 0x1.000001p0 -min_subnorm
sub -0x1.000001p0 min
sub -0x1.000001p0 -min
sub -0x1.000001p0 min_subnorm
sub -0x1.000001p0 -min_subnorm
sub 0x1.00000000000008p0 min
sub 0x1.00000000000008p0 -min
sub 0x1.00000000000008p0 min_subnorm
sub 0x1.00000000000008p0 -min_subnorm
sub -0x1.00000000000008p0 min
sub -0x1.00000000000008p0 -min
sub -0x1.00000000000008p0 min_subnorm
sub -0x1.00000000000008p0 -min_subnorm
sub 0x1.0000000000000001p0 min
sub 0x1.0000000000000001p0 -min
sub 0x1.0000000000000001p0 min_subnorm
sub 0x1.0000000000000001p0 -min_subnorm
sub -0x1.0000000000000001p0 min
sub -0x1.0000000000000001p0 -min
sub -0x1.0000000000000001p0 min_subnorm
sub -0x1.0000000000000001p0 -min_subnorm
# Two values representable in the same format with difference very
# close to half way between two representable values. In particular,
# verify that there is no intermediate rounding to 64 bits when
# subtracting two double values.
sub 1 0x1.000002p-24
sub 1 -0x1.000002p-24
sub 1 0x0.ffffffp-24
sub 1 -0x0.ffffffp-24
sub 0x1.000002p0 0x1.000002p-24
sub 0x1.000002p0 -0x1.000002p-24
sub 0x1.000002p0 0x0.ffffffp-24
sub 0x1.000002p0 -0x0.ffffffp-24
sub 1 0x1.0000000000001p-53
sub 1 -0x1.0000000000001p-53
sub 1 0x0.fffffffffffff8p-53
sub 1 -0x0.fffffffffffff8p-53
sub 0x1.0000000000001p0 0x1.0000000000001p-53
sub 0x1.0000000000001p0 -0x1.0000000000001p-53
sub 0x1.0000000000001p0 0x0.fffffffffffff8p-53
sub 0x1.0000000000001p0 -0x0.fffffffffffff8p-53
sub 1 0x1.0000000000000002p-64
sub 1 -0x1.0000000000000002p-64
sub 1 0x0.ffffffffffffffffp-64
sub 1 -0x0.ffffffffffffffffp-64
sub 0x1.0000000000000002p0 0x1.0000000000000002p-64
sub 0x1.0000000000000002p0 -0x1.0000000000000002p-64
sub 0x1.0000000000000002p0 0x0.ffffffffffffffffp-64
sub 0x1.0000000000000002p0 -0x0.ffffffffffffffffp-64
Start generating libm tests automatically with MPFR.
2013-11-29 16:27:55 +00:00
sqrt 0
sqrt -0
sqrt 2209
sqrt 4
sqrt 2
sqrt 0.25
sqrt 6642.25
sqrt 15190.5625
sqrt 0.75
Fix x86 sqrt rounding (bug 14032).
2013-11-29 16:31:16 +00:00
sqrt 0x1.fffffffffffffp+1023
sqrt 0x1.ffffffffffffbp+1023
sqrt 0x1.ffffffffffff7p+1023
sqrt 0x1.ffffffffffff3p+1023
sqrt 0x1.fffffffffffefp+1023
sqrt 0x1.fffffffffffebp+1023
sqrt 0x1.fffffffffffe7p+1023
sqrt 0x1.fffffffffffe3p+1023
sqrt 0x1.fffffffffffdfp+1023
sqrt 0x1.fffffffffffdbp+1023
sqrt 0x1.fffffffffffd7p+1023
sqrt 0x1.0000000000003p-1022
sqrt 0x1.0000000000007p-1022
sqrt 0x1.000000000000bp-1022
sqrt 0x1.000000000000fp-1022
sqrt 0x1.0000000000013p-1022
sqrt 0x1.0000000000017p-1022
sqrt 0x1.000000000001bp-1022
sqrt 0x1.000000000001fp-1022
sqrt 0x1.0000000000023p-1022
sqrt 0x1.0000000000027p-1022
sqrt 0x1.000000000002bp-1022
sqrt 0x1.000000000002fp-1022
sqrt 0x1.0000000000033p-1022
sqrt 0x1.0000000000037p-1022
sqrt 0x1.7167bc36eaa3bp+6
sqrt 0x1.7570994273ad7p+6
sqrt 0x1.7dae969442fe6p+6
sqrt 0x1.7f8444fcf67e5p+6
sqrt 0x1.8364650e63a54p+6
sqrt 0x1.85bedd274edd8p+6
sqrt 0x1.8609cf496ab77p+6
sqrt 0x1.873849c70a375p+6
sqrt 0x1.8919c962cbaaep+6
sqrt 0x1.8de4493e22dc6p+6
sqrt 0x1.924829a17a288p+6
sqrt 0x1.92702cd992f12p+6
sqrt 0x1.92b763a8311fdp+6
sqrt 0x1.947da013c7293p+6
sqrt 0x1.9536091c494d2p+6
sqrt 0x1.61b04c6p-1019
sqrt 0x1.93789f1p-1018
sqrt 0x1.a1989b4p-1018
sqrt 0x1.f93bc9p-1018
sqrt 0x1.2f675e3p-1017
sqrt 0x1.a158508p-1017
sqrt 0x1.cd31f078p-1017
sqrt 0x1.33b43b08p-1016
sqrt 0x1.6e66a858p-1016
sqrt 0x1.8661cbf8p-1016
sqrt 0x1.bbb221b4p-1016
sqrt 0x1.c4942f3cp-1016
sqrt 0x1.dbb258c8p-1016
sqrt 0x1.57103ea4p-1015
sqrt 0x1.9b294f88p-1015
sqrt 0x1.0000000000001p+0
sqrt 0x1.fffffffffffffp-1
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
sqrt max
sqrt min
sqrt min_subnorm
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tan 0
tan -0
tan pi/4
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
tan pi/2
tan -pi/2
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tan 0.75
tan 0x1p65
tan -0x1p65
Fix tanf spurious underflows (bug 18221). The flt-32 implementation of tanf produces spurious underflow exceptions for some small arguments, through computing values on the order of x^5. This patch fixes this by adjusting the threshold for returning x (or, as applicable, +/- 1/x) to 2**-13 (the next term in the power series being x^3/3). Tested for x86_64 and x86. [BZ #18221] * sysdeps/ieee754/flt-32/k_tanf.c (__kernel_tanf): Use 2**-13 not 2**-28 as threshold for returning x or +/- 1/x. * math/auto-libm-test-in: Add more tests of tan. * math/auto-libm-test-out: Regenerated.
2015-05-15 17:47:29 +00:00
tan 0x1p-27
tan -0x1p-27
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tan 0xc.9p-4
tan 0xc.908p-4
tan 0xc.90cp-4
tan 0xc.90ep-4
tan 0xc.90fp-4
tan 0xc.90f8p-4
tan 0xc.90fcp-4
tan 0xc.90fdp-4
tan 0xc.90fd8p-4
tan 0xc.90fdap-4
tan 0xc.ap-4
tan 0xc.98p-4
tan 0xc.94p-4
tan 0xc.92p-4
tan 0xc.91p-4
tan 0xc.90fep-4
tan 0xc.90fdcp-4
tan 0xc.90fdbp-4
tan -0xc.9p-4
tan -0xc.908p-4
tan -0xc.90cp-4
tan -0xc.90ep-4
tan -0xc.90fp-4
tan -0xc.90f8p-4
tan -0xc.90fcp-4
tan -0xc.90fdp-4
tan -0xc.90fd8p-4
tan -0xc.90fdap-4
tan -0xc.ap-4
tan -0xc.98p-4
tan -0xc.94p-4
tan -0xc.92p-4
tan -0xc.91p-4
tan -0xc.90fep-4
tan -0xc.90fdcp-4
tan -0xc.90fdbp-4
tan 1e22
tan 0x1p1023
tan 0x1p16383
tan 1
tan 2
tan 3
tan 4
tan 5
tan 6
tan 7
tan 8
tan 9
tan 10
Add more tests of tan. This patch adds some randomly-generated tests of tan that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of tan. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 22:54:39 +00:00
tan -0x1.062a48p+0
tan -0x1.4f69cp+0
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
tan 0x1.6ca7e8p+0
tan -0x1.b569cp+0
tan -0x2.12bafcp+0
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
tan 0x2.091d68p+0
tan -0x5.302ab9b18593264p+0
tan 0x1.1ad374p+0
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
tan -0x1.0d55b8p+0
Improve float range reduction accuracy near pi/2 (bug 21094). Bug 21094 reports 3ulp errors of cosf and tanf for certain arguments near pi/2 arising from the use of an insufficiently accurate range reduction. (To be clear, this is a quality-of-implementation issue relating to the apparent intent of those particular cosf and tanf implementations; 3ulp is within the general glibc accuracy goals, so not inherently a bug.) This patch fixes that error by making a wider range of cases use the existing more accurate range reduction for arguments close to pi/2. The wider range of values is still narrow enough for the "z -= pio2_2;" in the more accurate case to be exact, as the code expects. Tested for x86_64, x86 and mips64; no ulps updates needed (but at least on mips64, the larger ulps were seen if the tests were added without the substantive fix). [BZ #21094] * sysdeps/ieee754/flt-32/e_rem_pio2f.c (__ieee754_rem_pio2f): Use 24+24+24-bit pi for wider range of values around pi/2. * math/auto-libm-test-in: Add more tests of cos and tan. * math/auto-libm-test-out-cos: Regenerated. * math/auto-libm-test-out-tan: Likewise.
2017-03-15 22:00:54 +00:00
tan 1.57079697
tan -1.57079697
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
tan 0x1p-5
tan 0x1p-10
tan 0x1p-15
tan 0x1p-20
tan 0x1p-25
tan 0x1p-30
tan 0x1p-35
tan 0x1p-40
tan 0x1p-45
tan 0x1p-50
tan 0x1p-55
tan 0x1p-60
tan 0x1p-100
tan 0x1p-600
tan 0x1p-10000
tan max
tan -max
Fix tan missing underflows (bug 16517). Similar to various other bugs in this area, some tan implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86, mips64 and powerpc. [BZ #16517] * sysdeps/ieee754/dbl-64/s_tan.c: Include <float.h>. (tan): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/k_tanf.c: Include <float.h>. (__kernel_tanf): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/k_tanl.c: Include <float.h>. (__kernel_tanl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/k_tanl.c: Include <float.h>. (__kernel_tanl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-96/k_tanl.c: Include <float.h>. (__kernel_tanl): Force underflow exception for arguments with small absolute value. * math/auto-libm-test-in: Add more tests of tan. * math/auto-libm-test-out: Regenerated.
2015-08-07 23:10:35 +00:00
tan min
tan -min
tan min_subnorm
tan -min_subnorm
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tanh 0
tanh -0
tanh 0.75
tanh -0.75
tanh 1.0
tanh -1.0
Fix ldbl-128ibm tanhl inaccuracy (bug 18790). ldbl-128ibm tanhl uses a too-small threshold to decide when to return +/-1, resulting in large errors. This patch changes it to a more appropriate threshold (the requirement is for 2*exp(-2|x|) to be small in terms of ulps of 1). Tested for x86_64, x86 and powerpc. [BZ #18790] * sysdeps/ieee754/ldbl-128ibm/s_tanhl.c (__tanhl): Increase threshold for returning +/- 1. * math/auto-libm-test-in: Add more tests of tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-08-10 20:35:30 +00:00
tanh 2
tanh -2
tanh 3
tanh -3
tanh 4
tanh -4
tanh 5
tanh -5
tanh 6
tanh -6
tanh 7
tanh -7
tanh 8
tanh -8
tanh 9
tanh -9
tanh 10
tanh -10
tanh 15
tanh -15
tanh 20
tanh -20
tanh 22
tanh -22
tanh 25
tanh -25
tanh 30
tanh -30
tanh 35
tanh -35
tanh 40
tanh -40
tanh 45
tanh -45
tanh 50
tanh -50
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tanh 0x1p-57
Add more tests of tanh. This patch adds some randomly-generated tests of tanh that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 23:06:44 +00:00
tanh 0xe.6c659p-4
tanh 0x8.c259ep-4
tanh 0x6.5821dp-4
tanh 0x8.7c9e5p-4
tanh -0x3.b60d7cp-4
tanh 0x7.b9985p-4
tanh 0x7.a18e8p-4
tanh -0x2.6082fp-4
tanh 0xe.05031p-16
tanh 0x3.c80eaa7adaa3p-4
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
tanh 0x2.00f9857616524p-4
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
tanh -0xe.9e035p+0
tanh -0x3.c0d8b54c5a488p-4
Add more random libm test inputs (mainly for ldbl-128). This patch adds more libm test inputs found through random test generation to increase previously known ulps. This particular test generation was run for mips64, so most of the increased ulps are for ldbl-128 (float and double having been fairly well covered by such testing for x86_64), but there's the odd ulps increase for other formats. Tested for x86_64, x86 and mips64. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, carg, cos, csqrt, erfc, exp, exp10, exp2, log, log1p, log2, pow, sin, sincos, sinh, tan and tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-09-12 00:01:38 +00:00
tanh -0x3.2f59p-4
tanh 0x2.e6f54cp-4
tanh 0x3.397f2f50241d031p-4
tanh 0x8.4024a11b6610672b2982b852e8p-4
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
tanh 0x1p-5
tanh 0x1p-10
tanh 0x1p-15
tanh 0x1p-20
tanh 0x1p-25
tanh 0x1p-30
tanh 0x1p-35
tanh 0x1p-40
tanh 0x1p-45
tanh 0x1p-50
tanh 0x1p-55
tanh 0x1p-60
tanh 0x1p-100
tanh 0x1p-600
tanh 0x1p-10000
Add inputs that generate larger error bounds (Using values from https://members.loria.fr/PZimmermann/papers/accuracy.pdf)
2021-02-26 07:00:03 +00:00
# the next value generates larger error bounds on x86_64 (ldbl-96)
tanh 0x3.b9979a543d0fbfa8p-4
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
tanh max
tanh -max
Fix tanh missing underflows (bug 16520). Similar to various other bugs in this area, some tanh implementations do not raise the underflow exception for subnormal arguments, when the result is tiny and inexact. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86, mips64 and powerpc. [BZ #16520] * sysdeps/ieee754/dbl-64/s_tanh.c: Include <float.h>. (__tanh): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/flt-32/s_tanhf.c: Include <float.h>. (__tanhf): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128/s_tanhl.c: Include <float.h>. (__tanhl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-128ibm/s_tanhl.c: Include <float.h>. (__tanhl): Force underflow exception for arguments with small absolute value. * sysdeps/ieee754/ldbl-96/s_tanhl.c: Include <float.h>. (__tanhl): Force underflow exception for arguments with small absolute value. * math/auto-libm-test-in: Add more tests of tanh. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-08-13 16:40:39 +00:00
tanh min
tanh -min
tanh min_subnorm
tanh -min_subnorm
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma 0.5
tgamma -0.5
tgamma 1
tgamma 2
tgamma 3
tgamma 4
tgamma 5
tgamma 6
tgamma 7
tgamma 8
tgamma 9
tgamma 10
tgamma 0.7
tgamma 1.2
tgamma 1.5
tgamma 2.5
tgamma 3.5
tgamma 4.5
tgamma 5.5
tgamma 6.5
tgamma 7.5
tgamma 8.5
tgamma 9.5
tgamma -1.5
tgamma -2.5
tgamma -3.5
tgamma -4.5
tgamma -5.5
tgamma -6.5
tgamma -7.5
tgamma -8.5
tgamma -9.5
tgamma 0x1p-24
tgamma -0x1p-24
tgamma 0x1p-53
tgamma -0x1p-53
tgamma 0x1p-64
tgamma -0x1p-64
tgamma 0x1p-106
tgamma -0x1p-106
tgamma 0x1p-113
tgamma -0x1p-113
tgamma 0x1p-127
tgamma -0x1p-127
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
# IEEE semantics mean overflow very close to the threshold depends on
# the rounding mode; gen-auto-libm-tests does not reflect that glibc
# does not try to achieve this.
Remove type specific information from auto-libm-test-in Apply the following sed regexes to auto-libm-test-in in order: s/flt-32/binary32/ s/dbl-64/binary64/ s/ldbl-96-intel/intel96/ s/ldbl-96-m68k/m68k96/ s/ldbl-128ibm/ibm128/ s/ldbl-128/binary128/ and fixup ldbl-96 comment manually.
2016-06-08 19:41:05 +00:00
tgamma 0x1p-128 spurious-overflow:binary32
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma -0x1p-128
tgamma 0x1p-149
tgamma -0x1p-149
tgamma 0x1p-1023
tgamma -0x1p-1023
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
tgamma 0x1p-1024 spurious-overflow:binary64 spurious-overflow:ibm128
tgamma -0x1p-1024
tgamma 0x1p-1074
tgamma -0x1p-1074
tgamma 0x1p-16383
tgamma -0x1p-16383
Remove type specific information from auto-libm-test-in Apply the following sed regexes to auto-libm-test-in in order: s/flt-32/binary32/ s/dbl-64/binary64/ s/ldbl-96-intel/intel96/ s/ldbl-96-m68k/m68k96/ s/ldbl-128ibm/ibm128/ s/ldbl-128/binary128/ and fixup ldbl-96 comment manually.
2016-06-08 19:41:05 +00:00
tgamma 0x1p-16384 spurious-overflow:intel96 spurious-overflow:m68k96 spurious-overflow:binary128
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
tgamma -0x1p-16384
tgamma 0x1p-16445
tgamma -0x1p-16445
tgamma 0x1p-16494
tgamma -0x1p-16494
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma 0x8.00001p0
tgamma 0x7.fffff8p0
tgamma 0x7.000008p0
tgamma 0x6.fffff8p0
tgamma 0x6.000008p0
tgamma 0x5.fffff8p0
tgamma 0x5.000008p0
tgamma 0x4.fffff8p0
tgamma 0x4.000008p0
tgamma 0x3.fffffcp0
tgamma 0x3.000004p0
tgamma 0x2.fffffcp0
tgamma 0x2.000004p0
tgamma 0x1.fffffep0
tgamma 0x1.000002p0
tgamma 0x0.ffffffp0
tgamma -0x0.ffffffp0
tgamma -0x1.000002p0
tgamma -0x1.fffffep0
tgamma -0x2.000004p0
tgamma -0x2.fffffcp0
tgamma -0x3.000004p0
tgamma -0x3.fffffcp0
tgamma -0x4.000008p0
tgamma -0x4.fffff8p0
tgamma -0x5.000008p0
tgamma -0x5.fffff8p0
tgamma -0x6.000008p0
tgamma -0x6.fffff8p0
tgamma -0x7.000008p0
tgamma -0x7.fffff8p0
tgamma -0x8.00001p0
tgamma -0x9.fffffp0
tgamma -0xa.00001p0
tgamma -0x13.ffffep0
tgamma -0x14.00002p0
tgamma -0x1d.ffffep0
tgamma -0x1e.00002p0
tgamma -0x27.ffffcp0
tgamma -0x28.00004p0
tgamma -0x28.ffffcp0
tgamma -0x29.00004p0
Fix tgamma errno setting on underflow (bug 6810).
2013-12-05 14:01:41 +00:00
tgamma -0x29.ffffcp0
tgamma -0x2a.00004p0
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma 0x8.0000000000008p0
tgamma 0x7.ffffffffffffcp0
tgamma 0x7.0000000000004p0
tgamma 0x6.ffffffffffffcp0
tgamma 0x6.0000000000004p0
tgamma 0x5.ffffffffffffcp0
tgamma 0x5.0000000000004p0
tgamma 0x4.ffffffffffffcp0
tgamma 0x4.0000000000004p0
tgamma 0x3.ffffffffffffep0
tgamma 0x3.0000000000002p0
tgamma 0x2.ffffffffffffep0
tgamma 0x2.0000000000002p0
tgamma 0x1.fffffffffffffp0
tgamma 0x1.0000000000001p0
tgamma 0x0.fffffffffffff8p0
tgamma -0x0.fffffffffffff8p0
tgamma -0x1.0000000000001p0
tgamma -0x1.fffffffffffffp0
tgamma -0x2.0000000000002p0
tgamma -0x2.ffffffffffffep0
tgamma -0x3.0000000000002p0
tgamma -0x3.ffffffffffffep0
tgamma -0x4.0000000000004p0
tgamma -0x4.ffffffffffffcp0
tgamma -0x5.0000000000004p0
tgamma -0x5.ffffffffffffcp0
tgamma -0x6.0000000000004p0
tgamma -0x6.ffffffffffffcp0
tgamma -0x7.0000000000004p0
tgamma -0x7.ffffffffffffcp0
tgamma -0x8.0000000000008p0
tgamma -0x9.ffffffffffff8p0
tgamma -0xa.0000000000008p0
tgamma -0x13.ffffffffffffp0
tgamma -0x14.000000000001p0
tgamma -0x1d.ffffffffffffp0
tgamma -0x1e.000000000001p0
tgamma -0x27.fffffffffffep0
tgamma -0x28.000000000002p0
tgamma -0x28.fffffffffffep0
tgamma -0x29.000000000002p0
Fix tgamma errno setting on underflow (bug 6810).
2013-12-05 14:01:41 +00:00
tgamma -0x29.fffffffffffep0
tgamma -0x2a.000000000002p0
tgamma -0x31.fffffffffffep0
tgamma -0x32.000000000002p0
tgamma -0x63.fffffffffffcp0
tgamma -0x64.000000000004p0
tgamma -0x95.fffffffffff8p0
tgamma -0x96.000000000008p0
tgamma -0xb4.fffffffffff8p0
tgamma -0xb5.000000000008p0
tgamma -0xb5.fffffffffff8p0
tgamma -0xb6.000000000008p0
tgamma -0xb6.fffffffffff8p0
tgamma -0xb7.000000000008p0
tgamma -0xb7.fffffffffff8p0
tgamma -0xb8.000000000008p0
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma 0x8.00000000000000000000000004p0
tgamma 0x7.fffffffffffffffffffffffffep0
tgamma 0x7.00000000000000000000000002p0
tgamma 0x6.fffffffffffffffffffffffffep0
tgamma 0x6.00000000000000000000000002p0
tgamma 0x5.fffffffffffffffffffffffffep0
tgamma 0x5.00000000000000000000000002p0
tgamma 0x4.fffffffffffffffffffffffffep0
tgamma 0x4.00000000000000000000000002p0
tgamma 0x3.ffffffffffffffffffffffffffp0
tgamma 0x3.00000000000000000000000001p0
tgamma 0x2.ffffffffffffffffffffffffffp0
tgamma 0x2.00000000000000000000000001p0
tgamma 0x1.ffffffffffffffffffffffffff8p0
tgamma 0x1.000000000000000000000000008p0
tgamma 0x0.ffffffffffffffffffffffffffcp0
tgamma -0x0.ffffffffffffffffffffffffffcp0
tgamma -0x1.000000000000000000000000008p0
tgamma -0x1.ffffffffffffffffffffffffff8p0
tgamma -0x2.00000000000000000000000001p0
tgamma -0x2.ffffffffffffffffffffffffffp0
tgamma -0x3.00000000000000000000000001p0
tgamma -0x3.ffffffffffffffffffffffffffp0
tgamma -0x4.00000000000000000000000002p0
tgamma -0x4.fffffffffffffffffffffffffep0
tgamma -0x5.00000000000000000000000002p0
tgamma -0x5.fffffffffffffffffffffffffep0
tgamma -0x6.00000000000000000000000002p0
tgamma -0x6.fffffffffffffffffffffffffep0
tgamma -0x7.00000000000000000000000002p0
tgamma -0x7.fffffffffffffffffffffffffep0
tgamma -0x8.00000000000000000000000004p0
tgamma -0x9.fffffffffffffffffffffffffcp0
tgamma -0xa.00000000000000000000000004p0
tgamma -0x13.fffffffffffffffffffffffff8p0
tgamma -0x14.00000000000000000000000008p0
tgamma -0x1d.fffffffffffffffffffffffff8p0
tgamma -0x1e.00000000000000000000000008p0
tgamma -0x27.fffffffffffffffffffffffffp0
tgamma -0x28.0000000000000000000000001p0
tgamma -0x28.fffffffffffffffffffffffffp0
tgamma -0x29.0000000000000000000000001p0
Fix tgamma errno setting on underflow (bug 6810).
2013-12-05 14:01:41 +00:00
tgamma -0x29.fffffffffffffffffffffffffp0
tgamma -0x2a.0000000000000000000000001p0
tgamma -0x31.fffffffffffffffffffffffffp0
tgamma -0x32.0000000000000000000000001p0
tgamma -0x63.ffffffffffffffffffffffffep0
tgamma -0x64.0000000000000000000000002p0
tgamma -0x95.ffffffffffffffffffffffffcp0
tgamma -0x96.0000000000000000000000004p0
tgamma -0xb4.ffffffffffffffffffffffffcp0
tgamma -0xb5.0000000000000000000000004p0
tgamma -0xb5.ffffffffffffffffffffffffcp0
tgamma -0xb6.0000000000000000000000004p0
tgamma -0xb6.ffffffffffffffffffffffffcp0
tgamma -0xb7.0000000000000000000000004p0
tgamma -0xb7.ffffffffffffffffffffffffcp0
tgamma -0xb8.0000000000000000000000004p0
tgamma -0xbb.ffffffffffffffffffffffffcp0
tgamma -0xbc.0000000000000000000000004p0
tgamma -0xbc.ffffffffffffffffffffffffcp0
tgamma -0xbd.0000000000000000000000004p0
tgamma -0xbd.ffffffffffffffffffffffffcp0
tgamma -0xbe.0000000000000000000000004p0
tgamma -0xbe.ffffffffffffffffffffffffcp0
tgamma -0xbf.0000000000000000000000004p0
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma 0x8.000000000000001p0
tgamma 0x7.fffffffffffffff8p0
tgamma 0x7.0000000000000008p0
tgamma 0x6.fffffffffffffff8p0
tgamma 0x6.0000000000000008p0
tgamma 0x5.fffffffffffffff8p0
tgamma 0x5.0000000000000008p0
tgamma 0x4.fffffffffffffff8p0
tgamma 0x4.0000000000000008p0
tgamma 0x3.fffffffffffffffcp0
tgamma 0x3.0000000000000004p0
tgamma 0x2.fffffffffffffffcp0
tgamma 0x2.0000000000000004p0
tgamma 0x1.fffffffffffffffep0
tgamma 0x1.0000000000000002p0
tgamma 0x0.ffffffffffffffffp0
tgamma -0x0.ffffffffffffffffp0
tgamma -0x1.0000000000000002p0
tgamma -0x1.fffffffffffffffep0
tgamma -0x2.0000000000000004p0
tgamma -0x2.fffffffffffffffcp0
tgamma -0x3.0000000000000004p0
tgamma -0x3.fffffffffffffffcp0
tgamma -0x4.0000000000000008p0
tgamma -0x4.fffffffffffffff8p0
tgamma -0x5.0000000000000008p0
tgamma -0x5.fffffffffffffff8p0
tgamma -0x6.0000000000000008p0
tgamma -0x6.fffffffffffffff8p0
tgamma -0x7.0000000000000008p0
tgamma -0x7.fffffffffffffff8p0
tgamma -0x8.000000000000001p0
tgamma -0x9.fffffffffffffffp0
tgamma -0xa.000000000000001p0
tgamma -0x13.ffffffffffffffep0
tgamma -0x14.000000000000002p0
tgamma -0x1d.ffffffffffffffep0
tgamma -0x1e.000000000000002p0
tgamma -0x27.ffffffffffffffcp0
tgamma -0x28.000000000000004p0
tgamma -0x28.ffffffffffffffcp0
tgamma -0x29.000000000000004p0
Fix tgamma errno setting on underflow (bug 6810).
2013-12-05 14:01:41 +00:00
tgamma -0x29.ffffffffffffffcp0
tgamma -0x2a.000000000000004p0
tgamma -0x31.ffffffffffffffcp0
tgamma -0x32.000000000000004p0
tgamma -0x63.ffffffffffffff8p0
tgamma -0x64.000000000000008p0
tgamma -0x95.ffffffffffffffp0
tgamma -0x96.00000000000001p0
tgamma -0xb4.ffffffffffffffp0
tgamma -0xb5.00000000000001p0
tgamma -0xb5.ffffffffffffffp0
tgamma -0xb6.00000000000001p0
tgamma -0xb6.ffffffffffffffp0
tgamma -0xb7.00000000000001p0
tgamma -0xb7.ffffffffffffffp0
tgamma -0xb8.00000000000001p0
tgamma -0xbb.ffffffffffffffp0
tgamma -0xbc.00000000000001p0
tgamma -0xbc.ffffffffffffffp0
tgamma -0xbd.00000000000001p0
tgamma -0xbd.ffffffffffffffp0
tgamma -0xbe.00000000000001p0
tgamma -0xbe.ffffffffffffffp0
tgamma -0xbf.00000000000001p0
tgamma -0xf9.ffffffffffffffp0
tgamma -0xfa.00000000000001p0
tgamma -0x1f3.fffffffffffffep0
tgamma -0x1f4.00000000000002p0
tgamma -0x2ed.fffffffffffffcp0
tgamma -0x2ee.00000000000004p0
tgamma -0x3e7.fffffffffffffcp0
tgamma -0x3e8.00000000000004p0
tgamma -0x4e1.fffffffffffff8p0
tgamma -0x4e2.00000000000008p0
tgamma -0x5db.fffffffffffff8p0
tgamma -0x5dc.00000000000008p0
tgamma -0x6d5.fffffffffffff8p0
tgamma -0x6d6.00000000000008p0
tgamma -0x6e2.fffffffffffff8p0
tgamma -0x6e3.00000000000008p0
tgamma -0x6e3.fffffffffffff8p0
tgamma -0x6e4.00000000000008p0
tgamma -0x6e4.fffffffffffff8p0
tgamma -0x6e5.00000000000008p0
tgamma -0x6e5.fffffffffffff8p0
tgamma -0x6e6.00000000000008p0
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma 0x8.0000000000000000000000000008p0
tgamma 0x7.fffffffffffffffffffffffffffcp0
tgamma 0x7.0000000000000000000000000004p0
tgamma 0x6.fffffffffffffffffffffffffffcp0
tgamma 0x6.0000000000000000000000000004p0
tgamma 0x5.fffffffffffffffffffffffffffcp0
tgamma 0x5.0000000000000000000000000004p0
tgamma 0x4.fffffffffffffffffffffffffffcp0
tgamma 0x4.0000000000000000000000000004p0
tgamma 0x3.fffffffffffffffffffffffffffep0
tgamma 0x3.0000000000000000000000000002p0
tgamma 0x2.fffffffffffffffffffffffffffep0
tgamma 0x2.0000000000000000000000000002p0
tgamma 0x1.ffffffffffffffffffffffffffffp0
tgamma 0x1.0000000000000000000000000001p0
tgamma 0x0.ffffffffffffffffffffffffffff8p0
tgamma -0x0.ffffffffffffffffffffffffffff8p0
tgamma -0x1.0000000000000000000000000001p0
tgamma -0x1.ffffffffffffffffffffffffffffp0
tgamma -0x2.0000000000000000000000000002p0
tgamma -0x2.fffffffffffffffffffffffffffep0
tgamma -0x3.0000000000000000000000000002p0
tgamma -0x3.fffffffffffffffffffffffffffep0
tgamma -0x4.0000000000000000000000000004p0
tgamma -0x4.fffffffffffffffffffffffffffcp0
tgamma -0x5.0000000000000000000000000004p0
tgamma -0x5.fffffffffffffffffffffffffffcp0
tgamma -0x6.0000000000000000000000000004p0
tgamma -0x6.fffffffffffffffffffffffffffcp0
tgamma -0x7.0000000000000000000000000004p0
tgamma -0x7.fffffffffffffffffffffffffffcp0
tgamma -0x8.0000000000000000000000000008p0
tgamma -0x9.fffffffffffffffffffffffffff8p0
tgamma -0xa.0000000000000000000000000008p0
tgamma -0x13.fffffffffffffffffffffffffffp0
tgamma -0x14.000000000000000000000000001p0
tgamma -0x1d.fffffffffffffffffffffffffffp0
tgamma -0x1e.000000000000000000000000001p0
tgamma -0x27.ffffffffffffffffffffffffffep0
tgamma -0x28.000000000000000000000000002p0
tgamma -0x28.ffffffffffffffffffffffffffep0
tgamma -0x29.000000000000000000000000002p0
Fix tgamma errno setting on underflow (bug 6810).
2013-12-05 14:01:41 +00:00
tgamma -0x29.ffffffffffffffffffffffffffep0
tgamma -0x2a.000000000000000000000000002p0
tgamma -0x31.ffffffffffffffffffffffffffep0
tgamma -0x32.000000000000000000000000002p0
tgamma -0x63.ffffffffffffffffffffffffffcp0
tgamma -0x64.000000000000000000000000004p0
tgamma -0x95.ffffffffffffffffffffffffff8p0
tgamma -0x96.000000000000000000000000008p0
tgamma -0xb4.ffffffffffffffffffffffffff8p0
tgamma -0xb5.000000000000000000000000008p0
tgamma -0xb5.ffffffffffffffffffffffffff8p0
tgamma -0xb6.000000000000000000000000008p0
tgamma -0xb6.ffffffffffffffffffffffffff8p0
tgamma -0xb7.000000000000000000000000008p0
tgamma -0xb7.ffffffffffffffffffffffffff8p0
tgamma -0xb8.000000000000000000000000008p0
tgamma -0xbb.ffffffffffffffffffffffffff8p0
tgamma -0xbc.000000000000000000000000008p0
tgamma -0xbc.ffffffffffffffffffffffffff8p0
tgamma -0xbd.000000000000000000000000008p0
tgamma -0xbd.ffffffffffffffffffffffffff8p0
tgamma -0xbe.000000000000000000000000008p0
tgamma -0xbe.ffffffffffffffffffffffffff8p0
tgamma -0xbf.000000000000000000000000008p0
tgamma -0xf9.ffffffffffffffffffffffffff8p0
tgamma -0xfa.000000000000000000000000008p0
tgamma -0x1f3.ffffffffffffffffffffffffffp0
tgamma -0x1f4.00000000000000000000000001p0
tgamma -0x2ed.fffffffffffffffffffffffffep0
tgamma -0x2ee.00000000000000000000000002p0
tgamma -0x3e7.fffffffffffffffffffffffffep0
tgamma -0x3e8.00000000000000000000000002p0
tgamma -0x4e1.fffffffffffffffffffffffffcp0
tgamma -0x4e2.00000000000000000000000004p0
tgamma -0x5db.fffffffffffffffffffffffffcp0
tgamma -0x5dc.00000000000000000000000004p0
tgamma -0x6d5.fffffffffffffffffffffffffcp0
tgamma -0x6d6.00000000000000000000000004p0
tgamma -0x6e2.fffffffffffffffffffffffffcp0
tgamma -0x6e3.00000000000000000000000004p0
tgamma -0x6e3.fffffffffffffffffffffffffcp0
tgamma -0x6e4.00000000000000000000000004p0
tgamma -0x6e4.fffffffffffffffffffffffffcp0
tgamma -0x6e5.00000000000000000000000004p0
tgamma -0x6e5.fffffffffffffffffffffffffcp0
tgamma -0x6e6.00000000000000000000000004p0
tgamma -0x6eb.fffffffffffffffffffffffffcp0
tgamma -0x6ec.00000000000000000000000004p0
tgamma -0x6ec.fffffffffffffffffffffffffcp0
tgamma -0x6ed.00000000000000000000000004p0
tgamma -0x6ed.fffffffffffffffffffffffffcp0
tgamma -0x6ee.00000000000000000000000004p0
tgamma -0x6ee.fffffffffffffffffffffffffcp0
tgamma -0x6ef.00000000000000000000000004p0
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma -0x1.0a32a2p+5
Fix tgamma errno setting on underflow (bug 6810).
2013-12-05 14:01:41 +00:00
tgamma -0x1.5800000080001p+7
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
tgamma 18.5
tgamma 19.5
tgamma 23.5
tgamma 29.5
tgamma 30.5
tgamma 31.5
tgamma 32.5
tgamma 33.5
tgamma 34.5
tgamma 0x2.30a43cp+4
tgamma 0x2.30a44p+4
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
tgamma 0xa.b9fd72b0fb238p+4
tgamma 0xa.b9fd72b0fb24p+4
tgamma 0xa.b9fd72b0fb23a9ddbf0d3804f4p+4
tgamma 0xa.b9fd72b0fb23a9ddbf0d3804f8p+4
tgamma 0x6.db8c603359a97108p+8
tgamma 0x6.db8c603359a9711p+8
tgamma 0x6.db8c603359a971081bc4a2e9dfdp+8
tgamma 0x6.db8c603359a971081bc4a2e9dfd4p+8
tgamma 1e3
Fix tgamma errno setting on underflow (bug 6810).
2013-12-05 14:01:41 +00:00
tgamma -100000.5
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
tgamma max
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
Fix tgamma missing underflows (bug 18951). Similar to various other bugs in this area, tgamma functions can fail to raise the underflow exception when the result is tiny and inexact but one or more low bits of the intermediate result that is scaled down are zero. This patch forces the exception in a similar way to previous fixes. Tested for x86_64, x86, mips64 and powerpc. [BZ #18951] * sysdeps/ieee754/dbl-64/e_gamma_r.c (__ieee754_gamma_r): Force underflow exception for small results. * sysdeps/ieee754/flt-32/e_gammaf_r.c (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (__ieee754_gammal_r): Likewise. * math/auto-libm-test-in: Add more tests of tgamma. * math/auto-libm-test-out: Regenerated.
2015-09-17 15:51:54 +00:00
tgamma -0x22.30p0
tgamma -0x22.31p0
tgamma -0x22.32p0
tgamma -0x22.33p0
tgamma -0x22.34p0
tgamma -0x22.35p0
tgamma -0x22.36p0
tgamma -0x22.37p0
tgamma -0xa3.70p0
tgamma -0xa3.71p0
tgamma -0xa3.72p0
tgamma -0xa3.73p0
tgamma -0xa3.74p0
tgamma -0xa3.75p0
tgamma -0xa3.76p0
tgamma -0xa3.77p0
tgamma -0xab.0d0p0
tgamma -0xab.0d1p0
tgamma -0xab.0d2p0
tgamma -0xab.0d3p0
tgamma -0xab.0d4p0
tgamma -0xab.0d5p0
tgamma -0xab.0d6p0
tgamma -0xab.0d7p0
tgamma -0x6db.030p0
tgamma -0x6db.031p0
tgamma -0x6db.032p0
tgamma -0x6db.033p0
tgamma -0x6db.034p0
tgamma -0x6db.035p0
tgamma -0x6db.036p0
tgamma -0x6db.037p0
tgamma -0x6db.050p0
tgamma -0x6db.051p0
tgamma -0x6db.052p0
tgamma -0x6db.053p0
tgamma -0x6db.054p0
tgamma -0x6db.055p0
tgamma -0x6db.056p0
tgamma -0x6db.057p0
Add more tests of tgamma. This patch adds some randomly-generated tests of tgamma that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 23:15:07 +00:00
tgamma -0x3.06644cp+0
tgamma -0x6.fe4636e0c5064p+0
tgamma -0x7.a13d7a2945cd5718p+0
Add more tests of libm functions. This patch adds more randomly-generated tests of various libm functions that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of atan, clog, clog10, cos, csqrt, erf, erfc, exp2, lgamma, log1p, sin, sincos, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-02 21:06:33 +00:00
tgamma -0x1.4a5caap+4
Add more tests of various libm functions. This patch adds more tests of various libm functions found through random test generation to give increased ulps on 32-bit x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acosh, asin, asinh, atanh, cabs, carg, cbrt, cosh, csqrt, erf, erfc, exp, exp10, expm1, hypot, log, log10, log1p, log2, pow, sinh, tan and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-11 00:58:28 +00:00
tgamma -0x9.2d3a5p+0
tgamma -0xb.0f63ep+0
tgamma -0x5.f0e02p+8
Add more random libm-test inputs. This patch adds more test inputs to various libm functions found through random generation to have larger ulps errors than previously listed in libm-test-ulp, on at least one of x86_64 and x86. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of acos, acosh, asin, asinh, atan, atan2, atanh, cabs, cbrt, cosh, csqrt, erf, erfc, exp, exp2, lgamma, log, log1p, log2, pow, sin, sincos, tan, tanh and tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-08-13 23:23:23 +00:00
tgamma -0xb.3123bp+0
tgamma -0x9.6d538p+0
tgamma -0xc.c2439p+0
tgamma -0xc.372f043322128p+0
tgamma -0xa.ccfcep+0
tgamma -0x9.418c8p+0
tgamma -0x6.ce9158p+0
tgamma -0xd.cbf53d0e7d06p+0
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
# the next value generates larger error bounds on x86_64 (binary32)
tgamma -0x3.0aa534p+0
Improve the accuracy of tgamma (BZ #26983) With this patch, the maximal known error for tgamma is now reduced to 9 ulps for dbl-64, for all rounding modes. Since exhaustive testing is not possible for dbl-64, it might be that there are still cases with an error larger than 9 ulps, but all known cases are fixed (intensive tests were done to find cases with large errors). Tested on x86_64 and powerpc (and by Adhemerval Zanella on aarch64, arm, s390x, sparc, and i686). Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-02 06:21:06 +00:00
# the next values generate large error bounds on x86_64 (binary64)
add inputs to auto-libm-test-in yielding larger errors (binary64, x86_64)
2020-11-25 18:03:33 +00:00
tgamma -0x1.62b8c36c7180bp+4
Improve the accuracy of tgamma (BZ #26983) With this patch, the maximal known error for tgamma is now reduced to 9 ulps for dbl-64, for all rounding modes. Since exhaustive testing is not possible for dbl-64, it might be that there are still cases with an error larger than 9 ulps, but all known cases are fixed (intensive tests were done to find cases with large errors). Tested on x86_64 and powerpc (and by Adhemerval Zanella on aarch64, arm, s390x, sparc, and i686). Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-02 06:21:06 +00:00
tgamma -0x1.62c4d519e8677p+3
tgamma -0x1.c033cc426752fp+2
tgamma -0x1.62cfd0d34ade2p+3
tgamma -0x1.8814da6eb7dbp+5
Add more tests of tgamma. This patch adds some randomly-generated tests of tgamma that are observed to increase ulps on x86_64. Tested for x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add more tests of tgamma. * math/auto-libm-test-out: Regenerated. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-05-01 23:15:07 +00:00
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
y0 0.125
y0 0.75
y0 1.0
y0 1.5
y0 2.0
y0 8.0
y0 10.0
y0 0x1.3ffp+74
y0 0x1.ff00000000002p+840
y0 0x1p1023
y0 0x1p16382
y0 0x1p16383
y0 0x1p-10
y0 0x1p-20
y0 0x1p-30
y0 0x1p-40
y0 0x1p-50
y0 0x1p-60
y0 0x1p-70
y0 0x1p-80
y0 0x1p-90
y0 0x1p-100
y0 0x1p-110
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
y0 0x1p-600
y0 0x1p-10000
Fix the inaccuracy of j0f/j1f/y0f/y1f [BZ #14469, #14470, #14471, #14472] For j0f/j1f/y0f/y1f, the largest error for all binary32 inputs is reduced to at most 9 ulps for all rounding modes. The new code is enabled only when there is a cancellation at the very end of the j0f/j1f/y0f/y1f computation, or for very large inputs, thus should not give any visible slowdown on average. Two different algorithms are used: * around the first 64 zeros of j0/j1/y0/y1, approximation polynomials of degree 3 are used, computed using the Sollya tool (https://www.sollya.org/) * for large inputs, an asymptotic formula from [1] is used [1] Fast and Accurate Bessel Function Computation, John Harrison, Proceedings of Arith 19, 2009. Inputs yielding the new largest errors are added to auto-libm-test-in, and ulps are regenerated for various targets (thanks Adhemerval Zanella). Tested on x86_64 with --disable-multi-arch and on powerpc64le-linux-gnu. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-01 06:14:10 +00:00
# the next values yield large errors for binary32
# (cf BZ #16492 for the xfail entries)
math: Add inputs that yield larger errors for float type (x86_64) The corner cases included were generated using exhaustive search for all float/binary32 values on x86_64 (comparing to MPFR for correct rounding to nearest). For the j0/j1/y0 functions, only cases with ulp error <= 9 were included. Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2020-03-30 13:55:55 +00:00
y0 0xd.3432bp-4
Fix the inaccuracy of j0f/j1f/y0f/y1f [BZ #14469, #14470, #14471, #14472] For j0f/j1f/y0f/y1f, the largest error for all binary32 inputs is reduced to at most 9 ulps for all rounding modes. The new code is enabled only when there is a cancellation at the very end of the j0f/j1f/y0f/y1f computation, or for very large inputs, thus should not give any visible slowdown on average. Two different algorithms are used: * around the first 64 zeros of j0/j1/y0/y1, approximation polynomials of degree 3 are used, computed using the Sollya tool (https://www.sollya.org/) * for large inputs, an asymptotic formula from [1] is used [1] Fast and Accurate Bessel Function Computation, John Harrison, Proceedings of Arith 19, 2009. Inputs yielding the new largest errors are added to auto-libm-test-in, and ulps are regenerated for various targets (thanks Adhemerval Zanella). Tested on x86_64 with --disable-multi-arch and on powerpc64le-linux-gnu. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-01 06:14:10 +00:00
y0 0x1.33eaacp+5 xfail:binary64 xfail:intel96 xfail-rounding:ibm128-libgcc
y0 0x1.a681cep-1 xfail-rounding:ibm128-libgcc
Set errno for y1 overflow (bug 17050). This patch fixes bug 17050, missing errno setting for y1 overflow (for small positive arguments). An appropriate check is added for overflow directly in the __ieee754_y1 implementation, similar to the check present for yn (doing it there rather than in the wrapper also avoids yn needing to repeat the check when called for order 1 or -1 and it uses __ieee754_y1). Tested x86_64 and x86; no ulps update needed. Also tested for mips64 to verify the ldbl-128 fix (the ldbl-128ibm code just #includes the ldbl-128 file). [BZ #17050] * sysdeps/ieee754/dbl-64/e_j1.c: Include <errno.h>. (__ieee754_y1): Set errno if return value overflows. * sysdeps/ieee754/flt-32/e_j1f.c: Include <errno.h>. (__ieee754_y1f): Set errno if return value overflows. * sysdeps/ieee754/ldbl-128/e_j1l.c: Include <errno.h>. (__ieee754_y1l): Set errno if return value overflows. * sysdeps/ieee754/ldbl-96/e_j1l.c: Include <errno.h>. (__ieee754_y1l): Set errno if return value overflows. * math/auto-libm-test-in: Add more tests of y0, y1 and yn. * math/auto-libm-test-out: Regenerated.
2014-06-23 20:17:13 +00:00
y0 min
y0 min_subnorm
Move TEST_f_f tests for [l-y]* functions from libm-test.inc to auto-libm-test-in.
2013-12-05 13:54:50 +00:00
y1 0.125
y1 0.75
y1 1.0
y1 1.5
y1 2.0
y1 8.0
y1 10.0
y1 0x1.27e204p+99
y1 0x1.001000001p+593
y1 0x1p1023
y1 0x1p16382
y1 0x1p16383
y1 0x1p-10
y1 0x1p-20
y1 0x1p-30
y1 0x1p-40
y1 0x1p-50
y1 0x1p-60
y1 0x1p-70
y1 0x1p-80
y1 0x1p-90
y1 0x1p-100
y1 0x1p-110
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
y1 0x1p-600
y1 0x1p-10000
Fix the inaccuracy of j0f/j1f/y0f/y1f [BZ #14469, #14470, #14471, #14472] For j0f/j1f/y0f/y1f, the largest error for all binary32 inputs is reduced to at most 9 ulps for all rounding modes. The new code is enabled only when there is a cancellation at the very end of the j0f/j1f/y0f/y1f computation, or for very large inputs, thus should not give any visible slowdown on average. Two different algorithms are used: * around the first 64 zeros of j0/j1/y0/y1, approximation polynomials of degree 3 are used, computed using the Sollya tool (https://www.sollya.org/) * for large inputs, an asymptotic formula from [1] is used [1] Fast and Accurate Bessel Function Computation, John Harrison, Proceedings of Arith 19, 2009. Inputs yielding the new largest errors are added to auto-libm-test-in, and ulps are regenerated for various targets (thanks Adhemerval Zanella). Tested on x86_64 with --disable-multi-arch and on powerpc64le-linux-gnu. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2021-04-01 06:14:10 +00:00
# the next three values yield the largest error in the binary32 format
# (cf BZ #27670 for the xfail entries)
y1 0x1.065194p+7 xfail-rounding:binary64 xfail-rounding:intel96 xfail-rounding:ibm128-libgcc
y1 0x1.c1badep+0 xfail-rounding:ibm128-libgcc
y1 0x1.c1bc2ep+0
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
y1 min
y1 min_subnorm
Move tests of jn and yn from libm-test.inc to auto-libm-test-in. This patch moves tests of jn and yn to auto-libm-test-in, adding the required support for gen-auto-libm-tests (and adding a missing assertion there and fixing logic that was broken for functions with integer arguments). Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of jn and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (jn_test_data): Use AUTO_TESTS_if_f. (yn_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_if_f. (func_calc_desc): Add mpfr_if_f union field. (FUNC_mpfr_if_f): New macro. (test_functions): Add jn and yn. (calc_generic_results): Assert type of second input for mpfr_ff_f. Handle mpfr_if_f. (output_for_one_input_case): Disable all checking for arguments fitting floating-point types in case of an integer argument. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-18 17:59:29 +00:00
# yn (0, x) == y0 (x).
yn 0 0.125
yn 0 0.75
yn 0 1.0
yn 0 1.5
yn 0 2.0
yn 0 8.0
yn 0 10.0
# yn (1, x) == y1 (x).
yn 1 0.125
yn 1 0.75
yn 1 1.0
yn 1 1.5
yn 1 2.0
yn 1 8.0
yn 1 10.0
# yn (-1, x) == -y1 (x).
yn -1 1.0
# yn (3, x).
yn 3 0.125
yn 3 0.75
yn 3 1.0
yn 3 2.0
yn 3 10.0
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
yn 3 0x1p-10
yn 3 0x1p-100
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
yn 3 0x1p-600
yn 3 0x1p-10000
Move tests of jn and yn from libm-test.inc to auto-libm-test-in. This patch moves tests of jn and yn to auto-libm-test-in, adding the required support for gen-auto-libm-tests (and adding a missing assertion there and fixing logic that was broken for functions with integer arguments). Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of jn and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (jn_test_data): Use AUTO_TESTS_if_f. (yn_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_if_f. (func_calc_desc): Add mpfr_if_f union field. (FUNC_mpfr_if_f): New macro. (test_functions): Add jn and yn. (calc_generic_results): Assert type of second input for mpfr_ff_f. Handle mpfr_if_f. (output_for_one_input_case): Disable all checking for arguments fitting floating-point types in case of an integer argument. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-18 17:59:29 +00:00
# yn (10, x).
yn 10 0.125
yn 10 0.75
yn 10 1.0
yn 10 2.0
yn 10 10.0
yn -10 1.0
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
yn 10 min
Move tests of jn and yn from libm-test.inc to auto-libm-test-in. This patch moves tests of jn and yn to auto-libm-test-in, adding the required support for gen-auto-libm-tests (and adding a missing assertion there and fixing logic that was broken for functions with integer arguments). Tested x86_64 and x86 and ulps updated accordingly. * math/auto-libm-test-in: Add tests of jn and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (jn_test_data): Use AUTO_TESTS_if_f. (yn_test_data): Likewise. * math/gen-auto-libm-tests.c (func_calc_method): Add value mpfr_if_f. (func_calc_desc): Add mpfr_if_f union field. (FUNC_mpfr_if_f): New macro. (test_functions): Add jn and yn. (calc_generic_results): Assert type of second input for mpfr_ff_f. Handle mpfr_if_f. (output_for_one_input_case): Disable all checking for arguments fitting floating-point types in case of an integer argument. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2013-12-18 17:59:29 +00:00
yn 2 0x1.ffff62p+99
yn 2 0x1p127
yn 2 0x1p1023
yn 2 0x1p16383
Add more libm tests (scalb*, signbit, sin, sincos, sinh, sqrt, tan, tanh, tgamma, y0, y1, yn, significand). This patch improves the libm test coverage for a few more functions. Tested for x86_64 and x86. * math/auto-libm-test-in: Add more tests of sin, sincos, sinh, sqrt, tan, tanh, y0, y1 and yn. * math/auto-libm-test-out: Regenerated. * math/libm-test.inc (scalb_test_data): Add more tests. (scalbn_test_data): Likewise. (scalbln_test_data): Likewise. (signbit_test_data): Likewise. (sin_test_data): Likewise. (sincos_test_data): Likewise. (sinh_test_data): Likewise. (sqrt_test_data): Likewise. (tan_test_data): Likewise. (tanh_test_data): Likewise. (tgamma_test_data): Likewise. (y0_test_data): Likewise. (y1_test_data): Likewise. (yn_test_data): Likewise. (significand_test_data): Likewise. * sysdeps/i386/fpu/libm-test-ulps: Update.
2015-11-04 00:45:23 +00:00
yn 2 0x1p-10
yn 2 0x1p-100
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
yn 2 0x1p-600
yn 2 0x1p-10000
Set errno for y1 overflow (bug 17050). This patch fixes bug 17050, missing errno setting for y1 overflow (for small positive arguments). An appropriate check is added for overflow directly in the __ieee754_y1 implementation, similar to the check present for yn (doing it there rather than in the wrapper also avoids yn needing to repeat the check when called for order 1 or -1 and it uses __ieee754_y1). Tested x86_64 and x86; no ulps update needed. Also tested for mips64 to verify the ldbl-128 fix (the ldbl-128ibm code just #includes the ldbl-128 file). [BZ #17050] * sysdeps/ieee754/dbl-64/e_j1.c: Include <errno.h>. (__ieee754_y1): Set errno if return value overflows. * sysdeps/ieee754/flt-32/e_j1f.c: Include <errno.h>. (__ieee754_y1f): Set errno if return value overflows. * sysdeps/ieee754/ldbl-128/e_j1l.c: Include <errno.h>. (__ieee754_y1l): Set errno if return value overflows. * sysdeps/ieee754/ldbl-96/e_j1l.c: Include <errno.h>. (__ieee754_y1l): Set errno if return value overflows. * math/auto-libm-test-in: Add more tests of y0, y1 and yn. * math/auto-libm-test-out: Regenerated.
2014-06-23 20:17:13 +00:00
yn 0 min
yn 0 min_subnorm
Improve libm-test XFAILing for ibm128-libgcc. This patch further improves XFAILing for ibm128-libgcc of tests in auto-libm-test-*. The bulk of the cases needing XFAILing are xfail-rounding:ibm128-libgcc used for inputs where (possibly after rounding the inputs to another floating-point type) the result overflows (and the result in non-default rounding modes may be wildly wrong with unpatched libgcc) or underflows near 0 (and the result in non-default rounding modes may end up having the wrong sign). This patch makes gen-auto-libm-tests detect such cases and apply xfail-rounding:ibm128-libgcc automatically to them, so most of the manual XFAILs in auto-libm-test-in are no longer needed (some are still needed if e.g. the result is very close to overflow, resulting in an internal overflow in libgcc in some rounding modes). A few manual XFAILs are added for cases not covered by this gen-auto-libm-tests change, and a few existing such XFAILs are left in. Tested for powerpc. * math/gen-auto-libm-tests.c (output_for_one_input_case): Apply xfail-rounding:ibm128-libgcc automatically to tests overflowing and those that can underflow to zero. * math/auto-libm-test-in: Remove most XFAILs for ibm128-libgcc and add others. * math/auto-libm-test-out: Regenerated.
2017-01-09 22:40:59 +00:00
yn 1 min
yn 1 min_subnorm
yn -1 min
yn -1 min_subnorm
yn 2 min
yn 2 min_subnorm
yn -2 min
yn -2 min_subnorm
yn 17 min
yn 17 min_subnorm
yn -17 min
yn -17 min_subnorm
yn 42 min
yn 42 min_subnorm
yn -42 min
yn -42 min_subnorm
Reference in New Issue Copy Permalink