[PATCH 12/15] math: Use cospif from CORE-MATH
Paul Zimmermann
Paul.Zimmermann@inria.fr
Mon Feb 3 07:16:16 GMT 2025
I confirm all binary32 inputs yield correct rounding for all rounding modes on x86-64.
Paul
> From: Adhemerval Zanella <adhemerval.zanella@linaro.org>
> Cc: DJ Delorie <dj@redhat.com>,
> Joseph Myers <josmyers@redhat.com>,
> Paul Zimmermann <Paul.Zimmermann@inria.fr>,
> Alexei Sibidanov <sibid@uvic.ca>
> Date: Fri, 31 Jan 2025 16:17:16 -0300
>
> The CORE-MATH implementation is correctly rounded (for any rounding mode)
> and shows better performance to the generic cospif.
>
> The code was adapted to glibc style and to use the definition of
> math_config.h (to handle errno, overflow, and underflow).
>
> Benchtest on x64_64 (Ryzen 9 5900X, gcc 14.2.1), aarch64 (Neoverse-N1,
> gcc 13.3.1), and powerpc (POWER10, gcc 13.2.1):
>
> latency master patched improvement
> x86_64 47.4679 38.4157 19.07%
> x86_64v2 46.9686 38.3329 18.39%
> x86_64v3 43.8929 31.8510 27.43%
> aarch64 (Neoverse) 18.8867 13.2089 30.06%
> power8 22.9435 7.8023 65.99%
> power10 15.4472 7.77505 49.67%
>
> reciprocal-throughput master patched improvement
> x86_64 20.9518 11.4991 45.12%
> x86_64v2 19.8699 10.5921 46.69%
> x86_64v3 19.3475 9.3998 51.42%
> aarch64 (Neoverse) 12.5767 6.2158 50.58%
> power8 15.0566 3.2654 78.31%
> power10 9.2866 3.1147 66.46%
> ---
> SHARED-FILES | 4 +
> sysdeps/aarch64/libm-test-ulps | 4 -
> sysdeps/arc/fpu/libm-test-ulps | 4 -
> sysdeps/arc/nofpu/libm-test-ulps | 1 -
> sysdeps/arm/libm-test-ulps | 4 -
> sysdeps/hppa/fpu/libm-test-ulps | 4 -
> sysdeps/i386/fpu/libm-test-ulps | 4 -
> .../i386/i686/fpu/multiarch/libm-test-ulps | 4 -
> sysdeps/ieee754/flt-32/s_cospif.c | 136 ++++++++++++++++++
> sysdeps/loongarch/lp64/libm-test-ulps | 4 -
> sysdeps/mips/mips64/libm-test-ulps | 4 -
> sysdeps/or1k/fpu/libm-test-ulps | 4 -
> sysdeps/or1k/nofpu/libm-test-ulps | 1 -
> sysdeps/powerpc/fpu/libm-test-ulps | 4 -
> sysdeps/riscv/nofpu/libm-test-ulps | 1 -
> sysdeps/riscv/rvd/libm-test-ulps | 4 -
> sysdeps/s390/fpu/libm-test-ulps | 4 -
> sysdeps/sparc/fpu/libm-test-ulps | 4 -
> sysdeps/x86_64/fpu/libm-test-ulps | 4 -
> 19 files changed, 140 insertions(+), 59 deletions(-)
> create mode 100644 sysdeps/ieee754/flt-32/s_cospif.c
>
> diff --git a/SHARED-FILES b/SHARED-FILES
> index 5702a2d1c3..3ce38d1542 100644
> --- a/SHARED-FILES
> +++ b/SHARED-FILES
> @@ -350,3 +350,7 @@ sysdeps/ieee754/flt-32/s_atanpif.c:
> (src/binary32/atanpi/atanpif.c in CORE-MATH)
> - the code was adapted to use glibc code style and internal
> functions to handle errno, overflow, and underflow.
> +sysdeps/ieee754/flt-32/s_cospif.c:
> + (src/binary32/cospi/cospif.c in CORE-MATH)
> + - the code was adapted to use glibc code style and internal
> + functions to handle errno, overflow, and underflow.
> diff --git a/sysdeps/aarch64/libm-test-ulps b/sysdeps/aarch64/libm-test-ulps
> index 10f182a211..a15f3effa0 100644
> --- a/sysdeps/aarch64/libm-test-ulps
> +++ b/sysdeps/aarch64/libm-test-ulps
> @@ -782,7 +782,6 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> ldouble: 2
>
> Function: "cospi_advsimd":
> @@ -791,7 +790,6 @@ float: 1
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: "cospi_sve":
> @@ -800,12 +798,10 @@ float: 1
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: Real part of "cpow":
> diff --git a/sysdeps/arc/fpu/libm-test-ulps b/sysdeps/arc/fpu/libm-test-ulps
> index 7fb407cecd..f2acbf453e 100644
> --- a/sysdeps/arc/fpu/libm-test-ulps
> +++ b/sysdeps/arc/fpu/libm-test-ulps
> @@ -553,19 +553,15 @@ double: 3
>
> Function: "cospi":
> double: 2
> -float: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
>
> Function: "cospi_upward":
> double: 1
> -float: 2
>
> Function: Real part of "cpow":
> double: 9
> diff --git a/sysdeps/arc/nofpu/libm-test-ulps b/sysdeps/arc/nofpu/libm-test-ulps
> index 1859c2bd4f..8716e5d29e 100644
> --- a/sysdeps/arc/nofpu/libm-test-ulps
> +++ b/sysdeps/arc/nofpu/libm-test-ulps
> @@ -134,7 +134,6 @@ double: 2
>
> Function: "cospi":
> double: 2
> -float: 2
>
> Function: Real part of "cpow":
> double: 2
> diff --git a/sysdeps/arm/libm-test-ulps b/sysdeps/arm/libm-test-ulps
> index fa9864adee..647f92944c 100644
> --- a/sysdeps/arm/libm-test-ulps
> +++ b/sysdeps/arm/libm-test-ulps
> @@ -545,19 +545,15 @@ double: 2
>
> Function: "cospi":
> double: 2
> -float: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
>
> Function: "cospi_upward":
> double: 1
> -float: 2
>
> Function: Real part of "cpow":
> double: 2
> diff --git a/sysdeps/hppa/fpu/libm-test-ulps b/sysdeps/hppa/fpu/libm-test-ulps
> index a59f61fc4e..88f7701c0e 100644
> --- a/sysdeps/hppa/fpu/libm-test-ulps
> +++ b/sysdeps/hppa/fpu/libm-test-ulps
> @@ -555,19 +555,15 @@ double: 2
>
> Function: "cospi":
> double: 2
> -float: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
>
> Function: "cospi_upward":
> double: 1
> -float: 2
>
> Function: Real part of "cpow":
> double: 2
> diff --git a/sysdeps/i386/fpu/libm-test-ulps b/sysdeps/i386/fpu/libm-test-ulps
> index 8aa52f4600..39066956b0 100644
> --- a/sysdeps/i386/fpu/libm-test-ulps
> +++ b/sysdeps/i386/fpu/libm-test-ulps
> @@ -854,25 +854,21 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> float128: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> float128: 2
> ldouble: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> float128: 2
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> float128: 2
> ldouble: 2
>
> diff --git a/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps b/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
> index 8032636808..a8c4723850 100644
> --- a/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
> +++ b/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
> @@ -854,25 +854,21 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> float128: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> float128: 2
> ldouble: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> float128: 2
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> float128: 2
> ldouble: 2
>
> diff --git a/sysdeps/ieee754/flt-32/s_cospif.c b/sysdeps/ieee754/flt-32/s_cospif.c
> new file mode 100644
> index 0000000000..d4c652f8c0
> --- /dev/null
> +++ b/sysdeps/ieee754/flt-32/s_cospif.c
> @@ -0,0 +1,136 @@
> +/* Correctly-rounded cosine of binary32 value for angles in half-revolutions
> +
> +Copyright (c) 2022-2025 Alexei Sibidanov.
> +
> +The original version of this file was copied from the CORE-MATH
> +project (src/binary32/cospi/cospif.c, revision f786e13).
> +
> +Permission is hereby granted, free of charge, to any person obtaining a copy
> +of this software and associated documentation files (the "Software"), to deal
> +in the Software without restriction, including without limitation the rights
> +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> +copies of the Software, and to permit persons to whom the Software is
> +furnished to do so, subject to the following conditions:
> +
> +The above copyright notice and this permission notice shall be included in all
> +copies or substantial portions of the Software.
> +
> +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
> +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
> +SOFTWARE.
> +
> +*/
> +
> +#include <math.h>
> +#include <stdint.h>
> +#include <libm-alias-float.h>
> +#include "math_config.h"
> +
> +float
> +__cospif (float x)
> +{
> + static const double sn[] =
> + {
> + 0x1.921fb54442d0fp-37, -0x1.4abbce6102b94p-112, 0x1.4669fa3c58463p-189
> + };
> + static const double cn[] =
> + {
> + -0x1.3bd3cc9be45cfp-74, 0x1.03c1f08088742p-150, -0x1.55d1e5eff55a5p-228
> + };
> + /* S[i] approximates sin(i*pi/2^6) */
> + static const double S[] =
> + {
> + 0x0p+0, 0x1.91f65f10dd814p-5, 0x1.917a6bc29b42cp-4,
> + 0x1.2c8106e8e613ap-3, 0x1.8f8b83c69a60bp-3, 0x1.f19f97b215f1bp-3,
> + 0x1.294062ed59f06p-2, 0x1.58f9a75ab1fddp-2, 0x1.87de2a6aea963p-2,
> + 0x1.b5d1009e15ccp-2, 0x1.e2b5d3806f63bp-2, 0x1.073879922ffeep-1,
> + 0x1.1c73b39ae68c8p-1, 0x1.30ff7fce17035p-1, 0x1.44cf325091dd6p-1,
> + 0x1.57d69348cecap-1, 0x1.6a09e667f3bcdp-1, 0x1.7b5df226aafafp-1,
> + 0x1.8bc806b151741p-1, 0x1.9b3e047f38741p-1, 0x1.a9b66290ea1a3p-1,
> + 0x1.b728345196e3ep-1, 0x1.c38b2f180bdb1p-1, 0x1.ced7af43cc773p-1,
> + 0x1.d906bcf328d46p-1, 0x1.e212104f686e5p-1, 0x1.e9f4156c62ddap-1,
> + 0x1.f0a7efb9230d7p-1, 0x1.f6297cff75cbp-1, 0x1.fa7557f08a517p-1,
> + 0x1.fd88da3d12526p-1, 0x1.ff621e3796d7ep-1, 0x1p+0,
> + 0x1.ff621e3796d7ep-1, 0x1.fd88da3d12526p-1, 0x1.fa7557f08a517p-1,
> + 0x1.f6297cff75cbp-1, 0x1.f0a7efb9230d7p-1, 0x1.e9f4156c62ddap-1,
> + 0x1.e212104f686e5p-1, 0x1.d906bcf328d46p-1, 0x1.ced7af43cc773p-1,
> + 0x1.c38b2f180bdb1p-1, 0x1.b728345196e3ep-1, 0x1.a9b66290ea1a3p-1,
> + 0x1.9b3e047f38741p-1, 0x1.8bc806b151741p-1, 0x1.7b5df226aafafp-1,
> + 0x1.6a09e667f3bcdp-1, 0x1.57d69348cecap-1, 0x1.44cf325091dd6p-1,
> + 0x1.30ff7fce17035p-1, 0x1.1c73b39ae68c8p-1, 0x1.073879922ffeep-1,
> + 0x1.e2b5d3806f63bp-2, 0x1.b5d1009e15ccp-2, 0x1.87de2a6aea963p-2,
> + 0x1.58f9a75ab1fddp-2, 0x1.294062ed59f06p-2, 0x1.f19f97b215f1bp-3,
> + 0x1.8f8b83c69a60bp-3, 0x1.2c8106e8e613ap-3, 0x1.917a6bc29b42cp-4,
> + 0x1.91f65f10dd814p-5, 0x0p+0, -0x1.91f65f10dd814p-5,
> + -0x1.917a6bc29b42cp-4, -0x1.2c8106e8e613ap-3, -0x1.8f8b83c69a60bp-3,
> + -0x1.f19f97b215f1bp-3, -0x1.294062ed59f06p-2, -0x1.58f9a75ab1fddp-2,
> + -0x1.87de2a6aea963p-2, -0x1.b5d1009e15ccp-2, -0x1.e2b5d3806f63bp-2,
> + -0x1.073879922ffeep-1, -0x1.1c73b39ae68c8p-1, -0x1.30ff7fce17035p-1,
> + -0x1.44cf325091dd6p-1, -0x1.57d69348cecap-1, -0x1.6a09e667f3bcdp-1,
> + -0x1.7b5df226aafafp-1, -0x1.8bc806b151741p-1, -0x1.9b3e047f38741p-1,
> + -0x1.a9b66290ea1a3p-1, -0x1.b728345196e3ep-1, -0x1.c38b2f180bdb1p-1,
> + -0x1.ced7af43cc773p-1, -0x1.d906bcf328d46p-1, -0x1.e212104f686e5p-1,
> + -0x1.e9f4156c62ddap-1, -0x1.f0a7efb9230d7p-1, -0x1.f6297cff75cbp-1,
> + -0x1.fa7557f08a517p-1, -0x1.fd88da3d12526p-1, -0x1.ff621e3796d7ep-1,
> + -0x1p+0, -0x1.ff621e3796d7ep-1, -0x1.fd88da3d12526p-1,
> + -0x1.fa7557f08a517p-1, -0x1.f6297cff75cbp-1, -0x1.f0a7efb9230d7p-1,
> + -0x1.e9f4156c62ddap-1, -0x1.e212104f686e5p-1, -0x1.d906bcf328d46p-1,
> + -0x1.ced7af43cc773p-1, -0x1.c38b2f180bdb1p-1, -0x1.b728345196e3ep-1,
> + -0x1.a9b66290ea1a3p-1, -0x1.9b3e047f38741p-1, -0x1.8bc806b151741p-1,
> + -0x1.7b5df226aafafp-1, -0x1.6a09e667f3bcdp-1, -0x1.57d69348cecap-1,
> + -0x1.44cf325091dd6p-1, -0x1.30ff7fce17035p-1, -0x1.1c73b39ae68c8p-1,
> + -0x1.073879922ffeep-1, -0x1.e2b5d3806f63bp-2, -0x1.b5d1009e15ccp-2,
> + -0x1.87de2a6aea963p-2, -0x1.58f9a75ab1fddp-2, -0x1.294062ed59f06p-2,
> + -0x1.f19f97b215f1bp-3, -0x1.8f8b83c69a60bp-3, -0x1.2c8106e8e613ap-3,
> + -0x1.917a6bc29b42cp-4, -0x1.91f65f10dd814p-5
> + };
> +
> + uint32_t ix = asuint (x);
> + int32_t e = (ix >> 23) & 0xff;
> + if (__glibc_unlikely (e == 0xff))
> + {
> + if (!(ix << 9))
> + return __math_invalidf (x);
> + return x + x; /* nan */
> + }
> + int32_t m = (ix & ~0u >> 9) | 1 << 23;
> + int32_t s = 143 - e;
> + int32_t p = e - 112;
> + if (__glibc_unlikely (p < 0)) /* |x| < 2^-15 */
> + {
> + uint32_t ax = ix & (~0u>>1);
> + /* Warning: -0x1.3bd3ccp+2f * x underflows for |x| < 0x1.9f03p-129 */
> + if (ax >= 0x19f030u)
> + return fmaf (-0x1.3bd3ccp+2f * x, x, 1.0f);
> + else /* |x| < 0x1.9f03p-129 */
> + return fmaf (-x, x, 1.0f);
> + }
> + if (__glibc_unlikely (p > 31))
> + {
> + if (__glibc_unlikely (p > 63))
> + return 1.0f;
> + int32_t iq = m << (p - 32);
> + return S[(iq + 32) & 127];
> + }
> + int32_t k = m << p;
> + if (__glibc_unlikely (k == 0))
> + {
> + int32_t iq = m >> (32 - p);
> + return S[(iq + 32) & 127];
> + }
> + double z = k;
> + double z2 = z * z;
> + double fs = sn[0] + z2 * (sn[1] + z2 * sn[2]);
> + double fc = cn[0] + z2 * (cn[1] + z2 * cn[2]);
> + uint32_t iq = m >> s;
> + iq = (iq + 1) >> 1;
> + uint32_t is = iq & 127, ic = (iq + 32) & 127;
> + double ts = S[ic], tc = S[is];
> + double r = ts + (ts * z2) * fc - (tc * z) * fs;
> + return r;
> +}
> +libm_alias_float (__cospi, cospi)
> diff --git a/sysdeps/loongarch/lp64/libm-test-ulps b/sysdeps/loongarch/lp64/libm-test-ulps
> index 0cac55cbe4..fe84c60913 100644
> --- a/sysdeps/loongarch/lp64/libm-test-ulps
> +++ b/sysdeps/loongarch/lp64/libm-test-ulps
> @@ -701,22 +701,18 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: Real part of "cpow":
> diff --git a/sysdeps/mips/mips64/libm-test-ulps b/sysdeps/mips/mips64/libm-test-ulps
> index 1b5bcff11e..ddc78d0239 100644
> --- a/sysdeps/mips/mips64/libm-test-ulps
> +++ b/sysdeps/mips/mips64/libm-test-ulps
> @@ -701,22 +701,18 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: Real part of "cpow":
> diff --git a/sysdeps/or1k/fpu/libm-test-ulps b/sysdeps/or1k/fpu/libm-test-ulps
> index a608e3c949..884b4cc361 100644
> --- a/sysdeps/or1k/fpu/libm-test-ulps
> +++ b/sysdeps/or1k/fpu/libm-test-ulps
> @@ -545,19 +545,15 @@ double: 2
>
> Function: "cospi":
> double: 2
> -float: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
>
> Function: "cospi_upward":
> double: 1
> -float: 2
>
> Function: Real part of "cpow":
> double: 2
> diff --git a/sysdeps/or1k/nofpu/libm-test-ulps b/sysdeps/or1k/nofpu/libm-test-ulps
> index 56986f0be0..aec66e0fa3 100644
> --- a/sysdeps/or1k/nofpu/libm-test-ulps
> +++ b/sysdeps/or1k/nofpu/libm-test-ulps
> @@ -509,7 +509,6 @@ double: 2
>
> Function: "cospi":
> double: 2
> -float: 2
>
> Function: Real part of "cpow":
> double: 2
> diff --git a/sysdeps/powerpc/fpu/libm-test-ulps b/sysdeps/powerpc/fpu/libm-test-ulps
> index 630111e6c4..bdf0c98dc7 100644
> --- a/sysdeps/powerpc/fpu/libm-test-ulps
> +++ b/sysdeps/powerpc/fpu/libm-test-ulps
> @@ -858,25 +858,21 @@ ldouble: 2
>
> Function: "cospi":
> double: 2
> -float: 2
> float128: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> float128: 2
> ldouble: 4
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> float128: 2
> ldouble: 6
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> float128: 2
> ldouble: 6
>
> diff --git a/sysdeps/riscv/nofpu/libm-test-ulps b/sysdeps/riscv/nofpu/libm-test-ulps
> index 087dcd79fc..08af2495f3 100644
> --- a/sysdeps/riscv/nofpu/libm-test-ulps
> +++ b/sysdeps/riscv/nofpu/libm-test-ulps
> @@ -650,7 +650,6 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> ldouble: 2
>
> Function: Real part of "cpow":
> diff --git a/sysdeps/riscv/rvd/libm-test-ulps b/sysdeps/riscv/rvd/libm-test-ulps
> index efd83affa4..6644e38ebc 100644
> --- a/sysdeps/riscv/rvd/libm-test-ulps
> +++ b/sysdeps/riscv/rvd/libm-test-ulps
> @@ -709,22 +709,18 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: Real part of "cpow":
> diff --git a/sysdeps/s390/fpu/libm-test-ulps b/sysdeps/s390/fpu/libm-test-ulps
> index 709debb205..6318760eb5 100644
> --- a/sysdeps/s390/fpu/libm-test-ulps
> +++ b/sysdeps/s390/fpu/libm-test-ulps
> @@ -701,22 +701,18 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: Real part of "cpow":
> diff --git a/sysdeps/sparc/fpu/libm-test-ulps b/sysdeps/sparc/fpu/libm-test-ulps
> index becf5da3d6..2c319f8ae2 100644
> --- a/sysdeps/sparc/fpu/libm-test-ulps
> +++ b/sysdeps/sparc/fpu/libm-test-ulps
> @@ -701,22 +701,18 @@ ldouble: 3
>
> Function: "cospi":
> double: 2
> -float: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> ldouble: 2
>
> Function: Real part of "cpow":
> diff --git a/sysdeps/x86_64/fpu/libm-test-ulps b/sysdeps/x86_64/fpu/libm-test-ulps
> index 8c5d4fd471..e2cf3e04b6 100644
> --- a/sysdeps/x86_64/fpu/libm-test-ulps
> +++ b/sysdeps/x86_64/fpu/libm-test-ulps
> @@ -1050,25 +1050,21 @@ float: 2
>
> Function: "cospi":
> double: 2
> -float: 2
> float128: 2
> ldouble: 2
>
> Function: "cospi_downward":
> double: 1
> -float: 2
> float128: 2
> ldouble: 2
>
> Function: "cospi_towardzero":
> double: 1
> -float: 1
> float128: 2
> ldouble: 2
>
> Function: "cospi_upward":
> double: 1
> -float: 2
> float128: 2
> ldouble: 2
>
> --
> 2.43.0
>
>
More information about the Libc-alpha
mailing list