From 42f07a44ef88e0c9ff1bd5343786013272467414 Mon Sep 17 00:00:00 2001 From: Adhemerval Zanella Date: Thu, 13 Nov 2025 09:58:19 -0300 Subject: math: Remove ldbl-96 fma implementation It is worse than the ldbl-64 version on recent x86 hardware. With Zen3 and gcc-15: ldbl-96 removal reciprocal-throughput master patched improvement x86_64 1176.2200 289.4640 4.06x i686 1476.0600 636.8660 2.32x latency master patched improvement x86_64 1176.2200 293.7360 4.00x i686 1480.0700 658.4160 2.25x Checked on x86_64-linux-gnu and i686-linux-gnu. Reviewed-by: Wilco Dijkstra --- sysdeps/i386/i686/multiarch/s_fma.c | 2 +- sysdeps/ieee754/ldbl-96/s_fma.c | 106 ------------------------------------ 2 files changed, 1 insertion(+), 107 deletions(-) delete mode 100644 sysdeps/ieee754/ldbl-96/s_fma.c diff --git a/sysdeps/i386/i686/multiarch/s_fma.c b/sysdeps/i386/i686/multiarch/s_fma.c index 70d0afe529..6dadb9439f 100644 --- a/sysdeps/i386/i686/multiarch/s_fma.c +++ b/sysdeps/i386/i686/multiarch/s_fma.c @@ -38,4 +38,4 @@ libm_alias_double_narrow (__fma, fma) #define __fma __fma_ia32 -#include +#include diff --git a/sysdeps/ieee754/ldbl-96/s_fma.c b/sysdeps/ieee754/ldbl-96/s_fma.c deleted file mode 100644 index a9abf8f1de..0000000000 --- a/sysdeps/ieee754/ldbl-96/s_fma.c +++ /dev/null @@ -1,106 +0,0 @@ -/* Compute x * y + z as ternary operation. - Copyright (C) 2010-2025 Free Software Foundation, Inc. - This file is part of the GNU C Library. - - The GNU C Library is free software; you can redistribute it and/or - modify it under the terms of the GNU Lesser General Public - License as published by the Free Software Foundation; either - version 2.1 of the License, or (at your option) any later version. - - The GNU C Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public - License along with the GNU C Library; if not, see - . */ - -#define NO_MATH_REDIRECT -#include -#define dfmal __hide_dfmal -#define f32xfmaf64 __hide_f32xfmaf64 -#include -#undef dfmal -#undef f32xfmaf64 -#include -#include -#include -#include -#include - -/* This implementation uses rounding to odd to avoid problems with - double rounding. See a paper by Boldo and Melquiond: - http://www.lri.fr/~melquion/doc/08-tc.pdf */ - -double -__fma (double x, double y, double z) -{ - if (__glibc_unlikely (!isfinite (x) || !isfinite (y))) - return x * y + z; - else if (__glibc_unlikely (!isfinite (z))) - /* If z is Inf, but x and y are finite, the result should be z - rather than NaN. */ - return (z + x) + y; - - /* Ensure correct sign of exact 0 + 0. */ - if (__glibc_unlikely ((x == 0 || y == 0) && z == 0)) - { - x = math_opt_barrier (x); - return x * y + z; - } - - fenv_t env; - __feholdexcept (&env); - __fesetround (FE_TONEAREST); - - /* Multiplication m1 + m2 = x * y using Dekker's algorithm. */ -#define C ((1ULL << (LDBL_MANT_DIG + 1) / 2) + 1) - long double x1 = (long double) x * C; - long double y1 = (long double) y * C; - long double m1 = (long double) x * y; - x1 = (x - x1) + x1; - y1 = (y - y1) + y1; - long double x2 = x - x1; - long double y2 = y - y1; - long double m2 = (((x1 * y1 - m1) + x1 * y2) + x2 * y1) + x2 * y2; - - /* Addition a1 + a2 = z + m1 using Knuth's algorithm. */ - long double a1 = z + m1; - long double t1 = a1 - z; - long double t2 = a1 - t1; - t1 = m1 - t1; - t2 = z - t2; - long double a2 = t1 + t2; - /* Ensure the arithmetic is not scheduled after feclearexcept call. */ - math_force_eval (m2); - math_force_eval (a2); - __feclearexcept (FE_INEXACT); - - /* If the result is an exact zero, ensure it has the correct sign. */ - if (a1 == 0 && m2 == 0) - { - __feupdateenv (&env); - /* Ensure that round-to-nearest value of z + m1 is not reused. */ - z = math_opt_barrier (z); - return z + m1; - } - - __fesetround (FE_TOWARDZERO); - /* Perform m2 + a2 addition with round to odd. */ - a2 = a2 + m2; - - /* Add that to a1 again using rounding to odd. */ - union ieee854_long_double u; - u.d = a1 + a2; - if ((u.ieee.mantissa1 & 1) == 0 && u.ieee.exponent != 0x7fff) - u.ieee.mantissa1 |= __fetestexcept (FE_INEXACT) != 0; - __feupdateenv (&env); - - /* Add finally round to double precision. */ - return u.d; -} -#ifndef __fma -libm_alias_double (__fma, fma) -libm_alias_double_narrow (__fma, fma) -#endif -- cgit v1.2.3