dnl AMD64 mpn_addlsh_n and mpn_rsblsh_n. R = V2^k +- U.
dnl Copyright 2006, 2010-2012 Free Software Foundation, Inc.
dnl This file is part of the GNU MP Library.
dnl
dnl The GNU MP Library is free software; you can redistribute it and/or modify
dnl it under the terms of either:
dnl
dnl * the GNU Lesser General Public License as published by the Free
dnl Software Foundation; either version 3 of the License, or (at your
dnl option) any later version.
dnl
dnl or
dnl
dnl * the GNU General Public License as published by the Free Software
dnl Foundation; either version 2 of the License, or (at your option) any
dnl later version.
dnl
dnl or both in parallel, as here.
dnl
dnl The GNU MP Library is distributed in the hope that it will be useful, but
dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
dnl for more details.
dnl
dnl You should have received copies of the GNU General Public License and the
dnl GNU Lesser General Public License along with the GNU MP Library. If not,
dnl see https://www.gnu.org/licenses/.
include(`../config.m4')
C cycles/limb
C AMD K8,K9 2.87 < 3.85for lshift + add_n
C AMD K10 2.75 < 3.85for lshift + add_n
C Intel P4 22 > 7.33for lshift + add_n
C Intel core2 4.1 > 3.27for lshift + add_n
C Intel NHM 4.4 > 3.75for lshift + add_n
C Intel SBR 3.17 < 3.46for lshift + add_n
C Intel atom ? ? 8.75for lshift + add_n
C VIA nano 4.7 < 6.25for lshift + add_n
C TODO
C * Can we propagate carry into rdx instead of using a special carry register?
C That could save enough insns to get to 10 cycles/iteration.
L(b0): mov %rax, %r10 mov %rdx, %r11 mov -8(vp,n,8), %rax mul %r8 or %rax, %r11
ADDSUB 16(up,n,8), %r10
ADCSBB 24(up,n,8), %r11 mov (vp,n,8), %rax
sbb R32(%rcx), R32(%rcx) mov %rdx, %rbx mul %r8 or %rax, %rbx mov8(vp,n,8), %rax add $4, n jz L(end)
ALIGN(8)
L(top): mov %rdx, %r9 mul %r8 or %rax, %r9 mov %r10, -16(rp,n,8)
L(lo3): mov %rdx, %r10 mov -16(vp,n,8), %rax mul %r8 or %rax, %r10 mov %r11, -8(rp,n,8)
L(lo2): mov %rdx, %r11 mov -8(vp,n,8), %rax mul %r8 or %rax, %r11 add R32(%rcx), R32(%rcx)
ADCSBB (up,n,8), %rbx
ADCSBB 8(up,n,8), %r9
ADCSBB 16(up,n,8), %r10
ADCSBB 24(up,n,8), %r11 mov (vp,n,8), %rax
sbb R32(%rcx), R32(%rcx) mov %rbx, (rp,n,8)
L(lo1): mov %rdx, %rbx mul %r8 or %rax, %rbx mov %r9, 8(rp,n,8)
L(lo0): mov8(vp,n,8), %rax add $4, n jnz L(top)
L(end): mov %rdx, %r9 mul %r8 or %rax, %r9 mov %r10, -16(rp,n,8)
L(cj3): mov %r11, -8(rp,n,8)
L(cj2): add R32(%rcx), R32(%rcx)
ADCSBB (up,n,8), %rbx
ADCSBB 8(up,n,8), %r9 mov %rbx, (rp,n,8)
L(cj1): mov %r9, 8(rp,n,8) mov %rdx, %rax
ADCSBB $0, %rax pop %rbx pop %rbp pop %r12
FUNC_EXIT() ret
EPILOGUE()
Messung V0.5 in Prozent
¤ Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.0.9Bemerkung:
(vorverarbeitet am 2026-06-04)
¤
Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.
Bemerkung:
Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.