1dnl AMD64 mpn_com. 2 3dnl Copyright 2004-2006, 2011, 2012 Free Software Foundation, Inc. 4 5dnl This file is part of the GNU MP Library. 6dnl 7dnl The GNU MP Library is free software; you can redistribute it and/or modify 8dnl it under the terms of either: 9dnl 10dnl * the GNU Lesser General Public License as published by the Free 11dnl Software Foundation; either version 3 of the License, or (at your 12dnl option) any later version. 13dnl 14dnl or 15dnl 16dnl * the GNU General Public License as published by the Free Software 17dnl Foundation; either version 2 of the License, or (at your option) any 18dnl later version. 19dnl 20dnl or both in parallel, as here. 21dnl 22dnl The GNU MP Library is distributed in the hope that it will be useful, but 23dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY 24dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 25dnl for more details. 26dnl 27dnl You should have received copies of the GNU General Public License and the 28dnl GNU Lesser General Public License along with the GNU MP Library. If not, 29dnl see https://www.gnu.org/licenses/. 30 31include(`../config.m4') 32 33 34C cycles/limb 35C AMD K8,K9 1.25 36C AMD K10 1.25 37C Intel P4 2.78 38C Intel core2 1.1 39C Intel corei 1.5 40C Intel atom ? 41C VIA nano 2 42 43C INPUT PARAMETERS 44define(`rp',`%rdi') 45define(`up',`%rsi') 46define(`n',`%rdx') 47 48ABI_SUPPORT(DOS64) 49ABI_SUPPORT(STD64) 50 51ASM_START() 52 TEXT 53 ALIGN(32) 54PROLOGUE(mpn_com) 55 FUNC_ENTRY(3) 56 movq (up), %r8 57 movl R32(%rdx), R32(%rax) 58 leaq (up,n,8), up 59 leaq (rp,n,8), rp 60 negq n 61 andl $3, R32(%rax) 62 je L(b00) 63 cmpl $2, R32(%rax) 64 jc L(b01) 65 je L(b10) 66 67L(b11): notq %r8 68 movq %r8, (rp,n,8) 69 decq n 70 jmp L(e11) 71L(b10): addq $-2, n 72 jmp L(e10) 73 .byte 0x90,0x90,0x90,0x90,0x90,0x90 74L(b01): notq %r8 75 movq %r8, (rp,n,8) 76 incq n 77 jz L(ret) 78 79L(oop): movq (up,n,8), %r8 80L(b00): movq 8(up,n,8), %r9 81 notq %r8 82 notq %r9 83 movq %r8, (rp,n,8) 84 movq %r9, 8(rp,n,8) 85L(e11): movq 16(up,n,8), %r8 86L(e10): movq 24(up,n,8), %r9 87 notq %r8 88 notq %r9 89 movq %r8, 16(rp,n,8) 90 movq %r9, 24(rp,n,8) 91 addq $4, n 92 jnc L(oop) 93L(ret): FUNC_EXIT() 94 ret 95EPILOGUE() 96