1 dnl x86 mpn_gcd_11 optimised for processors with fast BSF. 2 3 dnl Based on the K7 gcd_1.asm, by Kevin Ryde. Rehacked by Torbjorn Granlund. 4 5 dnl Copyright 2000-2002, 2005, 2009, 2011, 2012, 2015 Free Software 6 dnl Foundation, Inc. 7 8 dnl This file is part of the GNU MP Library. 9 dnl 10 dnl The GNU MP Library is free software; you can redistribute it and/or modify 11 dnl it under the terms of either: 12 dnl 13 dnl * the GNU Lesser General Public License as published by the Free 14 dnl Software Foundation; either version 3 of the License, or (at your 15 dnl option) any later version. 16 dnl 17 dnl or 18 dnl 19 dnl * the GNU General Public License as published by the Free Software 20 dnl Foundation; either version 2 of the License, or (at your option) any 21 dnl later version. 22 dnl 23 dnl or both in parallel, as here. 24 dnl 25 dnl The GNU MP Library is distributed in the hope that it will be useful, but 26 dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY 27 dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 28 dnl for more details. 29 dnl 30 dnl You should have received copies of the GNU General Public License and the 31 dnl GNU Lesser General Public License along with the GNU MP Library. If not, 32 dnl see https://www.gnu.org/licenses/. 33 34 include(`../config.m4') 35 36 37 C cycles/bit (approx) 38 C AMD K7 7.80 39 C AMD K8,K9 7.79 40 C AMD K10 4.08 41 C AMD bd1 ? 42 C AMD bobcat 7.82 43 C Intel P4-2 14.9 44 C Intel P4-3/4 14.0 45 C Intel P6/13 5.09 46 C Intel core2 4.22 47 C Intel NHM 5.00 48 C Intel SBR 5.00 49 C Intel atom 17.1 50 C VIA nano ? 51 C Numbers measured with: speed -CD -s16-32 -t16 mpn_gcd_1 52 53 54 define(`u0', `%eax') 55 define(`v0', `%edx') 56 57 ASM_START() 58 TEXT 59 ALIGN(16) 60 PROLOGUE(mpn_gcd_11) 61 push %edi 62 push %esi 63 64 mov 12(%esp), %eax 65 mov 16(%esp), %edx 66 jmp L(odd) 67 68 ALIGN(16) C K10 BD C2 NHM SBR 69 L(top): cmovc( %esi, %eax) C u = |v - u| 0,3 0,3 0,6 0,5 0,5 70 cmovc( %edi, %edx) C v = min(u,v) 0,3 0,3 2,8 1,7 1,7 71 shr %cl, %eax C 1,7 1,6 2,8 2,8 2,8 72 L(odd): mov %edx, %esi C 1 1 4 3 3 73 sub %eax, %esi C 2 2 5 4 4 74 bsf %esi, %ecx C 3 3 6 5 5 75 mov %eax, %edi C 2 2 3 3 4 76 sub %edx, %eax C 2 2 4 3 4 77 jnz L(top) C 78 79 L(end): mov %edx, %eax 80 pop %esi 81 pop %edi 82 ret 83 EPILOGUE() 84