1dnl ARM64 mpn_cnd_add_n, mpn_cnd_sub_n 2 3dnl Contributed to the GNU project by Torbjörn Granlund. 4 5dnl Copyright 2012, 2013, 2017 Free Software Foundation, Inc. 6 7dnl This file is part of the GNU MP Library. 8dnl 9dnl The GNU MP Library is free software; you can redistribute it and/or modify 10dnl it under the terms of either: 11dnl 12dnl * the GNU Lesser General Public License as published by the Free 13dnl Software Foundation; either version 3 of the License, or (at your 14dnl option) any later version. 15dnl 16dnl or 17dnl 18dnl * the GNU General Public License as published by the Free Software 19dnl Foundation; either version 2 of the License, or (at your option) any 20dnl later version. 21dnl 22dnl or both in parallel, as here. 23dnl 24dnl The GNU MP Library is distributed in the hope that it will be useful, but 25dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY 26dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 27dnl for more details. 28dnl 29dnl You should have received copies of the GNU General Public License and the 30dnl GNU Lesser General Public License along with the GNU MP Library. If not, 31dnl see https://www.gnu.org/licenses/. 32 33include(`../config.m4') 34 35C cycles/limb 36C Cortex-A53 3.5-4 37C Cortex-A57 2.25 38C X-Gene 3.5 39 40changecom(blah) 41 42define(`cnd', `x0') 43define(`rp', `x1') 44define(`up', `x2') 45define(`vp', `x3') 46define(`n', `x4') 47 48ifdef(`OPERATION_cnd_add_n', ` 49 define(`ADDSUBC', adcs) 50 define(`CLRCY', `cmn xzr, xzr') 51 define(`RETVAL', `cset x0, cs') 52 define(`func', mpn_cnd_add_n)') 53ifdef(`OPERATION_cnd_sub_n', ` 54 define(`ADDSUBC', sbcs) 55 define(`CLRCY', `cmp xzr, xzr') 56 define(`RETVAL', `cset x0, cc') 57 define(`func', mpn_cnd_sub_n)') 58 59MULFUNC_PROLOGUE(mpn_cnd_add_n mpn_cnd_sub_n) 60 61ASM_START() 62PROLOGUE(func) 63 cmp cnd, #1 64 sbc cnd, cnd, cnd 65 66 CLRCY C really only needed for n = 0 (mod 4) 67 68 tbz n, #0, L(1) 69 ldr x10, [up], #8 70 ldr x12, [vp], #8 71 bic x6, x12, cnd 72 ADDSUBC x8, x10, x6 73 sub n, n, #1 74 str x8, [rp], #8 75 cbz n, L(rt) 76 77L(1): ldp x10, x11, [up], #16 78 ldp x12, x13, [vp], #16 79 sub n, n, #2 80 cbz n, L(end) 81 82L(top): bic x6, x12, cnd 83 bic x7, x13, cnd 84 ldp x12, x13, [vp], #16 85 ADDSUBC x8, x10, x6 86 ADDSUBC x9, x11, x7 87 ldp x10, x11, [up], #16 88 sub n, n, #2 89 stp x8, x9, [rp], #16 90 cbnz n, L(top) 91 92L(end): bic x6, x12, cnd 93 bic x7, x13, cnd 94 ADDSUBC x8, x10, x6 95 ADDSUBC x9, x11, x7 96 stp x8, x9, [rp] 97L(rt): RETVAL 98 ret 99EPILOGUE() 100