1dnl S/390-32/esame mpn_mul_basecase. 2 3dnl Copyright 2011 Free Software Foundation, Inc. 4 5dnl This file is part of the GNU MP Library. 6 7dnl The GNU MP Library is free software; you can redistribute it and/or modify 8dnl it under the terms of the GNU Lesser General Public License as published 9dnl by the Free Software Foundation; either version 3 of the License, or (at 10dnl your option) any later version. 11 12dnl The GNU MP Library is distributed in the hope that it will be useful, but 13dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY 14dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public 15dnl License for more details. 16 17dnl You should have received a copy of the GNU Lesser General Public License 18dnl along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. 19 20include(`../config.m4') 21 22C cycles/limb 23C z900 ? 24C z990 ? 25C z9 ? 26C z10 ? 27C z196 ? 28 29C TODO 30C * Perhaps add special case for un <= 2. 31C * Replace loops by faster code. The mul_1 and addmul_1 loops could be sped 32C up by about 10%. 33 34C INPUT PARAMETERS 35define(`rp', `%r2') 36define(`up', `%r3') 37define(`un', `%r4') 38define(`vp', `%r5') 39define(`vn', `%r6') 40 41define(`zero', `%r8') 42 43ASM_START() 44PROLOGUE(mpn_mul_basecase) 45 chi un, 2 46 jhe L(ge2) 47 48C un = vn = 1 49 l %r1, 0(vp) 50 ml %r0, 0(up) 51 st %r1, 0(rp) 52 st %r0, 4(rp) 53 br %r14 54 55L(ge2): C jne L(gen) 56 57 58L(gen): 59C mul_1 ======================================================================= 60 61 stm %r6, %r12, 24(%r15) 62 lhi zero, 0 63 ahi un, -1 64 65 l %r7, 0(vp) 66 l %r11, 0(up) 67 lhi %r12, 4 C init index register 68 mlr %r10, %r7 69 lr %r9, un 70 st %r11, 0(rp) 71 cr %r15, %r15 C clear carry flag 72 73L(tm): l %r1, 0(%r12,up) 74 mlr %r0, %r7 75 alcr %r1, %r10 76 lr %r10, %r0 C copy high part to carry limb 77 st %r1, 0(%r12,rp) 78 la %r12, 4(%r12) 79 brct %r9, L(tm) 80 81 alcr %r0, zero 82 st %r0, 0(%r12,rp) 83 84C addmul_1 loop =============================================================== 85 86 ahi vn, -1 87 je L(outer_end) 88L(outer_loop): 89 90 la rp, 4(rp) C rp += 1 91 la vp, 4(vp) C up += 1 92 l %r7, 0(vp) 93 l %r11, 0(up) 94 lhi %r12, 4 C init index register 95 mlr %r10, %r7 96 lr %r9, un 97 al %r11, 0(rp) 98 st %r11, 0(rp) 99 100L(tam): l %r1, 0(%r12,up) 101 l %r11, 0(%r12,rp) 102 mlr %r0, %r7 103 alcr %r1, %r11 104 alcr %r0, zero 105 alr %r1, %r10 106 lr %r10, %r0 107 st %r1, 0(%r12,rp) 108 la %r12, 4(%r12) 109 brct %r9, L(tam) 110 111 alcr %r0, zero 112 st %r0, 0(%r12,rp) 113 114 brct vn, L(outer_loop) 115L(outer_end): 116 117 lm %r6, %r12, 24(%r15) 118 br %r14 119EPILOGUE() 120