1 1.1 mrg dnl SPARC v9 32-bit mpn_addmul_1 -- Multiply a limb vector with a limb and add 2 1.1 mrg dnl the result to a second limb vector. 3 1.1 mrg 4 1.1 mrg dnl Copyright 1998, 2000, 2001, 2003 Free Software Foundation, Inc. 5 1.1 mrg 6 1.1 mrg dnl This file is part of the GNU MP Library. 7 1.1.1.2 mrg dnl 8 1.1 mrg dnl The GNU MP Library is free software; you can redistribute it and/or modify 9 1.1.1.2 mrg dnl it under the terms of either: 10 1.1.1.2 mrg dnl 11 1.1.1.2 mrg dnl * the GNU Lesser General Public License as published by the Free 12 1.1.1.2 mrg dnl Software Foundation; either version 3 of the License, or (at your 13 1.1.1.2 mrg dnl option) any later version. 14 1.1.1.2 mrg dnl 15 1.1.1.2 mrg dnl or 16 1.1.1.2 mrg dnl 17 1.1.1.2 mrg dnl * the GNU General Public License as published by the Free Software 18 1.1.1.2 mrg dnl Foundation; either version 2 of the License, or (at your option) any 19 1.1.1.2 mrg dnl later version. 20 1.1.1.2 mrg dnl 21 1.1.1.2 mrg dnl or both in parallel, as here. 22 1.1.1.2 mrg dnl 23 1.1 mrg dnl The GNU MP Library is distributed in the hope that it will be useful, but 24 1.1 mrg dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY 25 1.1.1.2 mrg dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 26 1.1.1.2 mrg dnl for more details. 27 1.1.1.2 mrg dnl 28 1.1.1.2 mrg dnl You should have received copies of the GNU General Public License and the 29 1.1.1.2 mrg dnl GNU Lesser General Public License along with the GNU MP Library. If not, 30 1.1.1.2 mrg dnl see https://www.gnu.org/licenses/. 31 1.1 mrg 32 1.1 mrg include(`../config.m4') 33 1.1 mrg 34 1.1 mrg C Algorithm: We use two floating-point multiplies per limb product, with the 35 1.1 mrg C invariant v operand split into two 16-bit pieces, and the u operand split 36 1.1 mrg C into 32-bit pieces. We convert the two 48-bit products and transfer them to 37 1.1 mrg C the integer unit. 38 1.1 mrg 39 1.1 mrg C cycles/limb 40 1.1 mrg C UltraSPARC 1&2: 6.5 41 1.1 mrg C UltraSPARC 3: ? 42 1.1 mrg 43 1.1 mrg C Possible optimizations: 44 1.1 mrg C 1. Combine 32-bit memory operations into 64-bit operations. Since we're 45 1.1 mrg C memory bandwidth limited, this could save 1.5 cycles/limb. 46 1.1 mrg C 2. Unroll the inner loop. Since we already use alternate temporary areas, 47 1.1 mrg C it is very straightforward to unroll, using an exit branch midways. 48 1.1 mrg C Unrolling would allow deeper scheduling which could improve speed for L2 49 1.1 mrg C cache case. 50 1.1 mrg C 3. For mpn_mul_1: Use more alternating temp areas. The std'es and ldx'es 51 1.1 mrg C aren't sufficiently apart-scheduled with just two temp areas. 52 1.1 mrg C 4. Specialize for particular v values. If its upper 16 bits are zero, we 53 1.1 mrg C could save many operations. 54 1.1 mrg 55 1.1 mrg C INPUT PARAMETERS 56 1.1 mrg C rp i0 57 1.1 mrg C up i1 58 1.1 mrg C n i2 59 1.1 mrg C v i3 60 1.1 mrg 61 1.1 mrg define(`FSIZE',224) 62 1.1 mrg 63 1.1 mrg ASM_START() 64 1.1 mrg PROLOGUE(mpn_addmul_1) 65 1.1 mrg add %sp, -FSIZE, %sp 66 1.1 mrg sethi %hi(0xffff), %g1 67 1.1 mrg srl %o3, 16, %g2 68 1.1 mrg or %g1, %lo(0xffff), %g1 69 1.1 mrg and %o3, %g1, %g1 70 1.1 mrg stx %g1, [%sp+104] 71 1.1 mrg stx %g2, [%sp+112] 72 1.1 mrg ldd [%sp+104], %f6 73 1.1 mrg ldd [%sp+112], %f8 74 1.1 mrg fxtod %f6, %f6 75 1.1 mrg fxtod %f8, %f8 76 1.1 mrg ld [%sp+104], %f10 C zero f10 77 1.1 mrg 78 1.1 mrg mov 0, %g3 C cy = 0 79 1.1 mrg 80 1.1 mrg define(`fanop', `fitod %f18, %f0') C A quasi nop running in the FA pipe 81 1.1 mrg 82 1.1 mrg add %sp, 160, %o5 C point in scratch area 83 1.1 mrg and %o5, -32, %o5 C align at 0 (mod 32) in scratch area 84 1.1 mrg 85 1.1 mrg subcc %o2, 1, %o2 86 1.1 mrg ld [%o1], %f11 C read up[i] 87 1.1 mrg add %o1, 4, %o1 C up++ 88 1.1 mrg bne,pt %icc, .L_two_or_more 89 1.1 mrg fxtod %f10, %f2 90 1.1 mrg 91 1.1 mrg fmuld %f2, %f8, %f16 92 1.1 mrg fmuld %f2, %f6, %f4 93 1.1 mrg fdtox %f16, %f14 94 1.1 mrg fdtox %f4, %f12 95 1.1 mrg std %f14, [%o5+16] 96 1.1 mrg std %f12, [%o5+24] 97 1.1 mrg ldx [%o5+16], %g2 C p16 98 1.1 mrg ldx [%o5+24], %g1 C p0 99 1.1 mrg lduw [%o0], %g5 C read rp[i] 100 1.1 mrg b .L1 101 1.1 mrg add %o0, -16, %o0 102 1.1 mrg 103 1.1 mrg .align 16 104 1.1 mrg .L_two_or_more: 105 1.1 mrg subcc %o2, 1, %o2 106 1.1 mrg ld [%o1], %f11 C read up[i] 107 1.1 mrg fmuld %f2, %f8, %f16 108 1.1 mrg fmuld %f2, %f6, %f4 109 1.1 mrg add %o1, 4, %o1 C up++ 110 1.1 mrg bne,pt %icc, .L_three_or_more 111 1.1 mrg fxtod %f10, %f2 112 1.1 mrg 113 1.1 mrg fdtox %f16, %f14 114 1.1 mrg fdtox %f4, %f12 115 1.1 mrg std %f14, [%o5+16] 116 1.1 mrg fmuld %f2, %f8, %f16 117 1.1 mrg std %f12, [%o5+24] 118 1.1 mrg fmuld %f2, %f6, %f4 119 1.1 mrg fdtox %f16, %f14 120 1.1 mrg fdtox %f4, %f12 121 1.1 mrg std %f14, [%o5+0] 122 1.1 mrg std %f12, [%o5+8] 123 1.1 mrg lduw [%o0], %g5 C read rp[i] 124 1.1 mrg ldx [%o5+16], %g2 C p16 125 1.1 mrg ldx [%o5+24], %g1 C p0 126 1.1 mrg b .L2 127 1.1 mrg add %o0, -12, %o0 128 1.1 mrg 129 1.1 mrg .align 16 130 1.1 mrg .L_three_or_more: 131 1.1 mrg subcc %o2, 1, %o2 132 1.1 mrg ld [%o1], %f11 C read up[i] 133 1.1 mrg fdtox %f16, %f14 134 1.1 mrg fdtox %f4, %f12 135 1.1 mrg std %f14, [%o5+16] 136 1.1 mrg fmuld %f2, %f8, %f16 137 1.1 mrg std %f12, [%o5+24] 138 1.1 mrg fmuld %f2, %f6, %f4 139 1.1 mrg add %o1, 4, %o1 C up++ 140 1.1 mrg bne,pt %icc, .L_four_or_more 141 1.1 mrg fxtod %f10, %f2 142 1.1 mrg 143 1.1 mrg fdtox %f16, %f14 144 1.1 mrg fdtox %f4, %f12 145 1.1 mrg std %f14, [%o5+0] 146 1.1 mrg fmuld %f2, %f8, %f16 147 1.1 mrg std %f12, [%o5+8] 148 1.1 mrg fmuld %f2, %f6, %f4 149 1.1 mrg fdtox %f16, %f14 150 1.1 mrg ldx [%o5+16], %g2 C p16 151 1.1 mrg fdtox %f4, %f12 152 1.1 mrg ldx [%o5+24], %g1 C p0 153 1.1 mrg std %f14, [%o5+16] 154 1.1 mrg std %f12, [%o5+24] 155 1.1 mrg lduw [%o0], %g5 C read rp[i] 156 1.1 mrg b .L3 157 1.1 mrg add %o0, -8, %o0 158 1.1 mrg 159 1.1 mrg .align 16 160 1.1 mrg .L_four_or_more: 161 1.1 mrg subcc %o2, 1, %o2 162 1.1 mrg ld [%o1], %f11 C read up[i] 163 1.1 mrg fdtox %f16, %f14 164 1.1 mrg fdtox %f4, %f12 165 1.1 mrg std %f14, [%o5+0] 166 1.1 mrg fmuld %f2, %f8, %f16 167 1.1 mrg std %f12, [%o5+8] 168 1.1 mrg fmuld %f2, %f6, %f4 169 1.1 mrg add %o1, 4, %o1 C up++ 170 1.1 mrg bne,pt %icc, .L_five_or_more 171 1.1 mrg fxtod %f10, %f2 172 1.1 mrg 173 1.1 mrg fdtox %f16, %f14 174 1.1 mrg ldx [%o5+16], %g2 C p16 175 1.1 mrg fdtox %f4, %f12 176 1.1 mrg ldx [%o5+24], %g1 C p0 177 1.1 mrg std %f14, [%o5+16] 178 1.1 mrg fmuld %f2, %f8, %f16 179 1.1 mrg std %f12, [%o5+24] 180 1.1 mrg fmuld %f2, %f6, %f4 181 1.1 mrg add %o1, 4, %o1 C up++ 182 1.1 mrg lduw [%o0], %g5 C read rp[i] 183 1.1 mrg b .L4 184 1.1 mrg add %o0, -4, %o0 185 1.1 mrg 186 1.1 mrg .align 16 187 1.1 mrg .L_five_or_more: 188 1.1 mrg subcc %o2, 1, %o2 189 1.1 mrg ld [%o1], %f11 C read up[i] 190 1.1 mrg fdtox %f16, %f14 191 1.1 mrg ldx [%o5+16], %g2 C p16 192 1.1 mrg fdtox %f4, %f12 193 1.1 mrg ldx [%o5+24], %g1 C p0 194 1.1 mrg std %f14, [%o5+16] 195 1.1 mrg fmuld %f2, %f8, %f16 196 1.1 mrg std %f12, [%o5+24] 197 1.1 mrg fmuld %f2, %f6, %f4 198 1.1 mrg add %o1, 4, %o1 C up++ 199 1.1 mrg lduw [%o0], %g5 C read rp[i] 200 1.1 mrg bne,pt %icc, .Loop 201 1.1 mrg fxtod %f10, %f2 202 1.1 mrg b,a .L5 203 1.1 mrg 204 1.1 mrg C BEGIN MAIN LOOP 205 1.1 mrg .align 16 206 1.1 mrg C -- 0 207 1.1 mrg .Loop: nop 208 1.1 mrg subcc %o2, 1, %o2 209 1.1 mrg ld [%o1], %f11 C read up[i] 210 1.1 mrg fdtox %f16, %f14 211 1.1 mrg C -- 1 212 1.1 mrg sllx %g2, 16, %g4 C (p16 << 16) 213 1.1 mrg add %o0, 4, %o0 C rp++ 214 1.1 mrg ldx [%o5+0], %g2 C p16 215 1.1 mrg fdtox %f4, %f12 216 1.1 mrg C -- 2 217 1.1 mrg nop 218 1.1 mrg add %g1, %g4, %g4 C p = p0 + (p16 << 16) 219 1.1 mrg ldx [%o5+8], %g1 C p0 220 1.1 mrg fanop 221 1.1 mrg C -- 3 222 1.1 mrg nop 223 1.1 mrg add %g3, %g4, %g4 C p += cy 224 1.1 mrg std %f14, [%o5+0] 225 1.1 mrg fmuld %f2, %f8, %f16 226 1.1 mrg C -- 4 227 1.1 mrg nop 228 1.1 mrg add %g5, %g4, %g4 C p += rp[i] 229 1.1 mrg std %f12, [%o5+8] 230 1.1 mrg fmuld %f2, %f6, %f4 231 1.1 mrg C -- 5 232 1.1 mrg xor %o5, 16, %o5 C alternate scratch variables 233 1.1 mrg add %o1, 4, %o1 C up++ 234 1.1 mrg stw %g4, [%o0-4] 235 1.1 mrg fanop 236 1.1 mrg C -- 6 237 1.1 mrg srlx %g4, 32, %g3 C new cy 238 1.1 mrg lduw [%o0], %g5 C read rp[i] 239 1.1 mrg bne,pt %icc, .Loop 240 1.1 mrg fxtod %f10, %f2 241 1.1 mrg C END MAIN LOOP 242 1.1 mrg 243 1.1 mrg .L5: fdtox %f16, %f14 244 1.1 mrg sllx %g2, 16, %g4 C (p16 << 16) 245 1.1 mrg ldx [%o5+0], %g2 C p16 246 1.1 mrg fdtox %f4, %f12 247 1.1 mrg add %g1, %g4, %g4 C p = p0 + (p16 << 16) 248 1.1 mrg ldx [%o5+8], %g1 C p0 249 1.1 mrg add %g4, %g3, %g4 C p += cy 250 1.1 mrg std %f14, [%o5+0] 251 1.1 mrg fmuld %f2, %f8, %f16 252 1.1 mrg add %g5, %g4, %g4 C p += rp[i] 253 1.1 mrg std %f12, [%o5+8] 254 1.1 mrg fmuld %f2, %f6, %f4 255 1.1 mrg xor %o5, 16, %o5 256 1.1 mrg stw %g4, [%o0+0] 257 1.1 mrg srlx %g4, 32, %g3 C new cy 258 1.1 mrg lduw [%o0+4], %g5 C read rp[i] 259 1.1 mrg 260 1.1 mrg .L4: fdtox %f16, %f14 261 1.1 mrg sllx %g2, 16, %g4 C (p16 << 16) 262 1.1 mrg ldx [%o5+0], %g2 C p16 263 1.1 mrg fdtox %f4, %f12 264 1.1 mrg add %g1, %g4, %g4 C p = p0 + (p16 << 16) 265 1.1 mrg ldx [%o5+8], %g1 C p0 266 1.1 mrg add %g3, %g4, %g4 C p += cy 267 1.1 mrg std %f14, [%o5+0] 268 1.1 mrg add %g5, %g4, %g4 C p += rp[i] 269 1.1 mrg std %f12, [%o5+8] 270 1.1 mrg xor %o5, 16, %o5 271 1.1 mrg stw %g4, [%o0+4] 272 1.1 mrg srlx %g4, 32, %g3 C new cy 273 1.1 mrg lduw [%o0+8], %g5 C read rp[i] 274 1.1 mrg 275 1.1 mrg .L3: sllx %g2, 16, %g4 C (p16 << 16) 276 1.1 mrg ldx [%o5+0], %g2 C p16 277 1.1 mrg add %g1, %g4, %g4 C p = p0 + (p16 << 16) 278 1.1 mrg ldx [%o5+8], %g1 C p0 279 1.1 mrg add %g3, %g4, %g4 C p += cy 280 1.1 mrg add %g5, %g4, %g4 C p += rp[i] 281 1.1 mrg xor %o5, 16, %o5 282 1.1 mrg stw %g4, [%o0+8] 283 1.1 mrg srlx %g4, 32, %g3 C new cy 284 1.1 mrg lduw [%o0+12], %g5 C read rp[i] 285 1.1 mrg 286 1.1 mrg .L2: sllx %g2, 16, %g4 C (p16 << 16) 287 1.1 mrg ldx [%o5+0], %g2 C p16 288 1.1 mrg add %g1, %g4, %g4 C p = p0 + (p16 << 16) 289 1.1 mrg ldx [%o5+8], %g1 C p0 290 1.1 mrg add %g3, %g4, %g4 C p += cy 291 1.1 mrg add %g5, %g4, %g4 C p += rp[i] 292 1.1 mrg stw %g4, [%o0+12] 293 1.1 mrg srlx %g4, 32, %g3 C new cy 294 1.1 mrg lduw [%o0+16], %g5 C read rp[i] 295 1.1 mrg 296 1.1 mrg .L1: sllx %g2, 16, %g4 C (p16 << 16) 297 1.1 mrg add %g1, %g4, %g4 C p = p0 + (p16 << 16) 298 1.1 mrg add %g3, %g4, %g4 C p += cy 299 1.1 mrg add %g5, %g4, %g4 C p += rp[i] 300 1.1 mrg stw %g4, [%o0+16] 301 1.1 mrg srlx %g4, 32, %g3 C new cy 302 1.1 mrg 303 1.1 mrg mov %g3, %o0 304 1.1 mrg retl 305 1.1 mrg sub %sp, -FSIZE, %sp 306 1.1 mrg EPILOGUE(mpn_addmul_1) 307
Indexes created Tue Apr 21 00:23:25 UTC 2026