addmul_1.asm revision 1.1
1 1.1 mrg dnl IA-64 mpn_addmul_1 -- Multiply a limb vector with a limb and add the 2 1.1 mrg dnl result to a second limb vector. 3 1.1 mrg 4 1.1 mrg dnl Copyright 2000, 2001, 2002, 2003, 2004, 2005, 2007 Free Software 5 1.1 mrg dnl Foundation, Inc. 6 1.1 mrg 7 1.1 mrg dnl This file is part of the GNU MP Library. 8 1.1 mrg 9 1.1 mrg dnl The GNU MP Library is free software; you can redistribute it and/or modify 10 1.1 mrg dnl it under the terms of the GNU Lesser General Public License as published 11 1.1 mrg dnl by the Free Software Foundation; either version 3 of the License, or (at 12 1.1 mrg dnl your option) any later version. 13 1.1 mrg 14 1.1 mrg dnl The GNU MP Library is distributed in the hope that it will be useful, but 15 1.1 mrg dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY 16 1.1 mrg dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public 17 1.1 mrg dnl License for more details. 18 1.1 mrg 19 1.1 mrg dnl You should have received a copy of the GNU Lesser General Public License 20 1.1 mrg dnl along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. 21 1.1 mrg 22 1.1 mrg include(`../config.m4') 23 1.1 mrg 24 1.1 mrg C cycles/limb 25 1.1 mrg C Itanium: 3.0 26 1.1 mrg C Itanium 2: 2.0 27 1.1 mrg 28 1.1 mrg C TODO 29 1.1 mrg C * Further optimize feed-in and wind-down code, both for speed and code size. 30 1.1 mrg C * Handle low limb input and results specially, using a common stf8 in the 31 1.1 mrg C epilogue. 32 1.1 mrg C * Use 1 c/l carry propagation scheme in wind-down code. 33 1.1 mrg C * Use extra pointer registers for `up' and rp to speed up feed-in loads. 34 1.1 mrg C * Work out final differences with mul_1.asm. That function is 300 bytes 35 1.1 mrg C smaller than this due to better loop scheduling and thus simpler feed-in 36 1.1 mrg C code. 37 1.1 mrg 38 1.1 mrg C INPUT PARAMETERS 39 1.1 mrg define(`rp', `r32') 40 1.1 mrg define(`up', `r33') 41 1.1 mrg define(`n', `r34') 42 1.1 mrg define(`vl', `r35') 43 1.1 mrg 44 1.1 mrg ASM_START() 45 1.1 mrg PROLOGUE(mpn_addmul_1) 46 1.1 mrg .prologue 47 1.1 mrg .save ar.lc, r2 48 1.1 mrg .body 49 1.1 mrg 50 1.1 mrg ifdef(`HAVE_ABI_32', 51 1.1 mrg ` addp4 rp = 0, rp C M I 52 1.1 mrg addp4 up = 0, up C M I 53 1.1 mrg zxt4 n = n C I 54 1.1 mrg ;; 55 1.1 mrg ') 56 1.1 mrg {.mmi 57 1.1 mrg adds r15 = -1, n C M I 58 1.1 mrg mov r20 = rp C M I 59 1.1 mrg mov.i r2 = ar.lc C I0 60 1.1 mrg } 61 1.1 mrg {.mmi 62 1.1 mrg ldf8 f7 = [up], 8 C M 63 1.1 mrg ldf8 f8 = [rp], 8 C M 64 1.1 mrg and r14 = 3, n C M I 65 1.1 mrg ;; 66 1.1 mrg } 67 1.1 mrg {.mmi 68 1.1 mrg setf.sig f6 = vl C M2 M3 69 1.1 mrg cmp.eq p10, p0 = 0, r14 C M I 70 1.1 mrg shr.u r31 = r15, 2 C I0 71 1.1 mrg } 72 1.1 mrg {.mmi 73 1.1 mrg cmp.eq p11, p0 = 2, r14 C M I 74 1.1 mrg cmp.eq p12, p0 = 3, r14 C M I 75 1.1 mrg nop.i 0 C I 76 1.1 mrg ;; 77 1.1 mrg } 78 1.1 mrg {.mii 79 1.1 mrg cmp.ne p6, p7 = r0, r0 C M I 80 1.1 mrg mov.i ar.lc = r31 C I0 81 1.1 mrg cmp.ne p8, p9 = r0, r0 C M I 82 1.1 mrg } 83 1.1 mrg {.bbb 84 1.1 mrg (p10) br.dptk .Lb00 C B 85 1.1 mrg (p11) br.dptk .Lb10 C B 86 1.1 mrg (p12) br.dptk .Lb11 C B 87 1.1 mrg ;; 88 1.1 mrg } 89 1.1 mrg 90 1.1 mrg .Lb01: br.cloop.dptk .grt1 C B 91 1.1 mrg 92 1.1 mrg xma.l f39 = f7, f6, f8 C F 93 1.1 mrg xma.hu f43 = f7, f6, f8 C F 94 1.1 mrg ;; 95 1.1 mrg getf.sig r8 = f43 C M2 96 1.1 mrg stf8 [r20] = f39 C M2 M3 97 1.1 mrg mov.i ar.lc = r2 C I0 98 1.1 mrg br.ret.sptk.many b0 C B 99 1.1 mrg 100 1.1 mrg .grt1: 101 1.1 mrg ldf8 f32 = [up], 8 102 1.1 mrg ldf8 f44 = [rp], 8 103 1.1 mrg ;; 104 1.1 mrg ldf8 f33 = [up], 8 105 1.1 mrg ldf8 f45 = [rp], 8 106 1.1 mrg ;; 107 1.1 mrg ldf8 f34 = [up], 8 108 1.1 mrg xma.l f39 = f7, f6, f8 109 1.1 mrg ldf8 f46 = [rp], 8 110 1.1 mrg xma.hu f43 = f7, f6, f8 111 1.1 mrg ;; 112 1.1 mrg ldf8 f35 = [up], 8 113 1.1 mrg ldf8 f47 = [rp], 8 114 1.1 mrg br.cloop.dptk .grt5 115 1.1 mrg 116 1.1 mrg xma.l f36 = f32, f6, f44 117 1.1 mrg xma.hu f40 = f32, f6, f44 118 1.1 mrg ;; 119 1.1 mrg stf8 [r20] = f39, 8 120 1.1 mrg xma.l f37 = f33, f6, f45 121 1.1 mrg xma.hu f41 = f33, f6, f45 122 1.1 mrg ;; 123 1.1 mrg getf.sig r31 = f43 124 1.1 mrg getf.sig r24 = f36 125 1.1 mrg xma.l f38 = f34, f6, f46 126 1.1 mrg xma.hu f42 = f34, f6, f46 127 1.1 mrg ;; 128 1.1 mrg getf.sig r28 = f40 129 1.1 mrg getf.sig r25 = f37 130 1.1 mrg xma.l f39 = f35, f6, f47 131 1.1 mrg xma.hu f43 = f35, f6, f47 132 1.1 mrg ;; 133 1.1 mrg getf.sig r29 = f41 134 1.1 mrg getf.sig r26 = f38 135 1.1 mrg br .Lcj5 136 1.1 mrg 137 1.1 mrg .grt5: 138 1.1 mrg mov r30 = 0 139 1.1 mrg xma.l f36 = f32, f6, f44 140 1.1 mrg xma.hu f40 = f32, f6, f44 141 1.1 mrg ;; 142 1.1 mrg ldf8 f32 = [up], 8 143 1.1 mrg xma.l f37 = f33, f6, f45 144 1.1 mrg ldf8 f44 = [rp], 8 145 1.1 mrg xma.hu f41 = f33, f6, f45 146 1.1 mrg ;; 147 1.1 mrg ldf8 f33 = [up], 8 148 1.1 mrg getf.sig r27 = f39 149 1.1 mrg ;; 150 1.1 mrg getf.sig r31 = f43 151 1.1 mrg xma.l f38 = f34, f6, f46 152 1.1 mrg ldf8 f45 = [rp], 8 153 1.1 mrg xma.hu f42 = f34, f6, f46 154 1.1 mrg ;; 155 1.1 mrg ldf8 f34 = [up], 8 156 1.1 mrg getf.sig r24 = f36 157 1.1 mrg ;; 158 1.1 mrg getf.sig r28 = f40 159 1.1 mrg xma.l f39 = f35, f6, f47 160 1.1 mrg ldf8 f46 = [rp], 8 161 1.1 mrg xma.hu f43 = f35, f6, f47 162 1.1 mrg ;; 163 1.1 mrg ldf8 f35 = [up], 8 164 1.1 mrg getf.sig r25 = f37 165 1.1 mrg br.cloop.dptk .Loop 166 1.1 mrg br .Le0 167 1.1 mrg 168 1.1 mrg 169 1.1 mrg .Lb10: ldf8 f35 = [up], 8 170 1.1 mrg ldf8 f47 = [rp], 8 171 1.1 mrg br.cloop.dptk .grt2 172 1.1 mrg 173 1.1 mrg xma.l f38 = f7, f6, f8 174 1.1 mrg xma.hu f42 = f7, f6, f8 175 1.1 mrg ;; 176 1.1 mrg xma.l f39 = f35, f6, f47 177 1.1 mrg xma.hu f43 = f35, f6, f47 178 1.1 mrg ;; 179 1.1 mrg getf.sig r30 = f42 180 1.1 mrg stf8 [r20] = f38, 8 181 1.1 mrg getf.sig r27 = f39 182 1.1 mrg getf.sig r8 = f43 183 1.1 mrg br .Lcj2 184 1.1 mrg 185 1.1 mrg .grt2: 186 1.1 mrg ldf8 f32 = [up], 8 187 1.1 mrg ldf8 f44 = [rp], 8 188 1.1 mrg ;; 189 1.1 mrg ldf8 f33 = [up], 8 190 1.1 mrg xma.l f38 = f7, f6, f8 191 1.1 mrg ldf8 f45 = [rp], 8 192 1.1 mrg xma.hu f42 = f7, f6, f8 193 1.1 mrg ;; 194 1.1 mrg ldf8 f34 = [up], 8 195 1.1 mrg xma.l f39 = f35, f6, f47 196 1.1 mrg ldf8 f46 = [rp], 8 197 1.1 mrg xma.hu f43 = f35, f6, f47 198 1.1 mrg ;; 199 1.1 mrg ldf8 f35 = [up], 8 200 1.1 mrg ldf8 f47 = [rp], 8 201 1.1 mrg br.cloop.dptk .grt6 202 1.1 mrg 203 1.1 mrg stf8 [r20] = f38, 8 204 1.1 mrg xma.l f36 = f32, f6, f44 205 1.1 mrg xma.hu f40 = f32, f6, f44 206 1.1 mrg ;; 207 1.1 mrg getf.sig r30 = f42 208 1.1 mrg getf.sig r27 = f39 209 1.1 mrg xma.l f37 = f33, f6, f45 210 1.1 mrg xma.hu f41 = f33, f6, f45 211 1.1 mrg ;; 212 1.1 mrg getf.sig r31 = f43 213 1.1 mrg getf.sig r24 = f36 214 1.1 mrg xma.l f38 = f34, f6, f46 215 1.1 mrg xma.hu f42 = f34, f6, f46 216 1.1 mrg ;; 217 1.1 mrg getf.sig r28 = f40 218 1.1 mrg getf.sig r25 = f37 219 1.1 mrg xma.l f39 = f35, f6, f47 220 1.1 mrg xma.hu f43 = f35, f6, f47 221 1.1 mrg br .Lcj6 222 1.1 mrg 223 1.1 mrg .grt6: 224 1.1 mrg mov r29 = 0 225 1.1 mrg xma.l f36 = f32, f6, f44 226 1.1 mrg xma.hu f40 = f32, f6, f44 227 1.1 mrg ;; 228 1.1 mrg ldf8 f32 = [up], 8 229 1.1 mrg getf.sig r26 = f38 230 1.1 mrg ;; 231 1.1 mrg getf.sig r30 = f42 232 1.1 mrg xma.l f37 = f33, f6, f45 233 1.1 mrg ldf8 f44 = [rp], 8 234 1.1 mrg xma.hu f41 = f33, f6, f45 235 1.1 mrg ;; 236 1.1 mrg ldf8 f33 = [up], 8 237 1.1 mrg getf.sig r27 = f39 238 1.1 mrg ;; 239 1.1 mrg getf.sig r31 = f43 240 1.1 mrg xma.l f38 = f34, f6, f46 241 1.1 mrg ldf8 f45 = [rp], 8 242 1.1 mrg xma.hu f42 = f34, f6, f46 243 1.1 mrg ;; 244 1.1 mrg ldf8 f34 = [up], 8 245 1.1 mrg getf.sig r24 = f36 246 1.1 mrg br .LL10 247 1.1 mrg 248 1.1 mrg 249 1.1 mrg .Lb11: ldf8 f34 = [up], 8 250 1.1 mrg ldf8 f46 = [rp], 8 251 1.1 mrg ;; 252 1.1 mrg ldf8 f35 = [up], 8 253 1.1 mrg ldf8 f47 = [rp], 8 254 1.1 mrg br.cloop.dptk .grt3 255 1.1 mrg ;; 256 1.1 mrg 257 1.1 mrg xma.l f37 = f7, f6, f8 258 1.1 mrg xma.hu f41 = f7, f6, f8 259 1.1 mrg xma.l f38 = f34, f6, f46 260 1.1 mrg xma.hu f42 = f34, f6, f46 261 1.1 mrg xma.l f39 = f35, f6, f47 262 1.1 mrg xma.hu f43 = f35, f6, f47 263 1.1 mrg ;; 264 1.1 mrg getf.sig r29 = f41 265 1.1 mrg stf8 [r20] = f37, 8 266 1.1 mrg getf.sig r26 = f38 267 1.1 mrg getf.sig r30 = f42 268 1.1 mrg getf.sig r27 = f39 269 1.1 mrg getf.sig r8 = f43 270 1.1 mrg br .Lcj3 271 1.1 mrg 272 1.1 mrg .grt3: 273 1.1 mrg ldf8 f32 = [up], 8 274 1.1 mrg xma.l f37 = f7, f6, f8 275 1.1 mrg ldf8 f44 = [rp], 8 276 1.1 mrg xma.hu f41 = f7, f6, f8 277 1.1 mrg ;; 278 1.1 mrg ldf8 f33 = [up], 8 279 1.1 mrg xma.l f38 = f34, f6, f46 280 1.1 mrg ldf8 f45 = [rp], 8 281 1.1 mrg xma.hu f42 = f34, f6, f46 282 1.1 mrg ;; 283 1.1 mrg ldf8 f34 = [up], 8 284 1.1 mrg xma.l f39 = f35, f6, f47 285 1.1 mrg ldf8 f46 = [rp], 8 286 1.1 mrg xma.hu f43 = f35, f6, f47 287 1.1 mrg ;; 288 1.1 mrg ldf8 f35 = [up], 8 289 1.1 mrg getf.sig r25 = f37 C FIXME 290 1.1 mrg ldf8 f47 = [rp], 8 291 1.1 mrg br.cloop.dptk .grt7 292 1.1 mrg 293 1.1 mrg getf.sig r29 = f41 294 1.1 mrg stf8 [r20] = f37, 8 C FIXME 295 1.1 mrg xma.l f36 = f32, f6, f44 296 1.1 mrg getf.sig r26 = f38 297 1.1 mrg xma.hu f40 = f32, f6, f44 298 1.1 mrg ;; 299 1.1 mrg getf.sig r30 = f42 300 1.1 mrg xma.l f37 = f33, f6, f45 301 1.1 mrg getf.sig r27 = f39 302 1.1 mrg xma.hu f41 = f33, f6, f45 303 1.1 mrg ;; 304 1.1 mrg getf.sig r31 = f43 305 1.1 mrg xma.l f38 = f34, f6, f46 306 1.1 mrg getf.sig r24 = f36 307 1.1 mrg xma.hu f42 = f34, f6, f46 308 1.1 mrg br .Lcj7 309 1.1 mrg 310 1.1 mrg .grt7: 311 1.1 mrg getf.sig r29 = f41 312 1.1 mrg xma.l f36 = f32, f6, f44 313 1.1 mrg mov r28 = 0 314 1.1 mrg xma.hu f40 = f32, f6, f44 315 1.1 mrg ;; 316 1.1 mrg ldf8 f32 = [up], 8 317 1.1 mrg getf.sig r26 = f38 318 1.1 mrg ;; 319 1.1 mrg getf.sig r30 = f42 320 1.1 mrg xma.l f37 = f33, f6, f45 321 1.1 mrg ldf8 f44 = [rp], 8 322 1.1 mrg xma.hu f41 = f33, f6, f45 323 1.1 mrg ;; 324 1.1 mrg ldf8 f33 = [up], 8 325 1.1 mrg getf.sig r27 = f39 326 1.1 mrg br .LL11 327 1.1 mrg 328 1.1 mrg 329 1.1 mrg .Lb00: ldf8 f33 = [up], 8 330 1.1 mrg ldf8 f45 = [rp], 8 331 1.1 mrg ;; 332 1.1 mrg ldf8 f34 = [up], 8 333 1.1 mrg ldf8 f46 = [rp], 8 334 1.1 mrg ;; 335 1.1 mrg ldf8 f35 = [up], 8 336 1.1 mrg xma.l f36 = f7, f6, f8 337 1.1 mrg ldf8 f47 = [rp], 8 338 1.1 mrg xma.hu f40 = f7, f6, f8 339 1.1 mrg br.cloop.dptk .grt4 340 1.1 mrg 341 1.1 mrg xma.l f37 = f33, f6, f45 342 1.1 mrg xma.hu f41 = f33, f6, f45 343 1.1 mrg xma.l f38 = f34, f6, f46 344 1.1 mrg xma.hu f42 = f34, f6, f46 345 1.1 mrg ;; 346 1.1 mrg getf.sig r28 = f40 347 1.1 mrg stf8 [r20] = f36, 8 348 1.1 mrg xma.l f39 = f35, f6, f47 349 1.1 mrg getf.sig r25 = f37 350 1.1 mrg xma.hu f43 = f35, f6, f47 351 1.1 mrg ;; 352 1.1 mrg getf.sig r29 = f41 353 1.1 mrg getf.sig r26 = f38 354 1.1 mrg getf.sig r30 = f42 355 1.1 mrg getf.sig r27 = f39 356 1.1 mrg br .Lcj4 357 1.1 mrg 358 1.1 mrg .grt4: 359 1.1 mrg ldf8 f32 = [up], 8 360 1.1 mrg xma.l f37 = f33, f6, f45 361 1.1 mrg ldf8 f44 = [rp], 8 362 1.1 mrg xma.hu f41 = f33, f6, f45 363 1.1 mrg ;; 364 1.1 mrg ldf8 f33 = [up], 8 365 1.1 mrg xma.l f38 = f34, f6, f46 366 1.1 mrg ldf8 f45 = [rp], 8 367 1.1 mrg xma.hu f42 = f34, f6, f46 368 1.1 mrg ;; 369 1.1 mrg ldf8 f34 = [up], 8 370 1.1 mrg getf.sig r24 = f36 C FIXME 371 1.1 mrg xma.l f39 = f35, f6, f47 372 1.1 mrg ldf8 f46 = [rp], 8 373 1.1 mrg getf.sig r28 = f40 374 1.1 mrg xma.hu f43 = f35, f6, f47 375 1.1 mrg ;; 376 1.1 mrg ldf8 f35 = [up], 8 377 1.1 mrg getf.sig r25 = f37 378 1.1 mrg ldf8 f47 = [rp], 8 379 1.1 mrg br.cloop.dptk .grt8 380 1.1 mrg 381 1.1 mrg getf.sig r29 = f41 382 1.1 mrg stf8 [r20] = f36, 8 C FIXME 383 1.1 mrg xma.l f36 = f32, f6, f44 384 1.1 mrg getf.sig r26 = f38 385 1.1 mrg getf.sig r30 = f42 386 1.1 mrg xma.hu f40 = f32, f6, f44 387 1.1 mrg ;; 388 1.1 mrg xma.l f37 = f33, f6, f45 389 1.1 mrg getf.sig r27 = f39 390 1.1 mrg xma.hu f41 = f33, f6, f45 391 1.1 mrg br .Lcj8 392 1.1 mrg 393 1.1 mrg .grt8: 394 1.1 mrg getf.sig r29 = f41 395 1.1 mrg xma.l f36 = f32, f6, f44 396 1.1 mrg mov r31 = 0 397 1.1 mrg xma.hu f40 = f32, f6, f44 398 1.1 mrg ;; 399 1.1 mrg ldf8 f32 = [up], 8 400 1.1 mrg getf.sig r26 = f38 401 1.1 mrg br .LL00 402 1.1 mrg 403 1.1 mrg 404 1.1 mrg C *** MAIN LOOP START *** 405 1.1 mrg ALIGN(32) C insn fed cycle # 406 1.1 mrg .Loop: 407 1.1 mrg .pred.rel "mutex", p6, p7 C num by i1 i2 408 1.1 mrg getf.sig r29 = f41 C 00 16 0 0 409 1.1 mrg xma.l f36 = f32, f6, f44 C 01 06,15 0 0 410 1.1 mrg (p6) add r14 = r30, r27, 1 C 02 0 0 411 1.1 mrg ldf8 f47 = [rp], 8 C 03 0 0 412 1.1 mrg xma.hu f40 = f32, f6, f44 C 04 06,15 0 0 413 1.1 mrg (p7) add r14 = r30, r27 C 05 0 0 414 1.1 mrg ;; 415 1.1 mrg .pred.rel "mutex", p6, p7 416 1.1 mrg ldf8 f32 = [up], 8 C 06 1 1 417 1.1 mrg (p6) cmp.leu p8, p9 = r14, r27 C 07 1 1 418 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r27 C 08 1 1 419 1.1 mrg getf.sig r26 = f38 C 09 25 2 1 420 1.1 mrg st8 [r20] = r14, 8 C 10 2 1 421 1.1 mrg nop.b 0 C 11 2 1 422 1.1 mrg ;; 423 1.1 mrg .LL00: 424 1.1 mrg .pred.rel "mutex", p8, p9 425 1.1 mrg getf.sig r30 = f42 C 12 28 3 2 426 1.1 mrg xma.l f37 = f33, f6, f45 C 13 18,27 3 2 427 1.1 mrg (p8) add r16 = r31, r24, 1 C 14 3 2 428 1.1 mrg ldf8 f44 = [rp], 8 C 15 3 2 429 1.1 mrg xma.hu f41 = f33, f6, f45 C 16 18,27 3 2 430 1.1 mrg (p9) add r16 = r31, r24 C 17 3 2 431 1.1 mrg ;; 432 1.1 mrg .pred.rel "mutex", p8, p9 433 1.1 mrg ldf8 f33 = [up], 8 C 18 4 3 434 1.1 mrg (p8) cmp.leu p6, p7 = r16, r24 C 19 4 3 435 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r24 C 20 4 3 436 1.1 mrg getf.sig r27 = f39 C 21 37 5 3 437 1.1 mrg st8 [r20] = r16, 8 C 22 5 3 438 1.1 mrg nop.b 0 C 23 5 3 439 1.1 mrg ;; 440 1.1 mrg .LL11: 441 1.1 mrg .pred.rel "mutex", p6, p7 442 1.1 mrg getf.sig r31 = f43 C 24 40 6 4 443 1.1 mrg xma.l f38 = f34, f6, f46 C 25 30,39 6 4 444 1.1 mrg (p6) add r14 = r28, r25, 1 C 26 6 4 445 1.1 mrg ldf8 f45 = [rp], 8 C 27 6 4 446 1.1 mrg xma.hu f42 = f34, f6, f46 C 28 30,39 6 4 447 1.1 mrg (p7) add r14 = r28, r25 C 29 6 4 448 1.1 mrg ;; 449 1.1 mrg .pred.rel "mutex", p6, p7 450 1.1 mrg ldf8 f34 = [up], 8 C 30 7 5 451 1.1 mrg (p6) cmp.leu p8, p9 = r14, r25 C 31 7 5 452 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r25 C 32 7 5 453 1.1 mrg getf.sig r24 = f36 C 33 01 8 5 454 1.1 mrg st8 [r20] = r14, 8 C 34 8 5 455 1.1 mrg nop.b 0 C 35 8 5 456 1.1 mrg ;; 457 1.1 mrg .LL10: 458 1.1 mrg .pred.rel "mutex", p8, p9 459 1.1 mrg getf.sig r28 = f40 C 36 04 9 6 460 1.1 mrg xma.l f39 = f35, f6, f47 C 37 42,03 9 6 461 1.1 mrg (p8) add r16 = r29, r26, 1 C 38 9 6 462 1.1 mrg ldf8 f46 = [rp], 8 C 39 9 6 463 1.1 mrg xma.hu f43 = f35, f6, f47 C 40 42,03 9 6 464 1.1 mrg (p9) add r16 = r29, r26 C 41 9 6 465 1.1 mrg ;; 466 1.1 mrg .pred.rel "mutex", p8, p9 467 1.1 mrg ldf8 f35 = [up], 8 C 42 10 7 468 1.1 mrg (p8) cmp.leu p6, p7 = r16, r26 C 43 10 7 469 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r26 C 44 10 7 470 1.1 mrg getf.sig r25 = f37 C 45 13 11 7 471 1.1 mrg st8 [r20] = r16, 8 C 46 11 7 472 1.1 mrg br.cloop.dptk .Loop C 47 11 7 473 1.1 mrg C *** MAIN LOOP END *** 474 1.1 mrg ;; 475 1.1 mrg .Le0: 476 1.1 mrg .pred.rel "mutex", p6, p7 477 1.1 mrg getf.sig r29 = f41 C 478 1.1 mrg xma.l f36 = f32, f6, f44 C 479 1.1 mrg (p6) add r14 = r30, r27, 1 C 480 1.1 mrg ldf8 f47 = [rp], 8 C 481 1.1 mrg xma.hu f40 = f32, f6, f44 C 482 1.1 mrg (p7) add r14 = r30, r27 C 483 1.1 mrg ;; 484 1.1 mrg .pred.rel "mutex", p6, p7 485 1.1 mrg (p6) cmp.leu p8, p9 = r14, r27 C 486 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r27 C 487 1.1 mrg getf.sig r26 = f38 C 488 1.1 mrg st8 [r20] = r14, 8 C 489 1.1 mrg ;; 490 1.1 mrg .pred.rel "mutex", p8, p9 491 1.1 mrg getf.sig r30 = f42 C 492 1.1 mrg xma.l f37 = f33, f6, f45 C 493 1.1 mrg (p8) add r16 = r31, r24, 1 C 494 1.1 mrg xma.hu f41 = f33, f6, f45 C 495 1.1 mrg (p9) add r16 = r31, r24 C 496 1.1 mrg ;; 497 1.1 mrg .pred.rel "mutex", p8, p9 498 1.1 mrg (p8) cmp.leu p6, p7 = r16, r24 C 499 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r24 C 500 1.1 mrg getf.sig r27 = f39 C 501 1.1 mrg st8 [r20] = r16, 8 C 502 1.1 mrg ;; 503 1.1 mrg .Lcj8: 504 1.1 mrg .pred.rel "mutex", p6, p7 505 1.1 mrg getf.sig r31 = f43 C 506 1.1 mrg xma.l f38 = f34, f6, f46 C 507 1.1 mrg (p6) add r14 = r28, r25, 1 C 508 1.1 mrg xma.hu f42 = f34, f6, f46 C 509 1.1 mrg (p7) add r14 = r28, r25 C 510 1.1 mrg ;; 511 1.1 mrg .pred.rel "mutex", p6, p7 512 1.1 mrg (p6) cmp.leu p8, p9 = r14, r25 C 513 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r25 C 514 1.1 mrg getf.sig r24 = f36 C 515 1.1 mrg st8 [r20] = r14, 8 C 516 1.1 mrg ;; 517 1.1 mrg .Lcj7: 518 1.1 mrg .pred.rel "mutex", p8, p9 519 1.1 mrg getf.sig r28 = f40 C 520 1.1 mrg xma.l f39 = f35, f6, f47 C 521 1.1 mrg (p8) add r16 = r29, r26, 1 C 522 1.1 mrg xma.hu f43 = f35, f6, f47 C 523 1.1 mrg (p9) add r16 = r29, r26 C 524 1.1 mrg ;; 525 1.1 mrg .pred.rel "mutex", p8, p9 526 1.1 mrg (p8) cmp.leu p6, p7 = r16, r26 C 527 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r26 C 528 1.1 mrg getf.sig r25 = f37 C 529 1.1 mrg st8 [r20] = r16, 8 C 530 1.1 mrg ;; 531 1.1 mrg .Lcj6: 532 1.1 mrg .pred.rel "mutex", p6, p7 533 1.1 mrg getf.sig r29 = f41 C 534 1.1 mrg (p6) add r14 = r30, r27, 1 C 535 1.1 mrg (p7) add r14 = r30, r27 C 536 1.1 mrg ;; 537 1.1 mrg .pred.rel "mutex", p6, p7 538 1.1 mrg (p6) cmp.leu p8, p9 = r14, r27 C 539 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r27 C 540 1.1 mrg getf.sig r26 = f38 C 541 1.1 mrg st8 [r20] = r14, 8 C 542 1.1 mrg ;; 543 1.1 mrg .Lcj5: 544 1.1 mrg .pred.rel "mutex", p8, p9 545 1.1 mrg getf.sig r30 = f42 C 546 1.1 mrg (p8) add r16 = r31, r24, 1 C 547 1.1 mrg (p9) add r16 = r31, r24 C 548 1.1 mrg ;; 549 1.1 mrg .pred.rel "mutex", p8, p9 550 1.1 mrg (p8) cmp.leu p6, p7 = r16, r24 C 551 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r24 C 552 1.1 mrg getf.sig r27 = f39 C 553 1.1 mrg st8 [r20] = r16, 8 C 554 1.1 mrg ;; 555 1.1 mrg .Lcj4: 556 1.1 mrg .pred.rel "mutex", p6, p7 557 1.1 mrg getf.sig r8 = f43 C 558 1.1 mrg (p6) add r14 = r28, r25, 1 C 559 1.1 mrg (p7) add r14 = r28, r25 C 560 1.1 mrg ;; 561 1.1 mrg .pred.rel "mutex", p6, p7 562 1.1 mrg st8 [r20] = r14, 8 C 563 1.1 mrg (p6) cmp.leu p8, p9 = r14, r25 C 564 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r25 C 565 1.1 mrg ;; 566 1.1 mrg .Lcj3: 567 1.1 mrg .pred.rel "mutex", p8, p9 568 1.1 mrg (p8) add r16 = r29, r26, 1 C 569 1.1 mrg (p9) add r16 = r29, r26 C 570 1.1 mrg ;; 571 1.1 mrg .pred.rel "mutex", p8, p9 572 1.1 mrg st8 [r20] = r16, 8 C 573 1.1 mrg (p8) cmp.leu p6, p7 = r16, r26 C 574 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r26 C 575 1.1 mrg ;; 576 1.1 mrg .Lcj2: 577 1.1 mrg .pred.rel "mutex", p6, p7 578 1.1 mrg (p6) add r14 = r30, r27, 1 C 579 1.1 mrg (p7) add r14 = r30, r27 C 580 1.1 mrg ;; 581 1.1 mrg .pred.rel "mutex", p6, p7 582 1.1 mrg st8 [r20] = r14 C 583 1.1 mrg (p6) cmp.leu p8, p9 = r14, r27 C 584 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r27 C 585 1.1 mrg ;; 586 1.1 mrg (p8) add r8 = 1, r8 C M I 587 1.1 mrg mov.i ar.lc = r2 C I0 588 1.1 mrg br.ret.sptk.many b0 C B 589 1.1 mrg EPILOGUE() 590 1.1 mrg ASM_END() 591
Indexes created Fri Mar 06 19:51:54 UTC 2026