addmul_1.asm revision 1.1.1.2
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 1.1.1.2 mrg dnl Contributed to the GNU project by Torbjorn Granlund. 5 1.1.1.2 mrg 6 1.1 mrg dnl Copyright 2000, 2001, 2002, 2003, 2004, 2005, 2007 Free Software 7 1.1 mrg dnl Foundation, Inc. 8 1.1 mrg 9 1.1 mrg dnl This file is part of the GNU MP Library. 10 1.1 mrg 11 1.1 mrg dnl The GNU MP Library is free software; you can redistribute it and/or modify 12 1.1 mrg dnl it under the terms of the GNU Lesser General Public License as published 13 1.1 mrg dnl by the Free Software Foundation; either version 3 of the License, or (at 14 1.1 mrg dnl your option) any later version. 15 1.1 mrg 16 1.1 mrg dnl The GNU MP Library is distributed in the hope that it will be useful, but 17 1.1 mrg dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY 18 1.1 mrg dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public 19 1.1 mrg dnl License for more details. 20 1.1 mrg 21 1.1 mrg dnl You should have received a copy of the GNU Lesser General Public License 22 1.1 mrg dnl along with the GNU MP Library. If not, see http://www.gnu.org/licenses/. 23 1.1 mrg 24 1.1 mrg include(`../config.m4') 25 1.1 mrg 26 1.1 mrg C cycles/limb 27 1.1 mrg C Itanium: 3.0 28 1.1 mrg C Itanium 2: 2.0 29 1.1 mrg 30 1.1 mrg C TODO 31 1.1 mrg C * Further optimize feed-in and wind-down code, both for speed and code size. 32 1.1 mrg C * Handle low limb input and results specially, using a common stf8 in the 33 1.1 mrg C epilogue. 34 1.1 mrg C * Use 1 c/l carry propagation scheme in wind-down code. 35 1.1 mrg C * Use extra pointer registers for `up' and rp to speed up feed-in loads. 36 1.1 mrg C * Work out final differences with mul_1.asm. That function is 300 bytes 37 1.1 mrg C smaller than this due to better loop scheduling and thus simpler feed-in 38 1.1 mrg C code. 39 1.1 mrg 40 1.1 mrg C INPUT PARAMETERS 41 1.1 mrg define(`rp', `r32') 42 1.1 mrg define(`up', `r33') 43 1.1 mrg define(`n', `r34') 44 1.1 mrg define(`vl', `r35') 45 1.1 mrg 46 1.1 mrg ASM_START() 47 1.1 mrg PROLOGUE(mpn_addmul_1) 48 1.1 mrg .prologue 49 1.1 mrg .save ar.lc, r2 50 1.1 mrg .body 51 1.1 mrg 52 1.1 mrg ifdef(`HAVE_ABI_32', 53 1.1 mrg ` addp4 rp = 0, rp C M I 54 1.1 mrg addp4 up = 0, up C M I 55 1.1 mrg zxt4 n = n C I 56 1.1 mrg ;; 57 1.1 mrg ') 58 1.1 mrg {.mmi 59 1.1 mrg adds r15 = -1, n C M I 60 1.1 mrg mov r20 = rp C M I 61 1.1 mrg mov.i r2 = ar.lc C I0 62 1.1 mrg } 63 1.1 mrg {.mmi 64 1.1 mrg ldf8 f7 = [up], 8 C M 65 1.1 mrg ldf8 f8 = [rp], 8 C M 66 1.1 mrg and r14 = 3, n C M I 67 1.1 mrg ;; 68 1.1 mrg } 69 1.1 mrg {.mmi 70 1.1 mrg setf.sig f6 = vl C M2 M3 71 1.1 mrg cmp.eq p10, p0 = 0, r14 C M I 72 1.1 mrg shr.u r31 = r15, 2 C I0 73 1.1 mrg } 74 1.1 mrg {.mmi 75 1.1 mrg cmp.eq p11, p0 = 2, r14 C M I 76 1.1 mrg cmp.eq p12, p0 = 3, r14 C M I 77 1.1 mrg nop.i 0 C I 78 1.1 mrg ;; 79 1.1 mrg } 80 1.1 mrg {.mii 81 1.1 mrg cmp.ne p6, p7 = r0, r0 C M I 82 1.1 mrg mov.i ar.lc = r31 C I0 83 1.1 mrg cmp.ne p8, p9 = r0, r0 C M I 84 1.1 mrg } 85 1.1 mrg {.bbb 86 1.1 mrg (p10) br.dptk .Lb00 C B 87 1.1 mrg (p11) br.dptk .Lb10 C B 88 1.1 mrg (p12) br.dptk .Lb11 C B 89 1.1 mrg ;; 90 1.1 mrg } 91 1.1 mrg 92 1.1 mrg .Lb01: br.cloop.dptk .grt1 C B 93 1.1 mrg 94 1.1 mrg xma.l f39 = f7, f6, f8 C F 95 1.1 mrg xma.hu f43 = f7, f6, f8 C F 96 1.1 mrg ;; 97 1.1 mrg getf.sig r8 = f43 C M2 98 1.1 mrg stf8 [r20] = f39 C M2 M3 99 1.1 mrg mov.i ar.lc = r2 C I0 100 1.1 mrg br.ret.sptk.many b0 C B 101 1.1 mrg 102 1.1 mrg .grt1: 103 1.1 mrg ldf8 f32 = [up], 8 104 1.1 mrg ldf8 f44 = [rp], 8 105 1.1 mrg ;; 106 1.1 mrg ldf8 f33 = [up], 8 107 1.1 mrg ldf8 f45 = [rp], 8 108 1.1 mrg ;; 109 1.1 mrg ldf8 f34 = [up], 8 110 1.1 mrg xma.l f39 = f7, f6, f8 111 1.1 mrg ldf8 f46 = [rp], 8 112 1.1 mrg xma.hu f43 = f7, f6, f8 113 1.1 mrg ;; 114 1.1 mrg ldf8 f35 = [up], 8 115 1.1 mrg ldf8 f47 = [rp], 8 116 1.1 mrg br.cloop.dptk .grt5 117 1.1 mrg 118 1.1 mrg xma.l f36 = f32, f6, f44 119 1.1 mrg xma.hu f40 = f32, f6, f44 120 1.1 mrg ;; 121 1.1 mrg stf8 [r20] = f39, 8 122 1.1 mrg xma.l f37 = f33, f6, f45 123 1.1 mrg xma.hu f41 = f33, f6, f45 124 1.1 mrg ;; 125 1.1 mrg getf.sig r31 = f43 126 1.1 mrg getf.sig r24 = f36 127 1.1 mrg xma.l f38 = f34, f6, f46 128 1.1 mrg xma.hu f42 = f34, f6, f46 129 1.1 mrg ;; 130 1.1 mrg getf.sig r28 = f40 131 1.1 mrg getf.sig r25 = f37 132 1.1 mrg xma.l f39 = f35, f6, f47 133 1.1 mrg xma.hu f43 = f35, f6, f47 134 1.1 mrg ;; 135 1.1 mrg getf.sig r29 = f41 136 1.1 mrg getf.sig r26 = f38 137 1.1 mrg br .Lcj5 138 1.1 mrg 139 1.1 mrg .grt5: 140 1.1 mrg mov r30 = 0 141 1.1 mrg xma.l f36 = f32, f6, f44 142 1.1 mrg xma.hu f40 = f32, f6, f44 143 1.1 mrg ;; 144 1.1 mrg ldf8 f32 = [up], 8 145 1.1 mrg xma.l f37 = f33, f6, f45 146 1.1 mrg ldf8 f44 = [rp], 8 147 1.1 mrg xma.hu f41 = f33, f6, f45 148 1.1 mrg ;; 149 1.1 mrg ldf8 f33 = [up], 8 150 1.1 mrg getf.sig r27 = f39 151 1.1 mrg ;; 152 1.1 mrg getf.sig r31 = f43 153 1.1 mrg xma.l f38 = f34, f6, f46 154 1.1 mrg ldf8 f45 = [rp], 8 155 1.1 mrg xma.hu f42 = f34, f6, f46 156 1.1 mrg ;; 157 1.1 mrg ldf8 f34 = [up], 8 158 1.1 mrg getf.sig r24 = f36 159 1.1 mrg ;; 160 1.1 mrg getf.sig r28 = f40 161 1.1 mrg xma.l f39 = f35, f6, f47 162 1.1 mrg ldf8 f46 = [rp], 8 163 1.1 mrg xma.hu f43 = f35, f6, f47 164 1.1 mrg ;; 165 1.1 mrg ldf8 f35 = [up], 8 166 1.1 mrg getf.sig r25 = f37 167 1.1 mrg br.cloop.dptk .Loop 168 1.1 mrg br .Le0 169 1.1 mrg 170 1.1 mrg 171 1.1 mrg .Lb10: ldf8 f35 = [up], 8 172 1.1 mrg ldf8 f47 = [rp], 8 173 1.1 mrg br.cloop.dptk .grt2 174 1.1 mrg 175 1.1 mrg xma.l f38 = f7, f6, f8 176 1.1 mrg xma.hu f42 = f7, f6, f8 177 1.1 mrg ;; 178 1.1 mrg xma.l f39 = f35, f6, f47 179 1.1 mrg xma.hu f43 = f35, f6, f47 180 1.1 mrg ;; 181 1.1 mrg getf.sig r30 = f42 182 1.1 mrg stf8 [r20] = f38, 8 183 1.1 mrg getf.sig r27 = f39 184 1.1 mrg getf.sig r8 = f43 185 1.1 mrg br .Lcj2 186 1.1 mrg 187 1.1 mrg .grt2: 188 1.1 mrg ldf8 f32 = [up], 8 189 1.1 mrg ldf8 f44 = [rp], 8 190 1.1 mrg ;; 191 1.1 mrg ldf8 f33 = [up], 8 192 1.1 mrg xma.l f38 = f7, f6, f8 193 1.1 mrg ldf8 f45 = [rp], 8 194 1.1 mrg xma.hu f42 = f7, f6, f8 195 1.1 mrg ;; 196 1.1 mrg ldf8 f34 = [up], 8 197 1.1 mrg xma.l f39 = f35, f6, f47 198 1.1 mrg ldf8 f46 = [rp], 8 199 1.1 mrg xma.hu f43 = f35, f6, f47 200 1.1 mrg ;; 201 1.1 mrg ldf8 f35 = [up], 8 202 1.1 mrg ldf8 f47 = [rp], 8 203 1.1 mrg br.cloop.dptk .grt6 204 1.1 mrg 205 1.1 mrg stf8 [r20] = f38, 8 206 1.1 mrg xma.l f36 = f32, f6, f44 207 1.1 mrg xma.hu f40 = f32, f6, f44 208 1.1 mrg ;; 209 1.1 mrg getf.sig r30 = f42 210 1.1 mrg getf.sig r27 = f39 211 1.1 mrg xma.l f37 = f33, f6, f45 212 1.1 mrg xma.hu f41 = f33, f6, f45 213 1.1 mrg ;; 214 1.1 mrg getf.sig r31 = f43 215 1.1 mrg getf.sig r24 = f36 216 1.1 mrg xma.l f38 = f34, f6, f46 217 1.1 mrg xma.hu f42 = f34, f6, f46 218 1.1 mrg ;; 219 1.1 mrg getf.sig r28 = f40 220 1.1 mrg getf.sig r25 = f37 221 1.1 mrg xma.l f39 = f35, f6, f47 222 1.1 mrg xma.hu f43 = f35, f6, f47 223 1.1 mrg br .Lcj6 224 1.1 mrg 225 1.1 mrg .grt6: 226 1.1 mrg mov r29 = 0 227 1.1 mrg xma.l f36 = f32, f6, f44 228 1.1 mrg xma.hu f40 = f32, f6, f44 229 1.1 mrg ;; 230 1.1 mrg ldf8 f32 = [up], 8 231 1.1 mrg getf.sig r26 = f38 232 1.1 mrg ;; 233 1.1 mrg getf.sig r30 = f42 234 1.1 mrg xma.l f37 = f33, f6, f45 235 1.1 mrg ldf8 f44 = [rp], 8 236 1.1 mrg xma.hu f41 = f33, f6, f45 237 1.1 mrg ;; 238 1.1 mrg ldf8 f33 = [up], 8 239 1.1 mrg getf.sig r27 = f39 240 1.1 mrg ;; 241 1.1 mrg getf.sig r31 = f43 242 1.1 mrg xma.l f38 = f34, f6, f46 243 1.1 mrg ldf8 f45 = [rp], 8 244 1.1 mrg xma.hu f42 = f34, f6, f46 245 1.1 mrg ;; 246 1.1 mrg ldf8 f34 = [up], 8 247 1.1 mrg getf.sig r24 = f36 248 1.1 mrg br .LL10 249 1.1 mrg 250 1.1 mrg 251 1.1 mrg .Lb11: ldf8 f34 = [up], 8 252 1.1 mrg ldf8 f46 = [rp], 8 253 1.1 mrg ;; 254 1.1 mrg ldf8 f35 = [up], 8 255 1.1 mrg ldf8 f47 = [rp], 8 256 1.1 mrg br.cloop.dptk .grt3 257 1.1 mrg ;; 258 1.1 mrg 259 1.1 mrg xma.l f37 = f7, f6, f8 260 1.1 mrg xma.hu f41 = f7, f6, f8 261 1.1 mrg xma.l f38 = f34, f6, f46 262 1.1 mrg xma.hu f42 = f34, f6, f46 263 1.1 mrg xma.l f39 = f35, f6, f47 264 1.1 mrg xma.hu f43 = f35, f6, f47 265 1.1 mrg ;; 266 1.1 mrg getf.sig r29 = f41 267 1.1 mrg stf8 [r20] = f37, 8 268 1.1 mrg getf.sig r26 = f38 269 1.1 mrg getf.sig r30 = f42 270 1.1 mrg getf.sig r27 = f39 271 1.1 mrg getf.sig r8 = f43 272 1.1 mrg br .Lcj3 273 1.1 mrg 274 1.1 mrg .grt3: 275 1.1 mrg ldf8 f32 = [up], 8 276 1.1 mrg xma.l f37 = f7, f6, f8 277 1.1 mrg ldf8 f44 = [rp], 8 278 1.1 mrg xma.hu f41 = f7, f6, f8 279 1.1 mrg ;; 280 1.1 mrg ldf8 f33 = [up], 8 281 1.1 mrg xma.l f38 = f34, f6, f46 282 1.1 mrg ldf8 f45 = [rp], 8 283 1.1 mrg xma.hu f42 = f34, f6, f46 284 1.1 mrg ;; 285 1.1 mrg ldf8 f34 = [up], 8 286 1.1 mrg xma.l f39 = f35, f6, f47 287 1.1 mrg ldf8 f46 = [rp], 8 288 1.1 mrg xma.hu f43 = f35, f6, f47 289 1.1 mrg ;; 290 1.1 mrg ldf8 f35 = [up], 8 291 1.1 mrg getf.sig r25 = f37 C FIXME 292 1.1 mrg ldf8 f47 = [rp], 8 293 1.1 mrg br.cloop.dptk .grt7 294 1.1 mrg 295 1.1 mrg getf.sig r29 = f41 296 1.1 mrg stf8 [r20] = f37, 8 C FIXME 297 1.1 mrg xma.l f36 = f32, f6, f44 298 1.1 mrg getf.sig r26 = f38 299 1.1 mrg xma.hu f40 = f32, f6, f44 300 1.1 mrg ;; 301 1.1 mrg getf.sig r30 = f42 302 1.1 mrg xma.l f37 = f33, f6, f45 303 1.1 mrg getf.sig r27 = f39 304 1.1 mrg xma.hu f41 = f33, f6, f45 305 1.1 mrg ;; 306 1.1 mrg getf.sig r31 = f43 307 1.1 mrg xma.l f38 = f34, f6, f46 308 1.1 mrg getf.sig r24 = f36 309 1.1 mrg xma.hu f42 = f34, f6, f46 310 1.1 mrg br .Lcj7 311 1.1 mrg 312 1.1 mrg .grt7: 313 1.1 mrg getf.sig r29 = f41 314 1.1 mrg xma.l f36 = f32, f6, f44 315 1.1 mrg mov r28 = 0 316 1.1 mrg xma.hu f40 = f32, f6, f44 317 1.1 mrg ;; 318 1.1 mrg ldf8 f32 = [up], 8 319 1.1 mrg getf.sig r26 = f38 320 1.1 mrg ;; 321 1.1 mrg getf.sig r30 = f42 322 1.1 mrg xma.l f37 = f33, f6, f45 323 1.1 mrg ldf8 f44 = [rp], 8 324 1.1 mrg xma.hu f41 = f33, f6, f45 325 1.1 mrg ;; 326 1.1 mrg ldf8 f33 = [up], 8 327 1.1 mrg getf.sig r27 = f39 328 1.1 mrg br .LL11 329 1.1 mrg 330 1.1 mrg 331 1.1 mrg .Lb00: ldf8 f33 = [up], 8 332 1.1 mrg ldf8 f45 = [rp], 8 333 1.1 mrg ;; 334 1.1 mrg ldf8 f34 = [up], 8 335 1.1 mrg ldf8 f46 = [rp], 8 336 1.1 mrg ;; 337 1.1 mrg ldf8 f35 = [up], 8 338 1.1 mrg xma.l f36 = f7, f6, f8 339 1.1 mrg ldf8 f47 = [rp], 8 340 1.1 mrg xma.hu f40 = f7, f6, f8 341 1.1 mrg br.cloop.dptk .grt4 342 1.1 mrg 343 1.1 mrg xma.l f37 = f33, f6, f45 344 1.1 mrg xma.hu f41 = f33, f6, f45 345 1.1 mrg xma.l f38 = f34, f6, f46 346 1.1 mrg xma.hu f42 = f34, f6, f46 347 1.1 mrg ;; 348 1.1 mrg getf.sig r28 = f40 349 1.1 mrg stf8 [r20] = f36, 8 350 1.1 mrg xma.l f39 = f35, f6, f47 351 1.1 mrg getf.sig r25 = f37 352 1.1 mrg xma.hu f43 = f35, f6, f47 353 1.1 mrg ;; 354 1.1 mrg getf.sig r29 = f41 355 1.1 mrg getf.sig r26 = f38 356 1.1 mrg getf.sig r30 = f42 357 1.1 mrg getf.sig r27 = f39 358 1.1 mrg br .Lcj4 359 1.1 mrg 360 1.1 mrg .grt4: 361 1.1 mrg ldf8 f32 = [up], 8 362 1.1 mrg xma.l f37 = f33, f6, f45 363 1.1 mrg ldf8 f44 = [rp], 8 364 1.1 mrg xma.hu f41 = f33, f6, f45 365 1.1 mrg ;; 366 1.1 mrg ldf8 f33 = [up], 8 367 1.1 mrg xma.l f38 = f34, f6, f46 368 1.1 mrg ldf8 f45 = [rp], 8 369 1.1 mrg xma.hu f42 = f34, f6, f46 370 1.1 mrg ;; 371 1.1 mrg ldf8 f34 = [up], 8 372 1.1 mrg getf.sig r24 = f36 C FIXME 373 1.1 mrg xma.l f39 = f35, f6, f47 374 1.1 mrg ldf8 f46 = [rp], 8 375 1.1 mrg getf.sig r28 = f40 376 1.1 mrg xma.hu f43 = f35, f6, f47 377 1.1 mrg ;; 378 1.1 mrg ldf8 f35 = [up], 8 379 1.1 mrg getf.sig r25 = f37 380 1.1 mrg ldf8 f47 = [rp], 8 381 1.1 mrg br.cloop.dptk .grt8 382 1.1 mrg 383 1.1 mrg getf.sig r29 = f41 384 1.1 mrg stf8 [r20] = f36, 8 C FIXME 385 1.1 mrg xma.l f36 = f32, f6, f44 386 1.1 mrg getf.sig r26 = f38 387 1.1 mrg getf.sig r30 = f42 388 1.1 mrg xma.hu f40 = f32, f6, f44 389 1.1 mrg ;; 390 1.1 mrg xma.l f37 = f33, f6, f45 391 1.1 mrg getf.sig r27 = f39 392 1.1 mrg xma.hu f41 = f33, f6, f45 393 1.1 mrg br .Lcj8 394 1.1 mrg 395 1.1 mrg .grt8: 396 1.1 mrg getf.sig r29 = f41 397 1.1 mrg xma.l f36 = f32, f6, f44 398 1.1 mrg mov r31 = 0 399 1.1 mrg xma.hu f40 = f32, f6, f44 400 1.1 mrg ;; 401 1.1 mrg ldf8 f32 = [up], 8 402 1.1 mrg getf.sig r26 = f38 403 1.1 mrg br .LL00 404 1.1 mrg 405 1.1 mrg 406 1.1 mrg C *** MAIN LOOP START *** 407 1.1 mrg ALIGN(32) C insn fed cycle # 408 1.1 mrg .Loop: 409 1.1 mrg .pred.rel "mutex", p6, p7 C num by i1 i2 410 1.1 mrg getf.sig r29 = f41 C 00 16 0 0 411 1.1 mrg xma.l f36 = f32, f6, f44 C 01 06,15 0 0 412 1.1 mrg (p6) add r14 = r30, r27, 1 C 02 0 0 413 1.1 mrg ldf8 f47 = [rp], 8 C 03 0 0 414 1.1 mrg xma.hu f40 = f32, f6, f44 C 04 06,15 0 0 415 1.1 mrg (p7) add r14 = r30, r27 C 05 0 0 416 1.1 mrg ;; 417 1.1 mrg .pred.rel "mutex", p6, p7 418 1.1 mrg ldf8 f32 = [up], 8 C 06 1 1 419 1.1 mrg (p6) cmp.leu p8, p9 = r14, r27 C 07 1 1 420 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r27 C 08 1 1 421 1.1 mrg getf.sig r26 = f38 C 09 25 2 1 422 1.1 mrg st8 [r20] = r14, 8 C 10 2 1 423 1.1 mrg nop.b 0 C 11 2 1 424 1.1 mrg ;; 425 1.1 mrg .LL00: 426 1.1 mrg .pred.rel "mutex", p8, p9 427 1.1 mrg getf.sig r30 = f42 C 12 28 3 2 428 1.1 mrg xma.l f37 = f33, f6, f45 C 13 18,27 3 2 429 1.1 mrg (p8) add r16 = r31, r24, 1 C 14 3 2 430 1.1 mrg ldf8 f44 = [rp], 8 C 15 3 2 431 1.1 mrg xma.hu f41 = f33, f6, f45 C 16 18,27 3 2 432 1.1 mrg (p9) add r16 = r31, r24 C 17 3 2 433 1.1 mrg ;; 434 1.1 mrg .pred.rel "mutex", p8, p9 435 1.1 mrg ldf8 f33 = [up], 8 C 18 4 3 436 1.1 mrg (p8) cmp.leu p6, p7 = r16, r24 C 19 4 3 437 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r24 C 20 4 3 438 1.1 mrg getf.sig r27 = f39 C 21 37 5 3 439 1.1 mrg st8 [r20] = r16, 8 C 22 5 3 440 1.1 mrg nop.b 0 C 23 5 3 441 1.1 mrg ;; 442 1.1 mrg .LL11: 443 1.1 mrg .pred.rel "mutex", p6, p7 444 1.1 mrg getf.sig r31 = f43 C 24 40 6 4 445 1.1 mrg xma.l f38 = f34, f6, f46 C 25 30,39 6 4 446 1.1 mrg (p6) add r14 = r28, r25, 1 C 26 6 4 447 1.1 mrg ldf8 f45 = [rp], 8 C 27 6 4 448 1.1 mrg xma.hu f42 = f34, f6, f46 C 28 30,39 6 4 449 1.1 mrg (p7) add r14 = r28, r25 C 29 6 4 450 1.1 mrg ;; 451 1.1 mrg .pred.rel "mutex", p6, p7 452 1.1 mrg ldf8 f34 = [up], 8 C 30 7 5 453 1.1 mrg (p6) cmp.leu p8, p9 = r14, r25 C 31 7 5 454 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r25 C 32 7 5 455 1.1 mrg getf.sig r24 = f36 C 33 01 8 5 456 1.1 mrg st8 [r20] = r14, 8 C 34 8 5 457 1.1 mrg nop.b 0 C 35 8 5 458 1.1 mrg ;; 459 1.1 mrg .LL10: 460 1.1 mrg .pred.rel "mutex", p8, p9 461 1.1 mrg getf.sig r28 = f40 C 36 04 9 6 462 1.1 mrg xma.l f39 = f35, f6, f47 C 37 42,03 9 6 463 1.1 mrg (p8) add r16 = r29, r26, 1 C 38 9 6 464 1.1 mrg ldf8 f46 = [rp], 8 C 39 9 6 465 1.1 mrg xma.hu f43 = f35, f6, f47 C 40 42,03 9 6 466 1.1 mrg (p9) add r16 = r29, r26 C 41 9 6 467 1.1 mrg ;; 468 1.1 mrg .pred.rel "mutex", p8, p9 469 1.1 mrg ldf8 f35 = [up], 8 C 42 10 7 470 1.1 mrg (p8) cmp.leu p6, p7 = r16, r26 C 43 10 7 471 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r26 C 44 10 7 472 1.1 mrg getf.sig r25 = f37 C 45 13 11 7 473 1.1 mrg st8 [r20] = r16, 8 C 46 11 7 474 1.1 mrg br.cloop.dptk .Loop C 47 11 7 475 1.1 mrg C *** MAIN LOOP END *** 476 1.1 mrg ;; 477 1.1 mrg .Le0: 478 1.1 mrg .pred.rel "mutex", p6, p7 479 1.1 mrg getf.sig r29 = f41 C 480 1.1 mrg xma.l f36 = f32, f6, f44 C 481 1.1 mrg (p6) add r14 = r30, r27, 1 C 482 1.1 mrg ldf8 f47 = [rp], 8 C 483 1.1 mrg xma.hu f40 = f32, f6, f44 C 484 1.1 mrg (p7) add r14 = r30, r27 C 485 1.1 mrg ;; 486 1.1 mrg .pred.rel "mutex", p6, p7 487 1.1 mrg (p6) cmp.leu p8, p9 = r14, r27 C 488 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r27 C 489 1.1 mrg getf.sig r26 = f38 C 490 1.1 mrg st8 [r20] = r14, 8 C 491 1.1 mrg ;; 492 1.1 mrg .pred.rel "mutex", p8, p9 493 1.1 mrg getf.sig r30 = f42 C 494 1.1 mrg xma.l f37 = f33, f6, f45 C 495 1.1 mrg (p8) add r16 = r31, r24, 1 C 496 1.1 mrg xma.hu f41 = f33, f6, f45 C 497 1.1 mrg (p9) add r16 = r31, r24 C 498 1.1 mrg ;; 499 1.1 mrg .pred.rel "mutex", p8, p9 500 1.1 mrg (p8) cmp.leu p6, p7 = r16, r24 C 501 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r24 C 502 1.1 mrg getf.sig r27 = f39 C 503 1.1 mrg st8 [r20] = r16, 8 C 504 1.1 mrg ;; 505 1.1 mrg .Lcj8: 506 1.1 mrg .pred.rel "mutex", p6, p7 507 1.1 mrg getf.sig r31 = f43 C 508 1.1 mrg xma.l f38 = f34, f6, f46 C 509 1.1 mrg (p6) add r14 = r28, r25, 1 C 510 1.1 mrg xma.hu f42 = f34, f6, f46 C 511 1.1 mrg (p7) add r14 = r28, r25 C 512 1.1 mrg ;; 513 1.1 mrg .pred.rel "mutex", p6, p7 514 1.1 mrg (p6) cmp.leu p8, p9 = r14, r25 C 515 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r25 C 516 1.1 mrg getf.sig r24 = f36 C 517 1.1 mrg st8 [r20] = r14, 8 C 518 1.1 mrg ;; 519 1.1 mrg .Lcj7: 520 1.1 mrg .pred.rel "mutex", p8, p9 521 1.1 mrg getf.sig r28 = f40 C 522 1.1 mrg xma.l f39 = f35, f6, f47 C 523 1.1 mrg (p8) add r16 = r29, r26, 1 C 524 1.1 mrg xma.hu f43 = f35, f6, f47 C 525 1.1 mrg (p9) add r16 = r29, r26 C 526 1.1 mrg ;; 527 1.1 mrg .pred.rel "mutex", p8, p9 528 1.1 mrg (p8) cmp.leu p6, p7 = r16, r26 C 529 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r26 C 530 1.1 mrg getf.sig r25 = f37 C 531 1.1 mrg st8 [r20] = r16, 8 C 532 1.1 mrg ;; 533 1.1 mrg .Lcj6: 534 1.1 mrg .pred.rel "mutex", p6, p7 535 1.1 mrg getf.sig r29 = f41 C 536 1.1 mrg (p6) add r14 = r30, r27, 1 C 537 1.1 mrg (p7) add r14 = r30, r27 C 538 1.1 mrg ;; 539 1.1 mrg .pred.rel "mutex", p6, p7 540 1.1 mrg (p6) cmp.leu p8, p9 = r14, r27 C 541 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r27 C 542 1.1 mrg getf.sig r26 = f38 C 543 1.1 mrg st8 [r20] = r14, 8 C 544 1.1 mrg ;; 545 1.1 mrg .Lcj5: 546 1.1 mrg .pred.rel "mutex", p8, p9 547 1.1 mrg getf.sig r30 = f42 C 548 1.1 mrg (p8) add r16 = r31, r24, 1 C 549 1.1 mrg (p9) add r16 = r31, r24 C 550 1.1 mrg ;; 551 1.1 mrg .pred.rel "mutex", p8, p9 552 1.1 mrg (p8) cmp.leu p6, p7 = r16, r24 C 553 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r24 C 554 1.1 mrg getf.sig r27 = f39 C 555 1.1 mrg st8 [r20] = r16, 8 C 556 1.1 mrg ;; 557 1.1 mrg .Lcj4: 558 1.1 mrg .pred.rel "mutex", p6, p7 559 1.1 mrg getf.sig r8 = f43 C 560 1.1 mrg (p6) add r14 = r28, r25, 1 C 561 1.1 mrg (p7) add r14 = r28, r25 C 562 1.1 mrg ;; 563 1.1 mrg .pred.rel "mutex", p6, p7 564 1.1 mrg st8 [r20] = r14, 8 C 565 1.1 mrg (p6) cmp.leu p8, p9 = r14, r25 C 566 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r25 C 567 1.1 mrg ;; 568 1.1 mrg .Lcj3: 569 1.1 mrg .pred.rel "mutex", p8, p9 570 1.1 mrg (p8) add r16 = r29, r26, 1 C 571 1.1 mrg (p9) add r16 = r29, r26 C 572 1.1 mrg ;; 573 1.1 mrg .pred.rel "mutex", p8, p9 574 1.1 mrg st8 [r20] = r16, 8 C 575 1.1 mrg (p8) cmp.leu p6, p7 = r16, r26 C 576 1.1 mrg (p9) cmp.ltu p6, p7 = r16, r26 C 577 1.1 mrg ;; 578 1.1 mrg .Lcj2: 579 1.1 mrg .pred.rel "mutex", p6, p7 580 1.1 mrg (p6) add r14 = r30, r27, 1 C 581 1.1 mrg (p7) add r14 = r30, r27 C 582 1.1 mrg ;; 583 1.1 mrg .pred.rel "mutex", p6, p7 584 1.1 mrg st8 [r20] = r14 C 585 1.1 mrg (p6) cmp.leu p8, p9 = r14, r27 C 586 1.1 mrg (p7) cmp.ltu p8, p9 = r14, r27 C 587 1.1 mrg ;; 588 1.1 mrg (p8) add r8 = 1, r8 C M I 589 1.1 mrg mov.i ar.lc = r2 C I0 590 1.1 mrg br.ret.sptk.many b0 C B 591 1.1 mrg EPILOGUE() 592 1.1 mrg ASM_END() 593
Indexes created Fri Mar 06 19:51:54 UTC 2026