1 /* 2 * Optimizations for Tiny Code Generator for QEMU 3 * 4 * Copyright (c) 2010 Samsung Electronics. 5 * Contributed by Kirill Batuzov <batuzovk@ispras.ru> 6 * 7 * Permission is hereby granted, free of charge, to any person obtaining a copy 8 * of this software and associated documentation files (the "Software"), to deal 9 * in the Software without restriction, including without limitation the rights 10 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 11 * copies of the Software, and to permit persons to whom the Software is 12 * furnished to do so, subject to the following conditions: 13 * 14 * The above copyright notice and this permission notice shall be included in 15 * all copies or substantial portions of the Software. 16 * 17 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 18 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 19 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 20 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 21 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 22 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN 23 * THE SOFTWARE. 24 */ 25 26 #include "config.h" 27 28 #include <stdlib.h> 29 #include <stdio.h> 30 31 #include "qemu-common.h" 32 #include "tcg-op.h" 33 34 #define CASE_OP_32_64(x) \ 35 glue(glue(case INDEX_op_, x), _i32): \ 36 glue(glue(case INDEX_op_, x), _i64) 37 38 typedef enum { 39 TCG_TEMP_UNDEF = 0, 40 TCG_TEMP_CONST, 41 TCG_TEMP_COPY, 42 } tcg_temp_state; 43 44 struct tcg_temp_info { 45 tcg_temp_state state; 46 uint16_t prev_copy; 47 uint16_t next_copy; 48 tcg_target_ulong val; 49 tcg_target_ulong mask; 50 }; 51 52 static struct tcg_temp_info temps[TCG_MAX_TEMPS]; 53 54 /* Reset TEMP's state to TCG_TEMP_UNDEF. If TEMP only had one copy, remove 55 the copy flag from the left temp. */ 56 static void reset_temp(TCGArg temp) 57 { 58 if (temps[temp].state == TCG_TEMP_COPY) { 59 if (temps[temp].prev_copy == temps[temp].next_copy) { 60 temps[temps[temp].next_copy].state = TCG_TEMP_UNDEF; 61 } else { 62 temps[temps[temp].next_copy].prev_copy = temps[temp].prev_copy; 63 temps[temps[temp].prev_copy].next_copy = temps[temp].next_copy; 64 } 65 } 66 temps[temp].state = TCG_TEMP_UNDEF; 67 temps[temp].mask = -1; 68 } 69 70 /* Reset all temporaries, given that there are NB_TEMPS of them. */ 71 static void reset_all_temps(int nb_temps) 72 { 73 int i; 74 for (i = 0; i < nb_temps; i++) { 75 temps[i].state = TCG_TEMP_UNDEF; 76 temps[i].mask = -1; 77 } 78 } 79 80 static int op_bits(TCGOpcode op) 81 { 82 const TCGOpDef *def = &tcg_op_defs[op]; 83 return def->flags & TCG_OPF_64BIT ? 64 : 32; 84 } 85 86 static TCGOpcode op_to_movi(TCGOpcode op) 87 { 88 switch (op_bits(op)) { 89 case 32: 90 return INDEX_op_movi_i32; 91 case 64: 92 return INDEX_op_movi_i64; 93 default: 94 fprintf(stderr, "op_to_movi: unexpected return value of " 95 "function op_bits.\n"); 96 tcg_abort(); 97 } 98 } 99 100 static TCGArg find_better_copy(TCGContext *s, TCGArg temp) 101 { 102 TCGArg i; 103 104 /* If this is already a global, we can't do better. */ 105 if (temp < s->nb_globals) { 106 return temp; 107 } 108 109 /* Search for a global first. */ 110 for (i = temps[temp].next_copy ; i != temp ; i = temps[i].next_copy) { 111 if (i < s->nb_globals) { 112 return i; 113 } 114 } 115 116 /* If it is a temp, search for a temp local. */ 117 if (!s->temps[temp].temp_local) { 118 for (i = temps[temp].next_copy ; i != temp ; i = temps[i].next_copy) { 119 if (s->temps[i].temp_local) { 120 return i; 121 } 122 } 123 } 124 125 /* Failure to find a better representation, return the same temp. */ 126 return temp; 127 } 128 129 static bool temps_are_copies(TCGArg arg1, TCGArg arg2) 130 { 131 TCGArg i; 132 133 if (arg1 == arg2) { 134 return true; 135 } 136 137 if (temps[arg1].state != TCG_TEMP_COPY 138 || temps[arg2].state != TCG_TEMP_COPY) { 139 return false; 140 } 141 142 for (i = temps[arg1].next_copy ; i != arg1 ; i = temps[i].next_copy) { 143 if (i == arg2) { 144 return true; 145 } 146 } 147 148 return false; 149 } 150 151 static void tcg_opt_gen_mov(TCGContext *s, TCGArg *gen_args, 152 TCGArg dst, TCGArg src) 153 { 154 reset_temp(dst); 155 temps[dst].mask = temps[src].mask; 156 assert(temps[src].state != TCG_TEMP_CONST); 157 158 if (s->temps[src].type == s->temps[dst].type) { 159 if (temps[src].state != TCG_TEMP_COPY) { 160 temps[src].state = TCG_TEMP_COPY; 161 temps[src].next_copy = src; 162 temps[src].prev_copy = src; 163 } 164 temps[dst].state = TCG_TEMP_COPY; 165 temps[dst].next_copy = temps[src].next_copy; 166 temps[dst].prev_copy = src; 167 temps[temps[dst].next_copy].prev_copy = dst; 168 temps[src].next_copy = dst; 169 } 170 171 gen_args[0] = dst; 172 gen_args[1] = src; 173 } 174 175 static void tcg_opt_gen_movi(TCGArg *gen_args, TCGArg dst, TCGArg val) 176 { 177 reset_temp(dst); 178 temps[dst].state = TCG_TEMP_CONST; 179 temps[dst].val = val; 180 temps[dst].mask = val; 181 gen_args[0] = dst; 182 gen_args[1] = val; 183 } 184 185 static TCGOpcode op_to_mov(TCGOpcode op) 186 { 187 switch (op_bits(op)) { 188 case 32: 189 return INDEX_op_mov_i32; 190 case 64: 191 return INDEX_op_mov_i64; 192 default: 193 fprintf(stderr, "op_to_mov: unexpected return value of " 194 "function op_bits.\n"); 195 tcg_abort(); 196 } 197 } 198 199 static TCGArg do_constant_folding_2(TCGOpcode op, TCGArg x, TCGArg y) 200 { 201 uint64_t l64, h64; 202 203 switch (op) { 204 CASE_OP_32_64(add): 205 return x + y; 206 207 CASE_OP_32_64(sub): 208 return x - y; 209 210 CASE_OP_32_64(mul): 211 return x * y; 212 213 CASE_OP_32_64(and): 214 return x & y; 215 216 CASE_OP_32_64(or): 217 return x | y; 218 219 CASE_OP_32_64(xor): 220 return x ^ y; 221 222 case INDEX_op_shl_i32: 223 return (uint32_t)x << (uint32_t)y; 224 225 case INDEX_op_shl_i64: 226 return (uint64_t)x << (uint64_t)y; 227 228 case INDEX_op_shr_i32: 229 return (uint32_t)x >> (uint32_t)y; 230 231 case INDEX_op_shr_i64: 232 return (uint64_t)x >> (uint64_t)y; 233 234 case INDEX_op_sar_i32: 235 return (int32_t)x >> (int32_t)y; 236 237 case INDEX_op_sar_i64: 238 return (int64_t)x >> (int64_t)y; 239 240 case INDEX_op_rotr_i32: 241 return ror32(x, y); 242 243 case INDEX_op_rotr_i64: 244 return ror64(x, y); 245 246 case INDEX_op_rotl_i32: 247 return rol32(x, y); 248 249 case INDEX_op_rotl_i64: 250 return rol64(x, y); 251 252 CASE_OP_32_64(not): 253 return ~x; 254 255 CASE_OP_32_64(neg): 256 return -x; 257 258 CASE_OP_32_64(andc): 259 return x & ~y; 260 261 CASE_OP_32_64(orc): 262 return x | ~y; 263 264 CASE_OP_32_64(eqv): 265 return ~(x ^ y); 266 267 CASE_OP_32_64(nand): 268 return ~(x & y); 269 270 CASE_OP_32_64(nor): 271 return ~(x | y); 272 273 CASE_OP_32_64(ext8s): 274 return (int8_t)x; 275 276 CASE_OP_32_64(ext16s): 277 return (int16_t)x; 278 279 CASE_OP_32_64(ext8u): 280 return (uint8_t)x; 281 282 CASE_OP_32_64(ext16u): 283 return (uint16_t)x; 284 285 case INDEX_op_ext32s_i64: 286 return (int32_t)x; 287 288 case INDEX_op_ext32u_i64: 289 return (uint32_t)x; 290 291 case INDEX_op_muluh_i32: 292 return ((uint64_t)(uint32_t)x * (uint32_t)y) >> 32; 293 case INDEX_op_mulsh_i32: 294 return ((int64_t)(int32_t)x * (int32_t)y) >> 32; 295 296 case INDEX_op_muluh_i64: 297 mulu64(&l64, &h64, x, y); 298 return h64; 299 case INDEX_op_mulsh_i64: 300 muls64(&l64, &h64, x, y); 301 return h64; 302 303 case INDEX_op_div_i32: 304 /* Avoid crashing on divide by zero, otherwise undefined. */ 305 return (int32_t)x / ((int32_t)y ? : 1); 306 case INDEX_op_divu_i32: 307 return (uint32_t)x / ((uint32_t)y ? : 1); 308 case INDEX_op_div_i64: 309 return (int64_t)x / ((int64_t)y ? : 1); 310 case INDEX_op_divu_i64: 311 return (uint64_t)x / ((uint64_t)y ? : 1); 312 313 case INDEX_op_rem_i32: 314 return (int32_t)x % ((int32_t)y ? : 1); 315 case INDEX_op_remu_i32: 316 return (uint32_t)x % ((uint32_t)y ? : 1); 317 case INDEX_op_rem_i64: 318 return (int64_t)x % ((int64_t)y ? : 1); 319 case INDEX_op_remu_i64: 320 return (uint64_t)x % ((uint64_t)y ? : 1); 321 322 default: 323 fprintf(stderr, 324 "Unrecognized operation %d in do_constant_folding.\n", op); 325 tcg_abort(); 326 } 327 } 328 329 static TCGArg do_constant_folding(TCGOpcode op, TCGArg x, TCGArg y) 330 { 331 TCGArg res = do_constant_folding_2(op, x, y); 332 if (op_bits(op) == 32) { 333 res &= 0xffffffff; 334 } 335 return res; 336 } 337 338 static bool do_constant_folding_cond_32(uint32_t x, uint32_t y, TCGCond c) 339 { 340 switch (c) { 341 case TCG_COND_EQ: 342 return x == y; 343 case TCG_COND_NE: 344 return x != y; 345 case TCG_COND_LT: 346 return (int32_t)x < (int32_t)y; 347 case TCG_COND_GE: 348 return (int32_t)x >= (int32_t)y; 349 case TCG_COND_LE: 350 return (int32_t)x <= (int32_t)y; 351 case TCG_COND_GT: 352 return (int32_t)x > (int32_t)y; 353 case TCG_COND_LTU: 354 return x < y; 355 case TCG_COND_GEU: 356 return x >= y; 357 case TCG_COND_LEU: 358 return x <= y; 359 case TCG_COND_GTU: 360 return x > y; 361 default: 362 tcg_abort(); 363 } 364 } 365 366 static bool do_constant_folding_cond_64(uint64_t x, uint64_t y, TCGCond c) 367 { 368 switch (c) { 369 case TCG_COND_EQ: 370 return x == y; 371 case TCG_COND_NE: 372 return x != y; 373 case TCG_COND_LT: 374 return (int64_t)x < (int64_t)y; 375 case TCG_COND_GE: 376 return (int64_t)x >= (int64_t)y; 377 case TCG_COND_LE: 378 return (int64_t)x <= (int64_t)y; 379 case TCG_COND_GT: 380 return (int64_t)x > (int64_t)y; 381 case TCG_COND_LTU: 382 return x < y; 383 case TCG_COND_GEU: 384 return x >= y; 385 case TCG_COND_LEU: 386 return x <= y; 387 case TCG_COND_GTU: 388 return x > y; 389 default: 390 tcg_abort(); 391 } 392 } 393 394 static bool do_constant_folding_cond_eq(TCGCond c) 395 { 396 switch (c) { 397 case TCG_COND_GT: 398 case TCG_COND_LTU: 399 case TCG_COND_LT: 400 case TCG_COND_GTU: 401 case TCG_COND_NE: 402 return 0; 403 case TCG_COND_GE: 404 case TCG_COND_GEU: 405 case TCG_COND_LE: 406 case TCG_COND_LEU: 407 case TCG_COND_EQ: 408 return 1; 409 default: 410 tcg_abort(); 411 } 412 } 413 414 /* Return 2 if the condition can't be simplified, and the result 415 of the condition (0 or 1) if it can */ 416 static TCGArg do_constant_folding_cond(TCGOpcode op, TCGArg x, 417 TCGArg y, TCGCond c) 418 { 419 if (temps[x].state == TCG_TEMP_CONST && temps[y].state == TCG_TEMP_CONST) { 420 switch (op_bits(op)) { 421 case 32: 422 return do_constant_folding_cond_32(temps[x].val, temps[y].val, c); 423 case 64: 424 return do_constant_folding_cond_64(temps[x].val, temps[y].val, c); 425 default: 426 tcg_abort(); 427 } 428 } else if (temps_are_copies(x, y)) { 429 return do_constant_folding_cond_eq(c); 430 } else if (temps[y].state == TCG_TEMP_CONST && temps[y].val == 0) { 431 switch (c) { 432 case TCG_COND_LTU: 433 return 0; 434 case TCG_COND_GEU: 435 return 1; 436 default: 437 return 2; 438 } 439 } else { 440 return 2; 441 } 442 } 443 444 /* Return 2 if the condition can't be simplified, and the result 445 of the condition (0 or 1) if it can */ 446 static TCGArg do_constant_folding_cond2(TCGArg *p1, TCGArg *p2, TCGCond c) 447 { 448 TCGArg al = p1[0], ah = p1[1]; 449 TCGArg bl = p2[0], bh = p2[1]; 450 451 if (temps[bl].state == TCG_TEMP_CONST 452 && temps[bh].state == TCG_TEMP_CONST) { 453 uint64_t b = ((uint64_t)temps[bh].val << 32) | (uint32_t)temps[bl].val; 454 455 if (temps[al].state == TCG_TEMP_CONST 456 && temps[ah].state == TCG_TEMP_CONST) { 457 uint64_t a; 458 a = ((uint64_t)temps[ah].val << 32) | (uint32_t)temps[al].val; 459 return do_constant_folding_cond_64(a, b, c); 460 } 461 if (b == 0) { 462 switch (c) { 463 case TCG_COND_LTU: 464 return 0; 465 case TCG_COND_GEU: 466 return 1; 467 default: 468 break; 469 } 470 } 471 } 472 if (temps_are_copies(al, bl) && temps_are_copies(ah, bh)) { 473 return do_constant_folding_cond_eq(c); 474 } 475 return 2; 476 } 477 478 static bool swap_commutative(TCGArg dest, TCGArg *p1, TCGArg *p2) 479 { 480 TCGArg a1 = *p1, a2 = *p2; 481 int sum = 0; 482 sum += temps[a1].state == TCG_TEMP_CONST; 483 sum -= temps[a2].state == TCG_TEMP_CONST; 484 485 /* Prefer the constant in second argument, and then the form 486 op a, a, b, which is better handled on non-RISC hosts. */ 487 if (sum > 0 || (sum == 0 && dest == a2)) { 488 *p1 = a2; 489 *p2 = a1; 490 return true; 491 } 492 return false; 493 } 494 495 static bool swap_commutative2(TCGArg *p1, TCGArg *p2) 496 { 497 int sum = 0; 498 sum += temps[p1[0]].state == TCG_TEMP_CONST; 499 sum += temps[p1[1]].state == TCG_TEMP_CONST; 500 sum -= temps[p2[0]].state == TCG_TEMP_CONST; 501 sum -= temps[p2[1]].state == TCG_TEMP_CONST; 502 if (sum > 0) { 503 TCGArg t; 504 t = p1[0], p1[0] = p2[0], p2[0] = t; 505 t = p1[1], p1[1] = p2[1], p2[1] = t; 506 return true; 507 } 508 return false; 509 } 510 511 /* Propagate constants and copies, fold constant expressions. */ 512 static TCGArg *tcg_constant_folding(TCGContext *s, uint16_t *tcg_opc_ptr, 513 TCGArg *args, TCGOpDef *tcg_op_defs) 514 { 515 int i, nb_ops, op_index, nb_temps, nb_globals, nb_call_args; 516 tcg_target_ulong mask, affected; 517 TCGOpcode op; 518 const TCGOpDef *def; 519 TCGArg *gen_args; 520 TCGArg tmp; 521 522 /* Array VALS has an element for each temp. 523 If this temp holds a constant then its value is kept in VALS' element. 524 If this temp is a copy of other ones then the other copies are 525 available through the doubly linked circular list. */ 526 527 nb_temps = s->nb_temps; 528 nb_globals = s->nb_globals; 529 reset_all_temps(nb_temps); 530 531 nb_ops = tcg_opc_ptr - s->gen_opc_buf; 532 gen_args = args; 533 for (op_index = 0; op_index < nb_ops; op_index++) { 534 op = s->gen_opc_buf[op_index]; 535 def = &tcg_op_defs[op]; 536 /* Do copy propagation */ 537 if (op == INDEX_op_call) { 538 int nb_oargs = args[0] >> 16; 539 int nb_iargs = args[0] & 0xffff; 540 for (i = nb_oargs + 1; i < nb_oargs + nb_iargs + 1; i++) { 541 if (temps[args[i]].state == TCG_TEMP_COPY) { 542 args[i] = find_better_copy(s, args[i]); 543 } 544 } 545 } else { 546 for (i = def->nb_oargs; i < def->nb_oargs + def->nb_iargs; i++) { 547 if (temps[args[i]].state == TCG_TEMP_COPY) { 548 args[i] = find_better_copy(s, args[i]); 549 } 550 } 551 } 552 553 /* For commutative operations make constant second argument */ 554 switch (op) { 555 CASE_OP_32_64(add): 556 CASE_OP_32_64(mul): 557 CASE_OP_32_64(and): 558 CASE_OP_32_64(or): 559 CASE_OP_32_64(xor): 560 CASE_OP_32_64(eqv): 561 CASE_OP_32_64(nand): 562 CASE_OP_32_64(nor): 563 CASE_OP_32_64(muluh): 564 CASE_OP_32_64(mulsh): 565 swap_commutative(args[0], &args[1], &args[2]); 566 break; 567 CASE_OP_32_64(brcond): 568 if (swap_commutative(-1, &args[0], &args[1])) { 569 args[2] = tcg_swap_cond(args[2]); 570 } 571 break; 572 CASE_OP_32_64(setcond): 573 if (swap_commutative(args[0], &args[1], &args[2])) { 574 args[3] = tcg_swap_cond(args[3]); 575 } 576 break; 577 CASE_OP_32_64(movcond): 578 if (swap_commutative(-1, &args[1], &args[2])) { 579 args[5] = tcg_swap_cond(args[5]); 580 } 581 /* For movcond, we canonicalize the "false" input reg to match 582 the destination reg so that the tcg backend can implement 583 a "move if true" operation. */ 584 if (swap_commutative(args[0], &args[4], &args[3])) { 585 args[5] = tcg_invert_cond(args[5]); 586 } 587 break; 588 CASE_OP_32_64(add2): 589 swap_commutative(args[0], &args[2], &args[4]); 590 swap_commutative(args[1], &args[3], &args[5]); 591 break; 592 CASE_OP_32_64(mulu2): 593 CASE_OP_32_64(muls2): 594 swap_commutative(args[0], &args[2], &args[3]); 595 break; 596 case INDEX_op_brcond2_i32: 597 if (swap_commutative2(&args[0], &args[2])) { 598 args[4] = tcg_swap_cond(args[4]); 599 } 600 break; 601 case INDEX_op_setcond2_i32: 602 if (swap_commutative2(&args[1], &args[3])) { 603 args[5] = tcg_swap_cond(args[5]); 604 } 605 break; 606 default: 607 break; 608 } 609 610 /* Simplify expressions for "shift/rot r, 0, a => movi r, 0", 611 and "sub r, 0, a => neg r, a" case. */ 612 switch (op) { 613 CASE_OP_32_64(shl): 614 CASE_OP_32_64(shr): 615 CASE_OP_32_64(sar): 616 CASE_OP_32_64(rotl): 617 CASE_OP_32_64(rotr): 618 if (temps[args[1]].state == TCG_TEMP_CONST 619 && temps[args[1]].val == 0) { 620 s->gen_opc_buf[op_index] = op_to_movi(op); 621 tcg_opt_gen_movi(gen_args, args[0], 0); 622 args += 3; 623 gen_args += 2; 624 continue; 625 } 626 break; 627 CASE_OP_32_64(sub): 628 { 629 TCGOpcode neg_op; 630 bool have_neg; 631 632 if (temps[args[2]].state == TCG_TEMP_CONST) { 633 /* Proceed with possible constant folding. */ 634 break; 635 } 636 if (op == INDEX_op_sub_i32) { 637 neg_op = INDEX_op_neg_i32; 638 have_neg = TCG_TARGET_HAS_neg_i32; 639 } else { 640 neg_op = INDEX_op_neg_i64; 641 have_neg = TCG_TARGET_HAS_neg_i64; 642 } 643 if (!have_neg) { 644 break; 645 } 646 if (temps[args[1]].state == TCG_TEMP_CONST 647 && temps[args[1]].val == 0) { 648 s->gen_opc_buf[op_index] = neg_op; 649 reset_temp(args[0]); 650 gen_args[0] = args[0]; 651 gen_args[1] = args[2]; 652 args += 3; 653 gen_args += 2; 654 continue; 655 } 656 } 657 break; 658 default: 659 break; 660 } 661 662 /* Simplify expression for "op r, a, 0 => mov r, a" cases */ 663 switch (op) { 664 CASE_OP_32_64(add): 665 CASE_OP_32_64(sub): 666 CASE_OP_32_64(shl): 667 CASE_OP_32_64(shr): 668 CASE_OP_32_64(sar): 669 CASE_OP_32_64(rotl): 670 CASE_OP_32_64(rotr): 671 CASE_OP_32_64(or): 672 CASE_OP_32_64(xor): 673 if (temps[args[1]].state == TCG_TEMP_CONST) { 674 /* Proceed with possible constant folding. */ 675 break; 676 } 677 if (temps[args[2]].state == TCG_TEMP_CONST 678 && temps[args[2]].val == 0) { 679 if (temps_are_copies(args[0], args[1])) { 680 s->gen_opc_buf[op_index] = INDEX_op_nop; 681 } else { 682 s->gen_opc_buf[op_index] = op_to_mov(op); 683 tcg_opt_gen_mov(s, gen_args, args[0], args[1]); 684 gen_args += 2; 685 } 686 args += 3; 687 continue; 688 } 689 break; 690 default: 691 break; 692 } 693 694 /* Simplify using known-zero bits */ 695 mask = -1; 696 affected = -1; 697 switch (op) { 698 CASE_OP_32_64(ext8s): 699 if ((temps[args[1]].mask & 0x80) != 0) { 700 break; 701 } 702 CASE_OP_32_64(ext8u): 703 mask = 0xff; 704 goto and_const; 705 CASE_OP_32_64(ext16s): 706 if ((temps[args[1]].mask & 0x8000) != 0) { 707 break; 708 } 709 CASE_OP_32_64(ext16u): 710 mask = 0xffff; 711 goto and_const; 712 case INDEX_op_ext32s_i64: 713 if ((temps[args[1]].mask & 0x80000000) != 0) { 714 break; 715 } 716 case INDEX_op_ext32u_i64: 717 mask = 0xffffffffU; 718 goto and_const; 719 720 CASE_OP_32_64(and): 721 mask = temps[args[2]].mask; 722 if (temps[args[2]].state == TCG_TEMP_CONST) { 723 and_const: 724 affected = temps[args[1]].mask & ~mask; 725 } 726 mask = temps[args[1]].mask & mask; 727 break; 728 729 CASE_OP_32_64(sar): 730 if (temps[args[2]].state == TCG_TEMP_CONST) { 731 mask = ((tcg_target_long)temps[args[1]].mask 732 >> temps[args[2]].val); 733 } 734 break; 735 736 CASE_OP_32_64(shr): 737 if (temps[args[2]].state == TCG_TEMP_CONST) { 738 mask = temps[args[1]].mask >> temps[args[2]].val; 739 } 740 break; 741 742 CASE_OP_32_64(shl): 743 if (temps[args[2]].state == TCG_TEMP_CONST) { 744 mask = temps[args[1]].mask << temps[args[2]].val; 745 } 746 break; 747 748 CASE_OP_32_64(neg): 749 /* Set to 1 all bits to the left of the rightmost. */ 750 mask = -(temps[args[1]].mask & -temps[args[1]].mask); 751 break; 752 753 CASE_OP_32_64(deposit): 754 tmp = ((1ull << args[4]) - 1); 755 mask = ((temps[args[1]].mask & ~(tmp << args[3])) 756 | ((temps[args[2]].mask & tmp) << args[3])); 757 break; 758 759 CASE_OP_32_64(or): 760 CASE_OP_32_64(xor): 761 mask = temps[args[1]].mask | temps[args[2]].mask; 762 break; 763 764 CASE_OP_32_64(setcond): 765 mask = 1; 766 break; 767 768 CASE_OP_32_64(movcond): 769 mask = temps[args[3]].mask | temps[args[4]].mask; 770 break; 771 772 default: 773 break; 774 } 775 776 if (mask == 0) { 777 assert(def->nb_oargs == 1); 778 s->gen_opc_buf[op_index] = op_to_movi(op); 779 tcg_opt_gen_movi(gen_args, args[0], 0); 780 args += def->nb_oargs + def->nb_iargs + def->nb_cargs; 781 gen_args += 2; 782 continue; 783 } 784 if (affected == 0) { 785 assert(def->nb_oargs == 1); 786 if (temps_are_copies(args[0], args[1])) { 787 s->gen_opc_buf[op_index] = INDEX_op_nop; 788 } else if (temps[args[1]].state != TCG_TEMP_CONST) { 789 s->gen_opc_buf[op_index] = op_to_mov(op); 790 tcg_opt_gen_mov(s, gen_args, args[0], args[1]); 791 gen_args += 2; 792 } else { 793 s->gen_opc_buf[op_index] = op_to_movi(op); 794 tcg_opt_gen_movi(gen_args, args[0], temps[args[1]].val); 795 gen_args += 2; 796 } 797 args += def->nb_iargs + 1; 798 continue; 799 } 800 801 /* Simplify expression for "op r, a, 0 => movi r, 0" cases */ 802 switch (op) { 803 CASE_OP_32_64(and): 804 CASE_OP_32_64(mul): 805 CASE_OP_32_64(muluh): 806 CASE_OP_32_64(mulsh): 807 if ((temps[args[2]].state == TCG_TEMP_CONST 808 && temps[args[2]].val == 0)) { 809 s->gen_opc_buf[op_index] = op_to_movi(op); 810 tcg_opt_gen_movi(gen_args, args[0], 0); 811 args += 3; 812 gen_args += 2; 813 continue; 814 } 815 break; 816 default: 817 break; 818 } 819 820 /* Simplify expression for "op r, a, a => mov r, a" cases */ 821 switch (op) { 822 CASE_OP_32_64(or): 823 CASE_OP_32_64(and): 824 if (temps_are_copies(args[1], args[2])) { 825 if (temps_are_copies(args[0], args[1])) { 826 s->gen_opc_buf[op_index] = INDEX_op_nop; 827 } else { 828 s->gen_opc_buf[op_index] = op_to_mov(op); 829 tcg_opt_gen_mov(s, gen_args, args[0], args[1]); 830 gen_args += 2; 831 } 832 args += 3; 833 continue; 834 } 835 break; 836 default: 837 break; 838 } 839 840 /* Simplify expression for "op r, a, a => movi r, 0" cases */ 841 switch (op) { 842 CASE_OP_32_64(sub): 843 CASE_OP_32_64(xor): 844 if (temps_are_copies(args[1], args[2])) { 845 s->gen_opc_buf[op_index] = op_to_movi(op); 846 tcg_opt_gen_movi(gen_args, args[0], 0); 847 gen_args += 2; 848 args += 3; 849 continue; 850 } 851 break; 852 default: 853 break; 854 } 855 856 /* Propagate constants through copy operations and do constant 857 folding. Constants will be substituted to arguments by register 858 allocator where needed and possible. Also detect copies. */ 859 switch (op) { 860 CASE_OP_32_64(mov): 861 if (temps_are_copies(args[0], args[1])) { 862 args += 2; 863 s->gen_opc_buf[op_index] = INDEX_op_nop; 864 break; 865 } 866 if (temps[args[1]].state != TCG_TEMP_CONST) { 867 tcg_opt_gen_mov(s, gen_args, args[0], args[1]); 868 gen_args += 2; 869 args += 2; 870 break; 871 } 872 /* Source argument is constant. Rewrite the operation and 873 let movi case handle it. */ 874 op = op_to_movi(op); 875 s->gen_opc_buf[op_index] = op; 876 args[1] = temps[args[1]].val; 877 /* fallthrough */ 878 CASE_OP_32_64(movi): 879 tcg_opt_gen_movi(gen_args, args[0], args[1]); 880 gen_args += 2; 881 args += 2; 882 break; 883 884 CASE_OP_32_64(not): 885 CASE_OP_32_64(neg): 886 CASE_OP_32_64(ext8s): 887 CASE_OP_32_64(ext8u): 888 CASE_OP_32_64(ext16s): 889 CASE_OP_32_64(ext16u): 890 case INDEX_op_ext32s_i64: 891 case INDEX_op_ext32u_i64: 892 if (temps[args[1]].state == TCG_TEMP_CONST) { 893 s->gen_opc_buf[op_index] = op_to_movi(op); 894 tmp = do_constant_folding(op, temps[args[1]].val, 0); 895 tcg_opt_gen_movi(gen_args, args[0], tmp); 896 gen_args += 2; 897 args += 2; 898 break; 899 } 900 goto do_default; 901 902 CASE_OP_32_64(add): 903 CASE_OP_32_64(sub): 904 CASE_OP_32_64(mul): 905 CASE_OP_32_64(or): 906 CASE_OP_32_64(and): 907 CASE_OP_32_64(xor): 908 CASE_OP_32_64(shl): 909 CASE_OP_32_64(shr): 910 CASE_OP_32_64(sar): 911 CASE_OP_32_64(rotl): 912 CASE_OP_32_64(rotr): 913 CASE_OP_32_64(andc): 914 CASE_OP_32_64(orc): 915 CASE_OP_32_64(eqv): 916 CASE_OP_32_64(nand): 917 CASE_OP_32_64(nor): 918 CASE_OP_32_64(muluh): 919 CASE_OP_32_64(mulsh): 920 CASE_OP_32_64(div): 921 CASE_OP_32_64(divu): 922 CASE_OP_32_64(rem): 923 CASE_OP_32_64(remu): 924 if (temps[args[1]].state == TCG_TEMP_CONST 925 && temps[args[2]].state == TCG_TEMP_CONST) { 926 s->gen_opc_buf[op_index] = op_to_movi(op); 927 tmp = do_constant_folding(op, temps[args[1]].val, 928 temps[args[2]].val); 929 tcg_opt_gen_movi(gen_args, args[0], tmp); 930 gen_args += 2; 931 args += 3; 932 break; 933 } 934 goto do_default; 935 936 CASE_OP_32_64(deposit): 937 if (temps[args[1]].state == TCG_TEMP_CONST 938 && temps[args[2]].state == TCG_TEMP_CONST) { 939 s->gen_opc_buf[op_index] = op_to_movi(op); 940 tmp = ((1ull << args[4]) - 1); 941 tmp = (temps[args[1]].val & ~(tmp << args[3])) 942 | ((temps[args[2]].val & tmp) << args[3]); 943 tcg_opt_gen_movi(gen_args, args[0], tmp); 944 gen_args += 2; 945 args += 5; 946 break; 947 } 948 goto do_default; 949 950 CASE_OP_32_64(setcond): 951 tmp = do_constant_folding_cond(op, args[1], args[2], args[3]); 952 if (tmp != 2) { 953 s->gen_opc_buf[op_index] = op_to_movi(op); 954 tcg_opt_gen_movi(gen_args, args[0], tmp); 955 gen_args += 2; 956 args += 4; 957 break; 958 } 959 goto do_default; 960 961 CASE_OP_32_64(brcond): 962 tmp = do_constant_folding_cond(op, args[0], args[1], args[2]); 963 if (tmp != 2) { 964 if (tmp) { 965 reset_all_temps(nb_temps); 966 s->gen_opc_buf[op_index] = INDEX_op_br; 967 gen_args[0] = args[3]; 968 gen_args += 1; 969 } else { 970 s->gen_opc_buf[op_index] = INDEX_op_nop; 971 } 972 args += 4; 973 break; 974 } 975 goto do_default; 976 977 CASE_OP_32_64(movcond): 978 tmp = do_constant_folding_cond(op, args[1], args[2], args[5]); 979 if (tmp != 2) { 980 if (temps_are_copies(args[0], args[4-tmp])) { 981 s->gen_opc_buf[op_index] = INDEX_op_nop; 982 } else if (temps[args[4-tmp]].state == TCG_TEMP_CONST) { 983 s->gen_opc_buf[op_index] = op_to_movi(op); 984 tcg_opt_gen_movi(gen_args, args[0], temps[args[4-tmp]].val); 985 gen_args += 2; 986 } else { 987 s->gen_opc_buf[op_index] = op_to_mov(op); 988 tcg_opt_gen_mov(s, gen_args, args[0], args[4-tmp]); 989 gen_args += 2; 990 } 991 args += 6; 992 break; 993 } 994 goto do_default; 995 996 case INDEX_op_add2_i32: 997 case INDEX_op_sub2_i32: 998 if (temps[args[2]].state == TCG_TEMP_CONST 999 && temps[args[3]].state == TCG_TEMP_CONST 1000 && temps[args[4]].state == TCG_TEMP_CONST 1001 && temps[args[5]].state == TCG_TEMP_CONST) { 1002 uint32_t al = temps[args[2]].val; 1003 uint32_t ah = temps[args[3]].val; 1004 uint32_t bl = temps[args[4]].val; 1005 uint32_t bh = temps[args[5]].val; 1006 uint64_t a = ((uint64_t)ah << 32) | al; 1007 uint64_t b = ((uint64_t)bh << 32) | bl; 1008 TCGArg rl, rh; 1009 1010 if (op == INDEX_op_add2_i32) { 1011 a += b; 1012 } else { 1013 a -= b; 1014 } 1015 1016 /* We emit the extra nop when we emit the add2/sub2. */ 1017 assert(s->gen_opc_buf[op_index + 1] == INDEX_op_nop); 1018 1019 rl = args[0]; 1020 rh = args[1]; 1021 s->gen_opc_buf[op_index] = INDEX_op_movi_i32; 1022 s->gen_opc_buf[++op_index] = INDEX_op_movi_i32; 1023 tcg_opt_gen_movi(&gen_args[0], rl, (uint32_t)a); 1024 tcg_opt_gen_movi(&gen_args[2], rh, (uint32_t)(a >> 32)); 1025 gen_args += 4; 1026 args += 6; 1027 break; 1028 } 1029 goto do_default; 1030 1031 case INDEX_op_mulu2_i32: 1032 if (temps[args[2]].state == TCG_TEMP_CONST 1033 && temps[args[3]].state == TCG_TEMP_CONST) { 1034 uint32_t a = temps[args[2]].val; 1035 uint32_t b = temps[args[3]].val; 1036 uint64_t r = (uint64_t)a * b; 1037 TCGArg rl, rh; 1038 1039 /* We emit the extra nop when we emit the mulu2. */ 1040 assert(s->gen_opc_buf[op_index + 1] == INDEX_op_nop); 1041 1042 rl = args[0]; 1043 rh = args[1]; 1044 s->gen_opc_buf[op_index] = INDEX_op_movi_i32; 1045 s->gen_opc_buf[++op_index] = INDEX_op_movi_i32; 1046 tcg_opt_gen_movi(&gen_args[0], rl, (uint32_t)r); 1047 tcg_opt_gen_movi(&gen_args[2], rh, (uint32_t)(r >> 32)); 1048 gen_args += 4; 1049 args += 4; 1050 break; 1051 } 1052 goto do_default; 1053 1054 case INDEX_op_brcond2_i32: 1055 tmp = do_constant_folding_cond2(&args[0], &args[2], args[4]); 1056 if (tmp != 2) { 1057 if (tmp) { 1058 reset_all_temps(nb_temps); 1059 s->gen_opc_buf[op_index] = INDEX_op_br; 1060 gen_args[0] = args[5]; 1061 gen_args += 1; 1062 } else { 1063 s->gen_opc_buf[op_index] = INDEX_op_nop; 1064 } 1065 } else if ((args[4] == TCG_COND_LT || args[4] == TCG_COND_GE) 1066 && temps[args[2]].state == TCG_TEMP_CONST 1067 && temps[args[3]].state == TCG_TEMP_CONST 1068 && temps[args[2]].val == 0 1069 && temps[args[3]].val == 0) { 1070 /* Simplify LT/GE comparisons vs zero to a single compare 1071 vs the high word of the input. */ 1072 reset_all_temps(nb_temps); 1073 s->gen_opc_buf[op_index] = INDEX_op_brcond_i32; 1074 gen_args[0] = args[1]; 1075 gen_args[1] = args[3]; 1076 gen_args[2] = args[4]; 1077 gen_args[3] = args[5]; 1078 gen_args += 4; 1079 } else { 1080 goto do_default; 1081 } 1082 args += 6; 1083 break; 1084 1085 case INDEX_op_setcond2_i32: 1086 tmp = do_constant_folding_cond2(&args[1], &args[3], args[5]); 1087 if (tmp != 2) { 1088 s->gen_opc_buf[op_index] = INDEX_op_movi_i32; 1089 tcg_opt_gen_movi(gen_args, args[0], tmp); 1090 gen_args += 2; 1091 } else if ((args[5] == TCG_COND_LT || args[5] == TCG_COND_GE) 1092 && temps[args[3]].state == TCG_TEMP_CONST 1093 && temps[args[4]].state == TCG_TEMP_CONST 1094 && temps[args[3]].val == 0 1095 && temps[args[4]].val == 0) { 1096 /* Simplify LT/GE comparisons vs zero to a single compare 1097 vs the high word of the input. */ 1098 s->gen_opc_buf[op_index] = INDEX_op_setcond_i32; 1099 reset_temp(args[0]); 1100 gen_args[0] = args[0]; 1101 gen_args[1] = args[2]; 1102 gen_args[2] = args[4]; 1103 gen_args[3] = args[5]; 1104 gen_args += 4; 1105 } else { 1106 goto do_default; 1107 } 1108 args += 6; 1109 break; 1110 1111 case INDEX_op_call: 1112 nb_call_args = (args[0] >> 16) + (args[0] & 0xffff); 1113 if (!(args[nb_call_args + 1] & (TCG_CALL_NO_READ_GLOBALS | 1114 TCG_CALL_NO_WRITE_GLOBALS))) { 1115 for (i = 0; i < nb_globals; i++) { 1116 reset_temp(i); 1117 } 1118 } 1119 for (i = 0; i < (args[0] >> 16); i++) { 1120 reset_temp(args[i + 1]); 1121 } 1122 i = nb_call_args + 3; 1123 while (i) { 1124 *gen_args = *args; 1125 args++; 1126 gen_args++; 1127 i--; 1128 } 1129 break; 1130 1131 default: 1132 do_default: 1133 /* Default case: we know nothing about operation (or were unable 1134 to compute the operation result) so no propagation is done. 1135 We trash everything if the operation is the end of a basic 1136 block, otherwise we only trash the output args. "mask" is 1137 the non-zero bits mask for the first output arg. */ 1138 if (def->flags & TCG_OPF_BB_END) { 1139 reset_all_temps(nb_temps); 1140 } else { 1141 for (i = 0; i < def->nb_oargs; i++) { 1142 reset_temp(args[i]); 1143 } 1144 } 1145 for (i = 0; i < def->nb_args; i++) { 1146 gen_args[i] = args[i]; 1147 } 1148 args += def->nb_args; 1149 gen_args += def->nb_args; 1150 break; 1151 } 1152 } 1153 1154 return gen_args; 1155 } 1156 1157 TCGArg *tcg_optimize(TCGContext *s, uint16_t *tcg_opc_ptr, 1158 TCGArg *args, TCGOpDef *tcg_op_defs) 1159 { 1160 TCGArg *res; 1161 res = tcg_constant_folding(s, tcg_opc_ptr, args, tcg_op_defs); 1162 return res; 1163 } 1164