1 /* 2 * QEMU float support 3 * 4 * Derived from SoftFloat. 5 */ 6 7 /*============================================================================ 8 9 This C source file is part of the SoftFloat IEC/IEEE Floating-point Arithmetic 10 Package, Release 2b. 11 12 Written by John R. Hauser. This work was made possible in part by the 13 International Computer Science Institute, located at Suite 600, 1947 Center 14 Street, Berkeley, California 94704. Funding was partially provided by the 15 National Science Foundation under grant MIP-9311980. The original version 16 of this code was written as part of a project to build a fixed-point vector 17 processor in collaboration with the University of California at Berkeley, 18 overseen by Profs. Nelson Morgan and John Wawrzynek. More information 19 is available through the Web page `http://www.cs.berkeley.edu/~jhauser/ 20 arithmetic/SoftFloat.html'. 21 22 THIS SOFTWARE IS DISTRIBUTED AS IS, FOR FREE. Although reasonable effort has 23 been made to avoid it, THIS SOFTWARE MAY CONTAIN FAULTS THAT WILL AT TIMES 24 RESULT IN INCORRECT BEHAVIOR. USE OF THIS SOFTWARE IS RESTRICTED TO PERSONS 25 AND ORGANIZATIONS WHO CAN AND WILL TAKE FULL RESPONSIBILITY FOR ALL LOSSES, 26 COSTS, OR OTHER PROBLEMS THEY INCUR DUE TO THE SOFTWARE, AND WHO FURTHERMORE 27 EFFECTIVELY INDEMNIFY JOHN HAUSER AND THE INTERNATIONAL COMPUTER SCIENCE 28 INSTITUTE (possibly via similar legal warning) AGAINST ALL LOSSES, COSTS, OR 29 OTHER PROBLEMS INCURRED BY THEIR CUSTOMERS AND CLIENTS DUE TO THE SOFTWARE. 30 31 Derivative works are acceptable, even for commercial purposes, so long as 32 (1) the source code for the derivative work includes prominent notice that 33 the work is derivative, and (2) the source code includes prominent notice with 34 these four paragraphs for those parts of this code that are retained. 35 36 =============================================================================*/ 37 38 /* softfloat (and in particular the code in softfloat-specialize.h) is 39 * target-dependent and needs the TARGET_* macros. 40 */ 41 #include "config.h" 42 43 #include "fpu/softfloat.h" 44 45 /* We only need stdlib for abort() */ 46 #include <stdlib.h> 47 48 /*---------------------------------------------------------------------------- 49 | Primitive arithmetic functions, including multi-word arithmetic, and 50 | division and square root approximations. (Can be specialized to target if 51 | desired.) 52 *----------------------------------------------------------------------------*/ 53 #include "softfloat-macros.h" 54 55 /*---------------------------------------------------------------------------- 56 | Functions and definitions to determine: (1) whether tininess for underflow 57 | is detected before or after rounding by default, (2) what (if anything) 58 | happens when exceptions are raised, (3) how signaling NaNs are distinguished 59 | from quiet NaNs, (4) the default generated quiet NaNs, and (5) how NaNs 60 | are propagated from function inputs to output. These details are target- 61 | specific. 62 *----------------------------------------------------------------------------*/ 63 #include "softfloat-specialize.h" 64 65 /*---------------------------------------------------------------------------- 66 | Returns the fraction bits of the half-precision floating-point value `a'. 67 *----------------------------------------------------------------------------*/ 68 69 static inline uint32_t extractFloat16Frac(float16 a) 70 { 71 return float16_val(a) & 0x3ff; 72 } 73 74 /*---------------------------------------------------------------------------- 75 | Returns the exponent bits of the half-precision floating-point value `a'. 76 *----------------------------------------------------------------------------*/ 77 78 static inline int_fast16_t extractFloat16Exp(float16 a) 79 { 80 return (float16_val(a) >> 10) & 0x1f; 81 } 82 83 /*---------------------------------------------------------------------------- 84 | Returns the sign bit of the single-precision floating-point value `a'. 85 *----------------------------------------------------------------------------*/ 86 87 static inline flag extractFloat16Sign(float16 a) 88 { 89 return float16_val(a)>>15; 90 } 91 92 /*---------------------------------------------------------------------------- 93 | Takes a 64-bit fixed-point value `absZ' with binary point between bits 6 94 | and 7, and returns the properly rounded 32-bit integer corresponding to the 95 | input. If `zSign' is 1, the input is negated before being converted to an 96 | integer. Bit 63 of `absZ' must be zero. Ordinarily, the fixed-point input 97 | is simply rounded to an integer, with the inexact exception raised if the 98 | input cannot be represented exactly as an integer. However, if the fixed- 99 | point input is too large, the invalid exception is raised and the largest 100 | positive or negative integer is returned. 101 *----------------------------------------------------------------------------*/ 102 103 static int32 roundAndPackInt32( flag zSign, uint64_t absZ STATUS_PARAM) 104 { 105 int8 roundingMode; 106 flag roundNearestEven; 107 int8 roundIncrement, roundBits; 108 int32_t z; 109 110 roundingMode = STATUS(float_rounding_mode); 111 roundNearestEven = ( roundingMode == float_round_nearest_even ); 112 switch (roundingMode) { 113 case float_round_nearest_even: 114 case float_round_ties_away: 115 roundIncrement = 0x40; 116 break; 117 case float_round_to_zero: 118 roundIncrement = 0; 119 break; 120 case float_round_up: 121 roundIncrement = zSign ? 0 : 0x7f; 122 break; 123 case float_round_down: 124 roundIncrement = zSign ? 0x7f : 0; 125 break; 126 default: 127 abort(); 128 } 129 roundBits = absZ & 0x7F; 130 absZ = ( absZ + roundIncrement )>>7; 131 absZ &= ~ ( ( ( roundBits ^ 0x40 ) == 0 ) & roundNearestEven ); 132 z = absZ; 133 if ( zSign ) z = - z; 134 if ( ( absZ>>32 ) || ( z && ( ( z < 0 ) ^ zSign ) ) ) { 135 float_raise( float_flag_invalid STATUS_VAR); 136 return zSign ? (int32_t) 0x80000000 : 0x7FFFFFFF; 137 } 138 if ( roundBits ) STATUS(float_exception_flags) |= float_flag_inexact; 139 return z; 140 141 } 142 143 /*---------------------------------------------------------------------------- 144 | Takes the 128-bit fixed-point value formed by concatenating `absZ0' and 145 | `absZ1', with binary point between bits 63 and 64 (between the input words), 146 | and returns the properly rounded 64-bit integer corresponding to the input. 147 | If `zSign' is 1, the input is negated before being converted to an integer. 148 | Ordinarily, the fixed-point input is simply rounded to an integer, with 149 | the inexact exception raised if the input cannot be represented exactly as 150 | an integer. However, if the fixed-point input is too large, the invalid 151 | exception is raised and the largest positive or negative integer is 152 | returned. 153 *----------------------------------------------------------------------------*/ 154 155 static int64 roundAndPackInt64( flag zSign, uint64_t absZ0, uint64_t absZ1 STATUS_PARAM) 156 { 157 int8 roundingMode; 158 flag roundNearestEven, increment; 159 int64_t z; 160 161 roundingMode = STATUS(float_rounding_mode); 162 roundNearestEven = ( roundingMode == float_round_nearest_even ); 163 switch (roundingMode) { 164 case float_round_nearest_even: 165 case float_round_ties_away: 166 increment = ((int64_t) absZ1 < 0); 167 break; 168 case float_round_to_zero: 169 increment = 0; 170 break; 171 case float_round_up: 172 increment = !zSign && absZ1; 173 break; 174 case float_round_down: 175 increment = zSign && absZ1; 176 break; 177 default: 178 abort(); 179 } 180 if ( increment ) { 181 ++absZ0; 182 if ( absZ0 == 0 ) goto overflow; 183 absZ0 &= ~ ( ( (uint64_t) ( absZ1<<1 ) == 0 ) & roundNearestEven ); 184 } 185 z = absZ0; 186 if ( zSign ) z = - z; 187 if ( z && ( ( z < 0 ) ^ zSign ) ) { 188 overflow: 189 float_raise( float_flag_invalid STATUS_VAR); 190 return 191 zSign ? (int64_t) LIT64( 0x8000000000000000 ) 192 : LIT64( 0x7FFFFFFFFFFFFFFF ); 193 } 194 if ( absZ1 ) STATUS(float_exception_flags) |= float_flag_inexact; 195 return z; 196 197 } 198 199 /*---------------------------------------------------------------------------- 200 | Takes the 128-bit fixed-point value formed by concatenating `absZ0' and 201 | `absZ1', with binary point between bits 63 and 64 (between the input words), 202 | and returns the properly rounded 64-bit unsigned integer corresponding to the 203 | input. Ordinarily, the fixed-point input is simply rounded to an integer, 204 | with the inexact exception raised if the input cannot be represented exactly 205 | as an integer. However, if the fixed-point input is too large, the invalid 206 | exception is raised and the largest unsigned integer is returned. 207 *----------------------------------------------------------------------------*/ 208 209 static int64 roundAndPackUint64(flag zSign, uint64_t absZ0, 210 uint64_t absZ1 STATUS_PARAM) 211 { 212 int8 roundingMode; 213 flag roundNearestEven, increment; 214 215 roundingMode = STATUS(float_rounding_mode); 216 roundNearestEven = (roundingMode == float_round_nearest_even); 217 switch (roundingMode) { 218 case float_round_nearest_even: 219 case float_round_ties_away: 220 increment = ((int64_t)absZ1 < 0); 221 break; 222 case float_round_to_zero: 223 increment = 0; 224 break; 225 case float_round_up: 226 increment = !zSign && absZ1; 227 break; 228 case float_round_down: 229 increment = zSign && absZ1; 230 break; 231 default: 232 abort(); 233 } 234 if (increment) { 235 ++absZ0; 236 if (absZ0 == 0) { 237 float_raise(float_flag_invalid STATUS_VAR); 238 return LIT64(0xFFFFFFFFFFFFFFFF); 239 } 240 absZ0 &= ~(((uint64_t)(absZ1<<1) == 0) & roundNearestEven); 241 } 242 243 if (zSign && absZ0) { 244 float_raise(float_flag_invalid STATUS_VAR); 245 return 0; 246 } 247 248 if (absZ1) { 249 STATUS(float_exception_flags) |= float_flag_inexact; 250 } 251 return absZ0; 252 } 253 254 /*---------------------------------------------------------------------------- 255 | Returns the fraction bits of the single-precision floating-point value `a'. 256 *----------------------------------------------------------------------------*/ 257 258 static inline uint32_t extractFloat32Frac( float32 a ) 259 { 260 261 return float32_val(a) & 0x007FFFFF; 262 263 } 264 265 /*---------------------------------------------------------------------------- 266 | Returns the exponent bits of the single-precision floating-point value `a'. 267 *----------------------------------------------------------------------------*/ 268 269 static inline int_fast16_t extractFloat32Exp(float32 a) 270 { 271 272 return ( float32_val(a)>>23 ) & 0xFF; 273 274 } 275 276 /*---------------------------------------------------------------------------- 277 | Returns the sign bit of the single-precision floating-point value `a'. 278 *----------------------------------------------------------------------------*/ 279 280 static inline flag extractFloat32Sign( float32 a ) 281 { 282 283 return float32_val(a)>>31; 284 285 } 286 287 /*---------------------------------------------------------------------------- 288 | If `a' is denormal and we are in flush-to-zero mode then set the 289 | input-denormal exception and return zero. Otherwise just return the value. 290 *----------------------------------------------------------------------------*/ 291 float32 float32_squash_input_denormal(float32 a STATUS_PARAM) 292 { 293 if (STATUS(flush_inputs_to_zero)) { 294 if (extractFloat32Exp(a) == 0 && extractFloat32Frac(a) != 0) { 295 float_raise(float_flag_input_denormal STATUS_VAR); 296 return make_float32(float32_val(a) & 0x80000000); 297 } 298 } 299 return a; 300 } 301 302 /*---------------------------------------------------------------------------- 303 | Normalizes the subnormal single-precision floating-point value represented 304 | by the denormalized significand `aSig'. The normalized exponent and 305 | significand are stored at the locations pointed to by `zExpPtr' and 306 | `zSigPtr', respectively. 307 *----------------------------------------------------------------------------*/ 308 309 static void 310 normalizeFloat32Subnormal(uint32_t aSig, int_fast16_t *zExpPtr, uint32_t *zSigPtr) 311 { 312 int8 shiftCount; 313 314 shiftCount = countLeadingZeros32( aSig ) - 8; 315 *zSigPtr = aSig<<shiftCount; 316 *zExpPtr = 1 - shiftCount; 317 318 } 319 320 /*---------------------------------------------------------------------------- 321 | Packs the sign `zSign', exponent `zExp', and significand `zSig' into a 322 | single-precision floating-point value, returning the result. After being 323 | shifted into the proper positions, the three fields are simply added 324 | together to form the result. This means that any integer portion of `zSig' 325 | will be added into the exponent. Since a properly normalized significand 326 | will have an integer portion equal to 1, the `zExp' input should be 1 less 327 | than the desired result exponent whenever `zSig' is a complete, normalized 328 | significand. 329 *----------------------------------------------------------------------------*/ 330 331 static inline float32 packFloat32(flag zSign, int_fast16_t zExp, uint32_t zSig) 332 { 333 334 return make_float32( 335 ( ( (uint32_t) zSign )<<31 ) + ( ( (uint32_t) zExp )<<23 ) + zSig); 336 337 } 338 339 /*---------------------------------------------------------------------------- 340 | Takes an abstract floating-point value having sign `zSign', exponent `zExp', 341 | and significand `zSig', and returns the proper single-precision floating- 342 | point value corresponding to the abstract input. Ordinarily, the abstract 343 | value is simply rounded and packed into the single-precision format, with 344 | the inexact exception raised if the abstract input cannot be represented 345 | exactly. However, if the abstract value is too large, the overflow and 346 | inexact exceptions are raised and an infinity or maximal finite value is 347 | returned. If the abstract value is too small, the input value is rounded to 348 | a subnormal number, and the underflow and inexact exceptions are raised if 349 | the abstract input cannot be represented exactly as a subnormal single- 350 | precision floating-point number. 351 | The input significand `zSig' has its binary point between bits 30 352 | and 29, which is 7 bits to the left of the usual location. This shifted 353 | significand must be normalized or smaller. If `zSig' is not normalized, 354 | `zExp' must be 0; in that case, the result returned is a subnormal number, 355 | and it must not require rounding. In the usual case that `zSig' is 356 | normalized, `zExp' must be 1 less than the ``true'' floating-point exponent. 357 | The handling of underflow and overflow follows the IEC/IEEE Standard for 358 | Binary Floating-Point Arithmetic. 359 *----------------------------------------------------------------------------*/ 360 361 static float32 roundAndPackFloat32(flag zSign, int_fast16_t zExp, uint32_t zSig STATUS_PARAM) 362 { 363 int8 roundingMode; 364 flag roundNearestEven; 365 int8 roundIncrement, roundBits; 366 flag isTiny; 367 368 roundingMode = STATUS(float_rounding_mode); 369 roundNearestEven = ( roundingMode == float_round_nearest_even ); 370 switch (roundingMode) { 371 case float_round_nearest_even: 372 case float_round_ties_away: 373 roundIncrement = 0x40; 374 break; 375 case float_round_to_zero: 376 roundIncrement = 0; 377 break; 378 case float_round_up: 379 roundIncrement = zSign ? 0 : 0x7f; 380 break; 381 case float_round_down: 382 roundIncrement = zSign ? 0x7f : 0; 383 break; 384 default: 385 abort(); 386 break; 387 } 388 roundBits = zSig & 0x7F; 389 if ( 0xFD <= (uint16_t) zExp ) { 390 if ( ( 0xFD < zExp ) 391 || ( ( zExp == 0xFD ) 392 && ( (int32_t) ( zSig + roundIncrement ) < 0 ) ) 393 ) { 394 float_raise( float_flag_overflow | float_flag_inexact STATUS_VAR); 395 return packFloat32( zSign, 0xFF, - ( roundIncrement == 0 )); 396 } 397 if ( zExp < 0 ) { 398 if (STATUS(flush_to_zero)) { 399 float_raise(float_flag_output_denormal STATUS_VAR); 400 return packFloat32(zSign, 0, 0); 401 } 402 isTiny = 403 ( STATUS(float_detect_tininess) == float_tininess_before_rounding ) 404 || ( zExp < -1 ) 405 || ( zSig + roundIncrement < 0x80000000 ); 406 shift32RightJamming( zSig, - zExp, &zSig ); 407 zExp = 0; 408 roundBits = zSig & 0x7F; 409 if ( isTiny && roundBits ) float_raise( float_flag_underflow STATUS_VAR); 410 } 411 } 412 if ( roundBits ) STATUS(float_exception_flags) |= float_flag_inexact; 413 zSig = ( zSig + roundIncrement )>>7; 414 zSig &= ~ ( ( ( roundBits ^ 0x40 ) == 0 ) & roundNearestEven ); 415 if ( zSig == 0 ) zExp = 0; 416 return packFloat32( zSign, zExp, zSig ); 417 418 } 419 420 /*---------------------------------------------------------------------------- 421 | Takes an abstract floating-point value having sign `zSign', exponent `zExp', 422 | and significand `zSig', and returns the proper single-precision floating- 423 | point value corresponding to the abstract input. This routine is just like 424 | `roundAndPackFloat32' except that `zSig' does not have to be normalized. 425 | Bit 31 of `zSig' must be zero, and `zExp' must be 1 less than the ``true'' 426 | floating-point exponent. 427 *----------------------------------------------------------------------------*/ 428 429 static float32 430 normalizeRoundAndPackFloat32(flag zSign, int_fast16_t zExp, uint32_t zSig STATUS_PARAM) 431 { 432 int8 shiftCount; 433 434 shiftCount = countLeadingZeros32( zSig ) - 1; 435 return roundAndPackFloat32( zSign, zExp - shiftCount, zSig<<shiftCount STATUS_VAR); 436 437 } 438 439 /*---------------------------------------------------------------------------- 440 | Returns the fraction bits of the double-precision floating-point value `a'. 441 *----------------------------------------------------------------------------*/ 442 443 static inline uint64_t extractFloat64Frac( float64 a ) 444 { 445 446 return float64_val(a) & LIT64( 0x000FFFFFFFFFFFFF ); 447 448 } 449 450 /*---------------------------------------------------------------------------- 451 | Returns the exponent bits of the double-precision floating-point value `a'. 452 *----------------------------------------------------------------------------*/ 453 454 static inline int_fast16_t extractFloat64Exp(float64 a) 455 { 456 457 return ( float64_val(a)>>52 ) & 0x7FF; 458 459 } 460 461 /*---------------------------------------------------------------------------- 462 | Returns the sign bit of the double-precision floating-point value `a'. 463 *----------------------------------------------------------------------------*/ 464 465 static inline flag extractFloat64Sign( float64 a ) 466 { 467 468 return float64_val(a)>>63; 469 470 } 471 472 /*---------------------------------------------------------------------------- 473 | If `a' is denormal and we are in flush-to-zero mode then set the 474 | input-denormal exception and return zero. Otherwise just return the value. 475 *----------------------------------------------------------------------------*/ 476 float64 float64_squash_input_denormal(float64 a STATUS_PARAM) 477 { 478 if (STATUS(flush_inputs_to_zero)) { 479 if (extractFloat64Exp(a) == 0 && extractFloat64Frac(a) != 0) { 480 float_raise(float_flag_input_denormal STATUS_VAR); 481 return make_float64(float64_val(a) & (1ULL << 63)); 482 } 483 } 484 return a; 485 } 486 487 /*---------------------------------------------------------------------------- 488 | Normalizes the subnormal double-precision floating-point value represented 489 | by the denormalized significand `aSig'. The normalized exponent and 490 | significand are stored at the locations pointed to by `zExpPtr' and 491 | `zSigPtr', respectively. 492 *----------------------------------------------------------------------------*/ 493 494 static void 495 normalizeFloat64Subnormal(uint64_t aSig, int_fast16_t *zExpPtr, uint64_t *zSigPtr) 496 { 497 int8 shiftCount; 498 499 shiftCount = countLeadingZeros64( aSig ) - 11; 500 *zSigPtr = aSig<<shiftCount; 501 *zExpPtr = 1 - shiftCount; 502 503 } 504 505 /*---------------------------------------------------------------------------- 506 | Packs the sign `zSign', exponent `zExp', and significand `zSig' into a 507 | double-precision floating-point value, returning the result. After being 508 | shifted into the proper positions, the three fields are simply added 509 | together to form the result. This means that any integer portion of `zSig' 510 | will be added into the exponent. Since a properly normalized significand 511 | will have an integer portion equal to 1, the `zExp' input should be 1 less 512 | than the desired result exponent whenever `zSig' is a complete, normalized 513 | significand. 514 *----------------------------------------------------------------------------*/ 515 516 static inline float64 packFloat64(flag zSign, int_fast16_t zExp, uint64_t zSig) 517 { 518 519 return make_float64( 520 ( ( (uint64_t) zSign )<<63 ) + ( ( (uint64_t) zExp )<<52 ) + zSig); 521 522 } 523 524 /*---------------------------------------------------------------------------- 525 | Takes an abstract floating-point value having sign `zSign', exponent `zExp', 526 | and significand `zSig', and returns the proper double-precision floating- 527 | point value corresponding to the abstract input. Ordinarily, the abstract 528 | value is simply rounded and packed into the double-precision format, with 529 | the inexact exception raised if the abstract input cannot be represented 530 | exactly. However, if the abstract value is too large, the overflow and 531 | inexact exceptions are raised and an infinity or maximal finite value is 532 | returned. If the abstract value is too small, the input value is rounded 533 | to a subnormal number, and the underflow and inexact exceptions are raised 534 | if the abstract input cannot be represented exactly as a subnormal double- 535 | precision floating-point number. 536 | The input significand `zSig' has its binary point between bits 62 537 | and 61, which is 10 bits to the left of the usual location. This shifted 538 | significand must be normalized or smaller. If `zSig' is not normalized, 539 | `zExp' must be 0; in that case, the result returned is a subnormal number, 540 | and it must not require rounding. In the usual case that `zSig' is 541 | normalized, `zExp' must be 1 less than the ``true'' floating-point exponent. 542 | The handling of underflow and overflow follows the IEC/IEEE Standard for 543 | Binary Floating-Point Arithmetic. 544 *----------------------------------------------------------------------------*/ 545 546 static float64 roundAndPackFloat64(flag zSign, int_fast16_t zExp, uint64_t zSig STATUS_PARAM) 547 { 548 int8 roundingMode; 549 flag roundNearestEven; 550 int_fast16_t roundIncrement, roundBits; 551 flag isTiny; 552 553 roundingMode = STATUS(float_rounding_mode); 554 roundNearestEven = ( roundingMode == float_round_nearest_even ); 555 switch (roundingMode) { 556 case float_round_nearest_even: 557 case float_round_ties_away: 558 roundIncrement = 0x200; 559 break; 560 case float_round_to_zero: 561 roundIncrement = 0; 562 break; 563 case float_round_up: 564 roundIncrement = zSign ? 0 : 0x3ff; 565 break; 566 case float_round_down: 567 roundIncrement = zSign ? 0x3ff : 0; 568 break; 569 default: 570 abort(); 571 } 572 roundBits = zSig & 0x3FF; 573 if ( 0x7FD <= (uint16_t) zExp ) { 574 if ( ( 0x7FD < zExp ) 575 || ( ( zExp == 0x7FD ) 576 && ( (int64_t) ( zSig + roundIncrement ) < 0 ) ) 577 ) { 578 float_raise( float_flag_overflow | float_flag_inexact STATUS_VAR); 579 return packFloat64( zSign, 0x7FF, - ( roundIncrement == 0 )); 580 } 581 if ( zExp < 0 ) { 582 if (STATUS(flush_to_zero)) { 583 float_raise(float_flag_output_denormal STATUS_VAR); 584 return packFloat64(zSign, 0, 0); 585 } 586 isTiny = 587 ( STATUS(float_detect_tininess) == float_tininess_before_rounding ) 588 || ( zExp < -1 ) 589 || ( zSig + roundIncrement < LIT64( 0x8000000000000000 ) ); 590 shift64RightJamming( zSig, - zExp, &zSig ); 591 zExp = 0; 592 roundBits = zSig & 0x3FF; 593 if ( isTiny && roundBits ) float_raise( float_flag_underflow STATUS_VAR); 594 } 595 } 596 if ( roundBits ) STATUS(float_exception_flags) |= float_flag_inexact; 597 zSig = ( zSig + roundIncrement )>>10; 598 zSig &= ~ ( ( ( roundBits ^ 0x200 ) == 0 ) & roundNearestEven ); 599 if ( zSig == 0 ) zExp = 0; 600 return packFloat64( zSign, zExp, zSig ); 601 602 } 603 604 /*---------------------------------------------------------------------------- 605 | Takes an abstract floating-point value having sign `zSign', exponent `zExp', 606 | and significand `zSig', and returns the proper double-precision floating- 607 | point value corresponding to the abstract input. This routine is just like 608 | `roundAndPackFloat64' except that `zSig' does not have to be normalized. 609 | Bit 63 of `zSig' must be zero, and `zExp' must be 1 less than the ``true'' 610 | floating-point exponent. 611 *----------------------------------------------------------------------------*/ 612 613 static float64 614 normalizeRoundAndPackFloat64(flag zSign, int_fast16_t zExp, uint64_t zSig STATUS_PARAM) 615 { 616 int8 shiftCount; 617 618 shiftCount = countLeadingZeros64( zSig ) - 1; 619 return roundAndPackFloat64( zSign, zExp - shiftCount, zSig<<shiftCount STATUS_VAR); 620 621 } 622 623 /*---------------------------------------------------------------------------- 624 | Returns the fraction bits of the extended double-precision floating-point 625 | value `a'. 626 *----------------------------------------------------------------------------*/ 627 628 static inline uint64_t extractFloatx80Frac( floatx80 a ) 629 { 630 631 return a.low; 632 633 } 634 635 /*---------------------------------------------------------------------------- 636 | Returns the exponent bits of the extended double-precision floating-point 637 | value `a'. 638 *----------------------------------------------------------------------------*/ 639 640 static inline int32 extractFloatx80Exp( floatx80 a ) 641 { 642 643 return a.high & 0x7FFF; 644 645 } 646 647 /*---------------------------------------------------------------------------- 648 | Returns the sign bit of the extended double-precision floating-point value 649 | `a'. 650 *----------------------------------------------------------------------------*/ 651 652 static inline flag extractFloatx80Sign( floatx80 a ) 653 { 654 655 return a.high>>15; 656 657 } 658 659 /*---------------------------------------------------------------------------- 660 | Normalizes the subnormal extended double-precision floating-point value 661 | represented by the denormalized significand `aSig'. The normalized exponent 662 | and significand are stored at the locations pointed to by `zExpPtr' and 663 | `zSigPtr', respectively. 664 *----------------------------------------------------------------------------*/ 665 666 static void 667 normalizeFloatx80Subnormal( uint64_t aSig, int32 *zExpPtr, uint64_t *zSigPtr ) 668 { 669 int8 shiftCount; 670 671 shiftCount = countLeadingZeros64( aSig ); 672 *zSigPtr = aSig<<shiftCount; 673 *zExpPtr = 1 - shiftCount; 674 675 } 676 677 /*---------------------------------------------------------------------------- 678 | Packs the sign `zSign', exponent `zExp', and significand `zSig' into an 679 | extended double-precision floating-point value, returning the result. 680 *----------------------------------------------------------------------------*/ 681 682 static inline floatx80 packFloatx80( flag zSign, int32 zExp, uint64_t zSig ) 683 { 684 floatx80 z; 685 686 z.low = zSig; 687 z.high = ( ( (uint16_t) zSign )<<15 ) + zExp; 688 return z; 689 690 } 691 692 /*---------------------------------------------------------------------------- 693 | Takes an abstract floating-point value having sign `zSign', exponent `zExp', 694 | and extended significand formed by the concatenation of `zSig0' and `zSig1', 695 | and returns the proper extended double-precision floating-point value 696 | corresponding to the abstract input. Ordinarily, the abstract value is 697 | rounded and packed into the extended double-precision format, with the 698 | inexact exception raised if the abstract input cannot be represented 699 | exactly. However, if the abstract value is too large, the overflow and 700 | inexact exceptions are raised and an infinity or maximal finite value is 701 | returned. If the abstract value is too small, the input value is rounded to 702 | a subnormal number, and the underflow and inexact exceptions are raised if 703 | the abstract input cannot be represented exactly as a subnormal extended 704 | double-precision floating-point number. 705 | If `roundingPrecision' is 32 or 64, the result is rounded to the same 706 | number of bits as single or double precision, respectively. Otherwise, the 707 | result is rounded to the full precision of the extended double-precision 708 | format. 709 | The input significand must be normalized or smaller. If the input 710 | significand is not normalized, `zExp' must be 0; in that case, the result 711 | returned is a subnormal number, and it must not require rounding. The 712 | handling of underflow and overflow follows the IEC/IEEE Standard for Binary 713 | Floating-Point Arithmetic. 714 *----------------------------------------------------------------------------*/ 715 716 static floatx80 717 roundAndPackFloatx80( 718 int8 roundingPrecision, flag zSign, int32 zExp, uint64_t zSig0, uint64_t zSig1 719 STATUS_PARAM) 720 { 721 int8 roundingMode; 722 flag roundNearestEven, increment, isTiny; 723 int64 roundIncrement, roundMask, roundBits; 724 725 roundingMode = STATUS(float_rounding_mode); 726 roundNearestEven = ( roundingMode == float_round_nearest_even ); 727 if ( roundingPrecision == 80 ) goto precision80; 728 if ( roundingPrecision == 64 ) { 729 roundIncrement = LIT64( 0x0000000000000400 ); 730 roundMask = LIT64( 0x00000000000007FF ); 731 } 732 else if ( roundingPrecision == 32 ) { 733 roundIncrement = LIT64( 0x0000008000000000 ); 734 roundMask = LIT64( 0x000000FFFFFFFFFF ); 735 } 736 else { 737 goto precision80; 738 } 739 zSig0 |= ( zSig1 != 0 ); 740 switch (roundingMode) { 741 case float_round_nearest_even: 742 case float_round_ties_away: 743 break; 744 case float_round_to_zero: 745 roundIncrement = 0; 746 break; 747 case float_round_up: 748 roundIncrement = zSign ? 0 : roundMask; 749 break; 750 case float_round_down: 751 roundIncrement = zSign ? roundMask : 0; 752 break; 753 default: 754 abort(); 755 } 756 roundBits = zSig0 & roundMask; 757 if ( 0x7FFD <= (uint32_t) ( zExp - 1 ) ) { 758 if ( ( 0x7FFE < zExp ) 759 || ( ( zExp == 0x7FFE ) && ( zSig0 + roundIncrement < zSig0 ) ) 760 ) { 761 goto overflow; 762 } 763 if ( zExp <= 0 ) { 764 if (STATUS(flush_to_zero)) { 765 float_raise(float_flag_output_denormal STATUS_VAR); 766 return packFloatx80(zSign, 0, 0); 767 } 768 isTiny = 769 ( STATUS(float_detect_tininess) == float_tininess_before_rounding ) 770 || ( zExp < 0 ) 771 || ( zSig0 <= zSig0 + roundIncrement ); 772 shift64RightJamming( zSig0, 1 - zExp, &zSig0 ); 773 zExp = 0; 774 roundBits = zSig0 & roundMask; 775 if ( isTiny && roundBits ) float_raise( float_flag_underflow STATUS_VAR); 776 if ( roundBits ) STATUS(float_exception_flags) |= float_flag_inexact; 777 zSig0 += roundIncrement; 778 if ( (int64_t) zSig0 < 0 ) zExp = 1; 779 roundIncrement = roundMask + 1; 780 if ( roundNearestEven && ( roundBits<<1 == roundIncrement ) ) { 781 roundMask |= roundIncrement; 782 } 783 zSig0 &= ~ roundMask; 784 return packFloatx80( zSign, zExp, zSig0 ); 785 } 786 } 787 if ( roundBits ) STATUS(float_exception_flags) |= float_flag_inexact; 788 zSig0 += roundIncrement; 789 if ( zSig0 < roundIncrement ) { 790 ++zExp; 791 zSig0 = LIT64( 0x8000000000000000 ); 792 } 793 roundIncrement = roundMask + 1; 794 if ( roundNearestEven && ( roundBits<<1 == roundIncrement ) ) { 795 roundMask |= roundIncrement; 796 } 797 zSig0 &= ~ roundMask; 798 if ( zSig0 == 0 ) zExp = 0; 799 return packFloatx80( zSign, zExp, zSig0 ); 800 precision80: 801 switch (roundingMode) { 802 case float_round_nearest_even: 803 case float_round_ties_away: 804 increment = ((int64_t)zSig1 < 0); 805 break; 806 case float_round_to_zero: 807 increment = 0; 808 break; 809 case float_round_up: 810 increment = !zSign && zSig1; 811 break; 812 case float_round_down: 813 increment = zSign && zSig1; 814 break; 815 default: 816 abort(); 817 } 818 if ( 0x7FFD <= (uint32_t) ( zExp - 1 ) ) { 819 if ( ( 0x7FFE < zExp ) 820 || ( ( zExp == 0x7FFE ) 821 && ( zSig0 == LIT64( 0xFFFFFFFFFFFFFFFF ) ) 822 && increment 823 ) 824 ) { 825 roundMask = 0; 826 overflow: 827 float_raise( float_flag_overflow | float_flag_inexact STATUS_VAR); 828 if ( ( roundingMode == float_round_to_zero ) 829 || ( zSign && ( roundingMode == float_round_up ) ) 830 || ( ! zSign && ( roundingMode == float_round_down ) ) 831 ) { 832 return packFloatx80( zSign, 0x7FFE, ~ roundMask ); 833 } 834 return packFloatx80( zSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 835 } 836 if ( zExp <= 0 ) { 837 isTiny = 838 ( STATUS(float_detect_tininess) == float_tininess_before_rounding ) 839 || ( zExp < 0 ) 840 || ! increment 841 || ( zSig0 < LIT64( 0xFFFFFFFFFFFFFFFF ) ); 842 shift64ExtraRightJamming( zSig0, zSig1, 1 - zExp, &zSig0, &zSig1 ); 843 zExp = 0; 844 if ( isTiny && zSig1 ) float_raise( float_flag_underflow STATUS_VAR); 845 if ( zSig1 ) STATUS(float_exception_flags) |= float_flag_inexact; 846 switch (roundingMode) { 847 case float_round_nearest_even: 848 case float_round_ties_away: 849 increment = ((int64_t)zSig1 < 0); 850 break; 851 case float_round_to_zero: 852 increment = 0; 853 break; 854 case float_round_up: 855 increment = !zSign && zSig1; 856 break; 857 case float_round_down: 858 increment = zSign && zSig1; 859 break; 860 default: 861 abort(); 862 } 863 if ( increment ) { 864 ++zSig0; 865 zSig0 &= 866 ~ ( ( (uint64_t) ( zSig1<<1 ) == 0 ) & roundNearestEven ); 867 if ( (int64_t) zSig0 < 0 ) zExp = 1; 868 } 869 return packFloatx80( zSign, zExp, zSig0 ); 870 } 871 } 872 if ( zSig1 ) STATUS(float_exception_flags) |= float_flag_inexact; 873 if ( increment ) { 874 ++zSig0; 875 if ( zSig0 == 0 ) { 876 ++zExp; 877 zSig0 = LIT64( 0x8000000000000000 ); 878 } 879 else { 880 zSig0 &= ~ ( ( (uint64_t) ( zSig1<<1 ) == 0 ) & roundNearestEven ); 881 } 882 } 883 else { 884 if ( zSig0 == 0 ) zExp = 0; 885 } 886 return packFloatx80( zSign, zExp, zSig0 ); 887 888 } 889 890 /*---------------------------------------------------------------------------- 891 | Takes an abstract floating-point value having sign `zSign', exponent 892 | `zExp', and significand formed by the concatenation of `zSig0' and `zSig1', 893 | and returns the proper extended double-precision floating-point value 894 | corresponding to the abstract input. This routine is just like 895 | `roundAndPackFloatx80' except that the input significand does not have to be 896 | normalized. 897 *----------------------------------------------------------------------------*/ 898 899 static floatx80 900 normalizeRoundAndPackFloatx80( 901 int8 roundingPrecision, flag zSign, int32 zExp, uint64_t zSig0, uint64_t zSig1 902 STATUS_PARAM) 903 { 904 int8 shiftCount; 905 906 if ( zSig0 == 0 ) { 907 zSig0 = zSig1; 908 zSig1 = 0; 909 zExp -= 64; 910 } 911 shiftCount = countLeadingZeros64( zSig0 ); 912 shortShift128Left( zSig0, zSig1, shiftCount, &zSig0, &zSig1 ); 913 zExp -= shiftCount; 914 return 915 roundAndPackFloatx80( roundingPrecision, zSign, zExp, zSig0, zSig1 STATUS_VAR); 916 917 } 918 919 /*---------------------------------------------------------------------------- 920 | Returns the least-significant 64 fraction bits of the quadruple-precision 921 | floating-point value `a'. 922 *----------------------------------------------------------------------------*/ 923 924 static inline uint64_t extractFloat128Frac1( float128 a ) 925 { 926 927 return a.low; 928 929 } 930 931 /*---------------------------------------------------------------------------- 932 | Returns the most-significant 48 fraction bits of the quadruple-precision 933 | floating-point value `a'. 934 *----------------------------------------------------------------------------*/ 935 936 static inline uint64_t extractFloat128Frac0( float128 a ) 937 { 938 939 return a.high & LIT64( 0x0000FFFFFFFFFFFF ); 940 941 } 942 943 /*---------------------------------------------------------------------------- 944 | Returns the exponent bits of the quadruple-precision floating-point value 945 | `a'. 946 *----------------------------------------------------------------------------*/ 947 948 static inline int32 extractFloat128Exp( float128 a ) 949 { 950 951 return ( a.high>>48 ) & 0x7FFF; 952 953 } 954 955 /*---------------------------------------------------------------------------- 956 | Returns the sign bit of the quadruple-precision floating-point value `a'. 957 *----------------------------------------------------------------------------*/ 958 959 static inline flag extractFloat128Sign( float128 a ) 960 { 961 962 return a.high>>63; 963 964 } 965 966 /*---------------------------------------------------------------------------- 967 | Normalizes the subnormal quadruple-precision floating-point value 968 | represented by the denormalized significand formed by the concatenation of 969 | `aSig0' and `aSig1'. The normalized exponent is stored at the location 970 | pointed to by `zExpPtr'. The most significant 49 bits of the normalized 971 | significand are stored at the location pointed to by `zSig0Ptr', and the 972 | least significant 64 bits of the normalized significand are stored at the 973 | location pointed to by `zSig1Ptr'. 974 *----------------------------------------------------------------------------*/ 975 976 static void 977 normalizeFloat128Subnormal( 978 uint64_t aSig0, 979 uint64_t aSig1, 980 int32 *zExpPtr, 981 uint64_t *zSig0Ptr, 982 uint64_t *zSig1Ptr 983 ) 984 { 985 int8 shiftCount; 986 987 if ( aSig0 == 0 ) { 988 shiftCount = countLeadingZeros64( aSig1 ) - 15; 989 if ( shiftCount < 0 ) { 990 *zSig0Ptr = aSig1>>( - shiftCount ); 991 *zSig1Ptr = aSig1<<( shiftCount & 63 ); 992 } 993 else { 994 *zSig0Ptr = aSig1<<shiftCount; 995 *zSig1Ptr = 0; 996 } 997 *zExpPtr = - shiftCount - 63; 998 } 999 else { 1000 shiftCount = countLeadingZeros64( aSig0 ) - 15; 1001 shortShift128Left( aSig0, aSig1, shiftCount, zSig0Ptr, zSig1Ptr ); 1002 *zExpPtr = 1 - shiftCount; 1003 } 1004 1005 } 1006 1007 /*---------------------------------------------------------------------------- 1008 | Packs the sign `zSign', the exponent `zExp', and the significand formed 1009 | by the concatenation of `zSig0' and `zSig1' into a quadruple-precision 1010 | floating-point value, returning the result. After being shifted into the 1011 | proper positions, the three fields `zSign', `zExp', and `zSig0' are simply 1012 | added together to form the most significant 32 bits of the result. This 1013 | means that any integer portion of `zSig0' will be added into the exponent. 1014 | Since a properly normalized significand will have an integer portion equal 1015 | to 1, the `zExp' input should be 1 less than the desired result exponent 1016 | whenever `zSig0' and `zSig1' concatenated form a complete, normalized 1017 | significand. 1018 *----------------------------------------------------------------------------*/ 1019 1020 static inline float128 1021 packFloat128( flag zSign, int32 zExp, uint64_t zSig0, uint64_t zSig1 ) 1022 { 1023 float128 z; 1024 1025 z.low = zSig1; 1026 z.high = ( ( (uint64_t) zSign )<<63 ) + ( ( (uint64_t) zExp )<<48 ) + zSig0; 1027 return z; 1028 1029 } 1030 1031 /*---------------------------------------------------------------------------- 1032 | Takes an abstract floating-point value having sign `zSign', exponent `zExp', 1033 | and extended significand formed by the concatenation of `zSig0', `zSig1', 1034 | and `zSig2', and returns the proper quadruple-precision floating-point value 1035 | corresponding to the abstract input. Ordinarily, the abstract value is 1036 | simply rounded and packed into the quadruple-precision format, with the 1037 | inexact exception raised if the abstract input cannot be represented 1038 | exactly. However, if the abstract value is too large, the overflow and 1039 | inexact exceptions are raised and an infinity or maximal finite value is 1040 | returned. If the abstract value is too small, the input value is rounded to 1041 | a subnormal number, and the underflow and inexact exceptions are raised if 1042 | the abstract input cannot be represented exactly as a subnormal quadruple- 1043 | precision floating-point number. 1044 | The input significand must be normalized or smaller. If the input 1045 | significand is not normalized, `zExp' must be 0; in that case, the result 1046 | returned is a subnormal number, and it must not require rounding. In the 1047 | usual case that the input significand is normalized, `zExp' must be 1 less 1048 | than the ``true'' floating-point exponent. The handling of underflow and 1049 | overflow follows the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 1050 *----------------------------------------------------------------------------*/ 1051 1052 static float128 1053 roundAndPackFloat128( 1054 flag zSign, int32 zExp, uint64_t zSig0, uint64_t zSig1, uint64_t zSig2 STATUS_PARAM) 1055 { 1056 int8 roundingMode; 1057 flag roundNearestEven, increment, isTiny; 1058 1059 roundingMode = STATUS(float_rounding_mode); 1060 roundNearestEven = ( roundingMode == float_round_nearest_even ); 1061 switch (roundingMode) { 1062 case float_round_nearest_even: 1063 case float_round_ties_away: 1064 increment = ((int64_t)zSig2 < 0); 1065 break; 1066 case float_round_to_zero: 1067 increment = 0; 1068 break; 1069 case float_round_up: 1070 increment = !zSign && zSig2; 1071 break; 1072 case float_round_down: 1073 increment = zSign && zSig2; 1074 break; 1075 default: 1076 abort(); 1077 } 1078 if ( 0x7FFD <= (uint32_t) zExp ) { 1079 if ( ( 0x7FFD < zExp ) 1080 || ( ( zExp == 0x7FFD ) 1081 && eq128( 1082 LIT64( 0x0001FFFFFFFFFFFF ), 1083 LIT64( 0xFFFFFFFFFFFFFFFF ), 1084 zSig0, 1085 zSig1 1086 ) 1087 && increment 1088 ) 1089 ) { 1090 float_raise( float_flag_overflow | float_flag_inexact STATUS_VAR); 1091 if ( ( roundingMode == float_round_to_zero ) 1092 || ( zSign && ( roundingMode == float_round_up ) ) 1093 || ( ! zSign && ( roundingMode == float_round_down ) ) 1094 ) { 1095 return 1096 packFloat128( 1097 zSign, 1098 0x7FFE, 1099 LIT64( 0x0000FFFFFFFFFFFF ), 1100 LIT64( 0xFFFFFFFFFFFFFFFF ) 1101 ); 1102 } 1103 return packFloat128( zSign, 0x7FFF, 0, 0 ); 1104 } 1105 if ( zExp < 0 ) { 1106 if (STATUS(flush_to_zero)) { 1107 float_raise(float_flag_output_denormal STATUS_VAR); 1108 return packFloat128(zSign, 0, 0, 0); 1109 } 1110 isTiny = 1111 ( STATUS(float_detect_tininess) == float_tininess_before_rounding ) 1112 || ( zExp < -1 ) 1113 || ! increment 1114 || lt128( 1115 zSig0, 1116 zSig1, 1117 LIT64( 0x0001FFFFFFFFFFFF ), 1118 LIT64( 0xFFFFFFFFFFFFFFFF ) 1119 ); 1120 shift128ExtraRightJamming( 1121 zSig0, zSig1, zSig2, - zExp, &zSig0, &zSig1, &zSig2 ); 1122 zExp = 0; 1123 if ( isTiny && zSig2 ) float_raise( float_flag_underflow STATUS_VAR); 1124 switch (roundingMode) { 1125 case float_round_nearest_even: 1126 case float_round_ties_away: 1127 increment = ((int64_t)zSig2 < 0); 1128 break; 1129 case float_round_to_zero: 1130 increment = 0; 1131 break; 1132 case float_round_up: 1133 increment = !zSign && zSig2; 1134 break; 1135 case float_round_down: 1136 increment = zSign && zSig2; 1137 break; 1138 default: 1139 abort(); 1140 } 1141 } 1142 } 1143 if ( zSig2 ) STATUS(float_exception_flags) |= float_flag_inexact; 1144 if ( increment ) { 1145 add128( zSig0, zSig1, 0, 1, &zSig0, &zSig1 ); 1146 zSig1 &= ~ ( ( zSig2 + zSig2 == 0 ) & roundNearestEven ); 1147 } 1148 else { 1149 if ( ( zSig0 | zSig1 ) == 0 ) zExp = 0; 1150 } 1151 return packFloat128( zSign, zExp, zSig0, zSig1 ); 1152 1153 } 1154 1155 /*---------------------------------------------------------------------------- 1156 | Takes an abstract floating-point value having sign `zSign', exponent `zExp', 1157 | and significand formed by the concatenation of `zSig0' and `zSig1', and 1158 | returns the proper quadruple-precision floating-point value corresponding 1159 | to the abstract input. This routine is just like `roundAndPackFloat128' 1160 | except that the input significand has fewer bits and does not have to be 1161 | normalized. In all cases, `zExp' must be 1 less than the ``true'' floating- 1162 | point exponent. 1163 *----------------------------------------------------------------------------*/ 1164 1165 static float128 1166 normalizeRoundAndPackFloat128( 1167 flag zSign, int32 zExp, uint64_t zSig0, uint64_t zSig1 STATUS_PARAM) 1168 { 1169 int8 shiftCount; 1170 uint64_t zSig2; 1171 1172 if ( zSig0 == 0 ) { 1173 zSig0 = zSig1; 1174 zSig1 = 0; 1175 zExp -= 64; 1176 } 1177 shiftCount = countLeadingZeros64( zSig0 ) - 15; 1178 if ( 0 <= shiftCount ) { 1179 zSig2 = 0; 1180 shortShift128Left( zSig0, zSig1, shiftCount, &zSig0, &zSig1 ); 1181 } 1182 else { 1183 shift128ExtraRightJamming( 1184 zSig0, zSig1, 0, - shiftCount, &zSig0, &zSig1, &zSig2 ); 1185 } 1186 zExp -= shiftCount; 1187 return roundAndPackFloat128( zSign, zExp, zSig0, zSig1, zSig2 STATUS_VAR); 1188 1189 } 1190 1191 /*---------------------------------------------------------------------------- 1192 | Returns the result of converting the 32-bit two's complement integer `a' 1193 | to the single-precision floating-point format. The conversion is performed 1194 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 1195 *----------------------------------------------------------------------------*/ 1196 1197 float32 int32_to_float32(int32_t a STATUS_PARAM) 1198 { 1199 flag zSign; 1200 1201 if ( a == 0 ) return float32_zero; 1202 if ( a == (int32_t) 0x80000000 ) return packFloat32( 1, 0x9E, 0 ); 1203 zSign = ( a < 0 ); 1204 return normalizeRoundAndPackFloat32( zSign, 0x9C, zSign ? - a : a STATUS_VAR ); 1205 1206 } 1207 1208 /*---------------------------------------------------------------------------- 1209 | Returns the result of converting the 32-bit two's complement integer `a' 1210 | to the double-precision floating-point format. The conversion is performed 1211 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 1212 *----------------------------------------------------------------------------*/ 1213 1214 float64 int32_to_float64(int32_t a STATUS_PARAM) 1215 { 1216 flag zSign; 1217 uint32 absA; 1218 int8 shiftCount; 1219 uint64_t zSig; 1220 1221 if ( a == 0 ) return float64_zero; 1222 zSign = ( a < 0 ); 1223 absA = zSign ? - a : a; 1224 shiftCount = countLeadingZeros32( absA ) + 21; 1225 zSig = absA; 1226 return packFloat64( zSign, 0x432 - shiftCount, zSig<<shiftCount ); 1227 1228 } 1229 1230 /*---------------------------------------------------------------------------- 1231 | Returns the result of converting the 32-bit two's complement integer `a' 1232 | to the extended double-precision floating-point format. The conversion 1233 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 1234 | Arithmetic. 1235 *----------------------------------------------------------------------------*/ 1236 1237 floatx80 int32_to_floatx80(int32_t a STATUS_PARAM) 1238 { 1239 flag zSign; 1240 uint32 absA; 1241 int8 shiftCount; 1242 uint64_t zSig; 1243 1244 if ( a == 0 ) return packFloatx80( 0, 0, 0 ); 1245 zSign = ( a < 0 ); 1246 absA = zSign ? - a : a; 1247 shiftCount = countLeadingZeros32( absA ) + 32; 1248 zSig = absA; 1249 return packFloatx80( zSign, 0x403E - shiftCount, zSig<<shiftCount ); 1250 1251 } 1252 1253 /*---------------------------------------------------------------------------- 1254 | Returns the result of converting the 32-bit two's complement integer `a' to 1255 | the quadruple-precision floating-point format. The conversion is performed 1256 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 1257 *----------------------------------------------------------------------------*/ 1258 1259 float128 int32_to_float128(int32_t a STATUS_PARAM) 1260 { 1261 flag zSign; 1262 uint32 absA; 1263 int8 shiftCount; 1264 uint64_t zSig0; 1265 1266 if ( a == 0 ) return packFloat128( 0, 0, 0, 0 ); 1267 zSign = ( a < 0 ); 1268 absA = zSign ? - a : a; 1269 shiftCount = countLeadingZeros32( absA ) + 17; 1270 zSig0 = absA; 1271 return packFloat128( zSign, 0x402E - shiftCount, zSig0<<shiftCount, 0 ); 1272 1273 } 1274 1275 /*---------------------------------------------------------------------------- 1276 | Returns the result of converting the 64-bit two's complement integer `a' 1277 | to the single-precision floating-point format. The conversion is performed 1278 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 1279 *----------------------------------------------------------------------------*/ 1280 1281 float32 int64_to_float32(int64_t a STATUS_PARAM) 1282 { 1283 flag zSign; 1284 uint64 absA; 1285 int8 shiftCount; 1286 1287 if ( a == 0 ) return float32_zero; 1288 zSign = ( a < 0 ); 1289 absA = zSign ? - a : a; 1290 shiftCount = countLeadingZeros64( absA ) - 40; 1291 if ( 0 <= shiftCount ) { 1292 return packFloat32( zSign, 0x95 - shiftCount, absA<<shiftCount ); 1293 } 1294 else { 1295 shiftCount += 7; 1296 if ( shiftCount < 0 ) { 1297 shift64RightJamming( absA, - shiftCount, &absA ); 1298 } 1299 else { 1300 absA <<= shiftCount; 1301 } 1302 return roundAndPackFloat32( zSign, 0x9C - shiftCount, absA STATUS_VAR ); 1303 } 1304 1305 } 1306 1307 float32 uint64_to_float32(uint64_t a STATUS_PARAM) 1308 { 1309 int8 shiftCount; 1310 1311 if ( a == 0 ) return float32_zero; 1312 shiftCount = countLeadingZeros64( a ) - 40; 1313 if ( 0 <= shiftCount ) { 1314 return packFloat32(0, 0x95 - shiftCount, a<<shiftCount); 1315 } 1316 else { 1317 shiftCount += 7; 1318 if ( shiftCount < 0 ) { 1319 shift64RightJamming( a, - shiftCount, &a ); 1320 } 1321 else { 1322 a <<= shiftCount; 1323 } 1324 return roundAndPackFloat32(0, 0x9C - shiftCount, a STATUS_VAR); 1325 } 1326 } 1327 1328 /*---------------------------------------------------------------------------- 1329 | Returns the result of converting the 64-bit two's complement integer `a' 1330 | to the double-precision floating-point format. The conversion is performed 1331 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 1332 *----------------------------------------------------------------------------*/ 1333 1334 float64 int64_to_float64(int64_t a STATUS_PARAM) 1335 { 1336 flag zSign; 1337 1338 if ( a == 0 ) return float64_zero; 1339 if ( a == (int64_t) LIT64( 0x8000000000000000 ) ) { 1340 return packFloat64( 1, 0x43E, 0 ); 1341 } 1342 zSign = ( a < 0 ); 1343 return normalizeRoundAndPackFloat64( zSign, 0x43C, zSign ? - a : a STATUS_VAR ); 1344 1345 } 1346 1347 float64 uint64_to_float64(uint64_t a STATUS_PARAM) 1348 { 1349 int exp = 0x43C; 1350 1351 if (a == 0) { 1352 return float64_zero; 1353 } 1354 if ((int64_t)a < 0) { 1355 shift64RightJamming(a, 1, &a); 1356 exp += 1; 1357 } 1358 return normalizeRoundAndPackFloat64(0, exp, a STATUS_VAR); 1359 } 1360 1361 /*---------------------------------------------------------------------------- 1362 | Returns the result of converting the 64-bit two's complement integer `a' 1363 | to the extended double-precision floating-point format. The conversion 1364 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 1365 | Arithmetic. 1366 *----------------------------------------------------------------------------*/ 1367 1368 floatx80 int64_to_floatx80(int64_t a STATUS_PARAM) 1369 { 1370 flag zSign; 1371 uint64 absA; 1372 int8 shiftCount; 1373 1374 if ( a == 0 ) return packFloatx80( 0, 0, 0 ); 1375 zSign = ( a < 0 ); 1376 absA = zSign ? - a : a; 1377 shiftCount = countLeadingZeros64( absA ); 1378 return packFloatx80( zSign, 0x403E - shiftCount, absA<<shiftCount ); 1379 1380 } 1381 1382 /*---------------------------------------------------------------------------- 1383 | Returns the result of converting the 64-bit two's complement integer `a' to 1384 | the quadruple-precision floating-point format. The conversion is performed 1385 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 1386 *----------------------------------------------------------------------------*/ 1387 1388 float128 int64_to_float128(int64_t a STATUS_PARAM) 1389 { 1390 flag zSign; 1391 uint64 absA; 1392 int8 shiftCount; 1393 int32 zExp; 1394 uint64_t zSig0, zSig1; 1395 1396 if ( a == 0 ) return packFloat128( 0, 0, 0, 0 ); 1397 zSign = ( a < 0 ); 1398 absA = zSign ? - a : a; 1399 shiftCount = countLeadingZeros64( absA ) + 49; 1400 zExp = 0x406E - shiftCount; 1401 if ( 64 <= shiftCount ) { 1402 zSig1 = 0; 1403 zSig0 = absA; 1404 shiftCount -= 64; 1405 } 1406 else { 1407 zSig1 = absA; 1408 zSig0 = 0; 1409 } 1410 shortShift128Left( zSig0, zSig1, shiftCount, &zSig0, &zSig1 ); 1411 return packFloat128( zSign, zExp, zSig0, zSig1 ); 1412 1413 } 1414 1415 float128 uint64_to_float128(uint64_t a STATUS_PARAM) 1416 { 1417 if (a == 0) { 1418 return float128_zero; 1419 } 1420 return normalizeRoundAndPackFloat128(0, 0x406E, a, 0 STATUS_VAR); 1421 } 1422 1423 /*---------------------------------------------------------------------------- 1424 | Returns the result of converting the single-precision floating-point value 1425 | `a' to the 32-bit two's complement integer format. The conversion is 1426 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1427 | Arithmetic---which means in particular that the conversion is rounded 1428 | according to the current rounding mode. If `a' is a NaN, the largest 1429 | positive integer is returned. Otherwise, if the conversion overflows, the 1430 | largest integer with the same sign as `a' is returned. 1431 *----------------------------------------------------------------------------*/ 1432 1433 int32 float32_to_int32( float32 a STATUS_PARAM ) 1434 { 1435 flag aSign; 1436 int_fast16_t aExp, shiftCount; 1437 uint32_t aSig; 1438 uint64_t aSig64; 1439 1440 a = float32_squash_input_denormal(a STATUS_VAR); 1441 aSig = extractFloat32Frac( a ); 1442 aExp = extractFloat32Exp( a ); 1443 aSign = extractFloat32Sign( a ); 1444 if ( ( aExp == 0xFF ) && aSig ) aSign = 0; 1445 if ( aExp ) aSig |= 0x00800000; 1446 shiftCount = 0xAF - aExp; 1447 aSig64 = aSig; 1448 aSig64 <<= 32; 1449 if ( 0 < shiftCount ) shift64RightJamming( aSig64, shiftCount, &aSig64 ); 1450 return roundAndPackInt32( aSign, aSig64 STATUS_VAR ); 1451 1452 } 1453 1454 /*---------------------------------------------------------------------------- 1455 | Returns the result of converting the single-precision floating-point value 1456 | `a' to the 32-bit two's complement integer format. The conversion is 1457 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1458 | Arithmetic, except that the conversion is always rounded toward zero. 1459 | If `a' is a NaN, the largest positive integer is returned. Otherwise, if 1460 | the conversion overflows, the largest integer with the same sign as `a' is 1461 | returned. 1462 *----------------------------------------------------------------------------*/ 1463 1464 int32 float32_to_int32_round_to_zero( float32 a STATUS_PARAM ) 1465 { 1466 flag aSign; 1467 int_fast16_t aExp, shiftCount; 1468 uint32_t aSig; 1469 int32_t z; 1470 a = float32_squash_input_denormal(a STATUS_VAR); 1471 1472 aSig = extractFloat32Frac( a ); 1473 aExp = extractFloat32Exp( a ); 1474 aSign = extractFloat32Sign( a ); 1475 shiftCount = aExp - 0x9E; 1476 if ( 0 <= shiftCount ) { 1477 if ( float32_val(a) != 0xCF000000 ) { 1478 float_raise( float_flag_invalid STATUS_VAR); 1479 if ( ! aSign || ( ( aExp == 0xFF ) && aSig ) ) return 0x7FFFFFFF; 1480 } 1481 return (int32_t) 0x80000000; 1482 } 1483 else if ( aExp <= 0x7E ) { 1484 if ( aExp | aSig ) STATUS(float_exception_flags) |= float_flag_inexact; 1485 return 0; 1486 } 1487 aSig = ( aSig | 0x00800000 )<<8; 1488 z = aSig>>( - shiftCount ); 1489 if ( (uint32_t) ( aSig<<( shiftCount & 31 ) ) ) { 1490 STATUS(float_exception_flags) |= float_flag_inexact; 1491 } 1492 if ( aSign ) z = - z; 1493 return z; 1494 1495 } 1496 1497 /*---------------------------------------------------------------------------- 1498 | Returns the result of converting the single-precision floating-point value 1499 | `a' to the 16-bit two's complement integer format. The conversion is 1500 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1501 | Arithmetic, except that the conversion is always rounded toward zero. 1502 | If `a' is a NaN, the largest positive integer is returned. Otherwise, if 1503 | the conversion overflows, the largest integer with the same sign as `a' is 1504 | returned. 1505 *----------------------------------------------------------------------------*/ 1506 1507 int_fast16_t float32_to_int16_round_to_zero(float32 a STATUS_PARAM) 1508 { 1509 flag aSign; 1510 int_fast16_t aExp, shiftCount; 1511 uint32_t aSig; 1512 int32 z; 1513 1514 aSig = extractFloat32Frac( a ); 1515 aExp = extractFloat32Exp( a ); 1516 aSign = extractFloat32Sign( a ); 1517 shiftCount = aExp - 0x8E; 1518 if ( 0 <= shiftCount ) { 1519 if ( float32_val(a) != 0xC7000000 ) { 1520 float_raise( float_flag_invalid STATUS_VAR); 1521 if ( ! aSign || ( ( aExp == 0xFF ) && aSig ) ) { 1522 return 0x7FFF; 1523 } 1524 } 1525 return (int32_t) 0xffff8000; 1526 } 1527 else if ( aExp <= 0x7E ) { 1528 if ( aExp | aSig ) { 1529 STATUS(float_exception_flags) |= float_flag_inexact; 1530 } 1531 return 0; 1532 } 1533 shiftCount -= 0x10; 1534 aSig = ( aSig | 0x00800000 )<<8; 1535 z = aSig>>( - shiftCount ); 1536 if ( (uint32_t) ( aSig<<( shiftCount & 31 ) ) ) { 1537 STATUS(float_exception_flags) |= float_flag_inexact; 1538 } 1539 if ( aSign ) { 1540 z = - z; 1541 } 1542 return z; 1543 1544 } 1545 1546 /*---------------------------------------------------------------------------- 1547 | Returns the result of converting the single-precision floating-point value 1548 | `a' to the 64-bit two's complement integer format. The conversion is 1549 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1550 | Arithmetic---which means in particular that the conversion is rounded 1551 | according to the current rounding mode. If `a' is a NaN, the largest 1552 | positive integer is returned. Otherwise, if the conversion overflows, the 1553 | largest integer with the same sign as `a' is returned. 1554 *----------------------------------------------------------------------------*/ 1555 1556 int64 float32_to_int64( float32 a STATUS_PARAM ) 1557 { 1558 flag aSign; 1559 int_fast16_t aExp, shiftCount; 1560 uint32_t aSig; 1561 uint64_t aSig64, aSigExtra; 1562 a = float32_squash_input_denormal(a STATUS_VAR); 1563 1564 aSig = extractFloat32Frac( a ); 1565 aExp = extractFloat32Exp( a ); 1566 aSign = extractFloat32Sign( a ); 1567 shiftCount = 0xBE - aExp; 1568 if ( shiftCount < 0 ) { 1569 float_raise( float_flag_invalid STATUS_VAR); 1570 if ( ! aSign || ( ( aExp == 0xFF ) && aSig ) ) { 1571 return LIT64( 0x7FFFFFFFFFFFFFFF ); 1572 } 1573 return (int64_t) LIT64( 0x8000000000000000 ); 1574 } 1575 if ( aExp ) aSig |= 0x00800000; 1576 aSig64 = aSig; 1577 aSig64 <<= 40; 1578 shift64ExtraRightJamming( aSig64, 0, shiftCount, &aSig64, &aSigExtra ); 1579 return roundAndPackInt64( aSign, aSig64, aSigExtra STATUS_VAR ); 1580 1581 } 1582 1583 /*---------------------------------------------------------------------------- 1584 | Returns the result of converting the single-precision floating-point value 1585 | `a' to the 64-bit unsigned integer format. The conversion is 1586 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1587 | Arithmetic---which means in particular that the conversion is rounded 1588 | according to the current rounding mode. If `a' is a NaN, the largest 1589 | unsigned integer is returned. Otherwise, if the conversion overflows, the 1590 | largest unsigned integer is returned. If the 'a' is negative, the result 1591 | is rounded and zero is returned; values that do not round to zero will 1592 | raise the inexact exception flag. 1593 *----------------------------------------------------------------------------*/ 1594 1595 uint64 float32_to_uint64(float32 a STATUS_PARAM) 1596 { 1597 flag aSign; 1598 int_fast16_t aExp, shiftCount; 1599 uint32_t aSig; 1600 uint64_t aSig64, aSigExtra; 1601 a = float32_squash_input_denormal(a STATUS_VAR); 1602 1603 aSig = extractFloat32Frac(a); 1604 aExp = extractFloat32Exp(a); 1605 aSign = extractFloat32Sign(a); 1606 if ((aSign) && (aExp > 126)) { 1607 float_raise(float_flag_invalid STATUS_VAR); 1608 if (float32_is_any_nan(a)) { 1609 return LIT64(0xFFFFFFFFFFFFFFFF); 1610 } else { 1611 return 0; 1612 } 1613 } 1614 shiftCount = 0xBE - aExp; 1615 if (aExp) { 1616 aSig |= 0x00800000; 1617 } 1618 if (shiftCount < 0) { 1619 float_raise(float_flag_invalid STATUS_VAR); 1620 return LIT64(0xFFFFFFFFFFFFFFFF); 1621 } 1622 1623 aSig64 = aSig; 1624 aSig64 <<= 40; 1625 shift64ExtraRightJamming(aSig64, 0, shiftCount, &aSig64, &aSigExtra); 1626 return roundAndPackUint64(aSign, aSig64, aSigExtra STATUS_VAR); 1627 } 1628 1629 /*---------------------------------------------------------------------------- 1630 | Returns the result of converting the single-precision floating-point value 1631 | `a' to the 64-bit unsigned integer format. The conversion is 1632 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1633 | Arithmetic, except that the conversion is always rounded toward zero. If 1634 | `a' is a NaN, the largest unsigned integer is returned. Otherwise, if the 1635 | conversion overflows, the largest unsigned integer is returned. If the 1636 | 'a' is negative, the result is rounded and zero is returned; values that do 1637 | not round to zero will raise the inexact flag. 1638 *----------------------------------------------------------------------------*/ 1639 1640 uint64 float32_to_uint64_round_to_zero(float32 a STATUS_PARAM) 1641 { 1642 signed char current_rounding_mode = STATUS(float_rounding_mode); 1643 set_float_rounding_mode(float_round_to_zero STATUS_VAR); 1644 int64_t v = float32_to_uint64(a STATUS_VAR); 1645 set_float_rounding_mode(current_rounding_mode STATUS_VAR); 1646 return v; 1647 } 1648 1649 /*---------------------------------------------------------------------------- 1650 | Returns the result of converting the single-precision floating-point value 1651 | `a' to the 64-bit two's complement integer format. The conversion is 1652 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1653 | Arithmetic, except that the conversion is always rounded toward zero. If 1654 | `a' is a NaN, the largest positive integer is returned. Otherwise, if the 1655 | conversion overflows, the largest integer with the same sign as `a' is 1656 | returned. 1657 *----------------------------------------------------------------------------*/ 1658 1659 int64 float32_to_int64_round_to_zero( float32 a STATUS_PARAM ) 1660 { 1661 flag aSign; 1662 int_fast16_t aExp, shiftCount; 1663 uint32_t aSig; 1664 uint64_t aSig64; 1665 int64 z; 1666 a = float32_squash_input_denormal(a STATUS_VAR); 1667 1668 aSig = extractFloat32Frac( a ); 1669 aExp = extractFloat32Exp( a ); 1670 aSign = extractFloat32Sign( a ); 1671 shiftCount = aExp - 0xBE; 1672 if ( 0 <= shiftCount ) { 1673 if ( float32_val(a) != 0xDF000000 ) { 1674 float_raise( float_flag_invalid STATUS_VAR); 1675 if ( ! aSign || ( ( aExp == 0xFF ) && aSig ) ) { 1676 return LIT64( 0x7FFFFFFFFFFFFFFF ); 1677 } 1678 } 1679 return (int64_t) LIT64( 0x8000000000000000 ); 1680 } 1681 else if ( aExp <= 0x7E ) { 1682 if ( aExp | aSig ) STATUS(float_exception_flags) |= float_flag_inexact; 1683 return 0; 1684 } 1685 aSig64 = aSig | 0x00800000; 1686 aSig64 <<= 40; 1687 z = aSig64>>( - shiftCount ); 1688 if ( (uint64_t) ( aSig64<<( shiftCount & 63 ) ) ) { 1689 STATUS(float_exception_flags) |= float_flag_inexact; 1690 } 1691 if ( aSign ) z = - z; 1692 return z; 1693 1694 } 1695 1696 /*---------------------------------------------------------------------------- 1697 | Returns the result of converting the single-precision floating-point value 1698 | `a' to the double-precision floating-point format. The conversion is 1699 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1700 | Arithmetic. 1701 *----------------------------------------------------------------------------*/ 1702 1703 float64 float32_to_float64( float32 a STATUS_PARAM ) 1704 { 1705 flag aSign; 1706 int_fast16_t aExp; 1707 uint32_t aSig; 1708 a = float32_squash_input_denormal(a STATUS_VAR); 1709 1710 aSig = extractFloat32Frac( a ); 1711 aExp = extractFloat32Exp( a ); 1712 aSign = extractFloat32Sign( a ); 1713 if ( aExp == 0xFF ) { 1714 if ( aSig ) return commonNaNToFloat64( float32ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 1715 return packFloat64( aSign, 0x7FF, 0 ); 1716 } 1717 if ( aExp == 0 ) { 1718 if ( aSig == 0 ) return packFloat64( aSign, 0, 0 ); 1719 normalizeFloat32Subnormal( aSig, &aExp, &aSig ); 1720 --aExp; 1721 } 1722 return packFloat64( aSign, aExp + 0x380, ( (uint64_t) aSig )<<29 ); 1723 1724 } 1725 1726 /*---------------------------------------------------------------------------- 1727 | Returns the result of converting the single-precision floating-point value 1728 | `a' to the extended double-precision floating-point format. The conversion 1729 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 1730 | Arithmetic. 1731 *----------------------------------------------------------------------------*/ 1732 1733 floatx80 float32_to_floatx80( float32 a STATUS_PARAM ) 1734 { 1735 flag aSign; 1736 int_fast16_t aExp; 1737 uint32_t aSig; 1738 1739 a = float32_squash_input_denormal(a STATUS_VAR); 1740 aSig = extractFloat32Frac( a ); 1741 aExp = extractFloat32Exp( a ); 1742 aSign = extractFloat32Sign( a ); 1743 if ( aExp == 0xFF ) { 1744 if ( aSig ) return commonNaNToFloatx80( float32ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 1745 return packFloatx80( aSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 1746 } 1747 if ( aExp == 0 ) { 1748 if ( aSig == 0 ) return packFloatx80( aSign, 0, 0 ); 1749 normalizeFloat32Subnormal( aSig, &aExp, &aSig ); 1750 } 1751 aSig |= 0x00800000; 1752 return packFloatx80( aSign, aExp + 0x3F80, ( (uint64_t) aSig )<<40 ); 1753 1754 } 1755 1756 /*---------------------------------------------------------------------------- 1757 | Returns the result of converting the single-precision floating-point value 1758 | `a' to the double-precision floating-point format. The conversion is 1759 | performed according to the IEC/IEEE Standard for Binary Floating-Point 1760 | Arithmetic. 1761 *----------------------------------------------------------------------------*/ 1762 1763 float128 float32_to_float128( float32 a STATUS_PARAM ) 1764 { 1765 flag aSign; 1766 int_fast16_t aExp; 1767 uint32_t aSig; 1768 1769 a = float32_squash_input_denormal(a STATUS_VAR); 1770 aSig = extractFloat32Frac( a ); 1771 aExp = extractFloat32Exp( a ); 1772 aSign = extractFloat32Sign( a ); 1773 if ( aExp == 0xFF ) { 1774 if ( aSig ) return commonNaNToFloat128( float32ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 1775 return packFloat128( aSign, 0x7FFF, 0, 0 ); 1776 } 1777 if ( aExp == 0 ) { 1778 if ( aSig == 0 ) return packFloat128( aSign, 0, 0, 0 ); 1779 normalizeFloat32Subnormal( aSig, &aExp, &aSig ); 1780 --aExp; 1781 } 1782 return packFloat128( aSign, aExp + 0x3F80, ( (uint64_t) aSig )<<25, 0 ); 1783 1784 } 1785 1786 /*---------------------------------------------------------------------------- 1787 | Rounds the single-precision floating-point value `a' to an integer, and 1788 | returns the result as a single-precision floating-point value. The 1789 | operation is performed according to the IEC/IEEE Standard for Binary 1790 | Floating-Point Arithmetic. 1791 *----------------------------------------------------------------------------*/ 1792 1793 float32 float32_round_to_int( float32 a STATUS_PARAM) 1794 { 1795 flag aSign; 1796 int_fast16_t aExp; 1797 uint32_t lastBitMask, roundBitsMask; 1798 uint32_t z; 1799 a = float32_squash_input_denormal(a STATUS_VAR); 1800 1801 aExp = extractFloat32Exp( a ); 1802 if ( 0x96 <= aExp ) { 1803 if ( ( aExp == 0xFF ) && extractFloat32Frac( a ) ) { 1804 return propagateFloat32NaN( a, a STATUS_VAR ); 1805 } 1806 return a; 1807 } 1808 if ( aExp <= 0x7E ) { 1809 if ( (uint32_t) ( float32_val(a)<<1 ) == 0 ) return a; 1810 STATUS(float_exception_flags) |= float_flag_inexact; 1811 aSign = extractFloat32Sign( a ); 1812 switch ( STATUS(float_rounding_mode) ) { 1813 case float_round_nearest_even: 1814 if ( ( aExp == 0x7E ) && extractFloat32Frac( a ) ) { 1815 return packFloat32( aSign, 0x7F, 0 ); 1816 } 1817 break; 1818 case float_round_ties_away: 1819 if (aExp == 0x7E) { 1820 return packFloat32(aSign, 0x7F, 0); 1821 } 1822 break; 1823 case float_round_down: 1824 return make_float32(aSign ? 0xBF800000 : 0); 1825 case float_round_up: 1826 return make_float32(aSign ? 0x80000000 : 0x3F800000); 1827 } 1828 return packFloat32( aSign, 0, 0 ); 1829 } 1830 lastBitMask = 1; 1831 lastBitMask <<= 0x96 - aExp; 1832 roundBitsMask = lastBitMask - 1; 1833 z = float32_val(a); 1834 switch (STATUS(float_rounding_mode)) { 1835 case float_round_nearest_even: 1836 z += lastBitMask>>1; 1837 if ((z & roundBitsMask) == 0) { 1838 z &= ~lastBitMask; 1839 } 1840 break; 1841 case float_round_ties_away: 1842 z += lastBitMask >> 1; 1843 break; 1844 case float_round_to_zero: 1845 break; 1846 case float_round_up: 1847 if (!extractFloat32Sign(make_float32(z))) { 1848 z += roundBitsMask; 1849 } 1850 break; 1851 case float_round_down: 1852 if (extractFloat32Sign(make_float32(z))) { 1853 z += roundBitsMask; 1854 } 1855 break; 1856 default: 1857 abort(); 1858 } 1859 z &= ~ roundBitsMask; 1860 if ( z != float32_val(a) ) STATUS(float_exception_flags) |= float_flag_inexact; 1861 return make_float32(z); 1862 1863 } 1864 1865 /*---------------------------------------------------------------------------- 1866 | Returns the result of adding the absolute values of the single-precision 1867 | floating-point values `a' and `b'. If `zSign' is 1, the sum is negated 1868 | before being returned. `zSign' is ignored if the result is a NaN. 1869 | The addition is performed according to the IEC/IEEE Standard for Binary 1870 | Floating-Point Arithmetic. 1871 *----------------------------------------------------------------------------*/ 1872 1873 static float32 addFloat32Sigs( float32 a, float32 b, flag zSign STATUS_PARAM) 1874 { 1875 int_fast16_t aExp, bExp, zExp; 1876 uint32_t aSig, bSig, zSig; 1877 int_fast16_t expDiff; 1878 1879 aSig = extractFloat32Frac( a ); 1880 aExp = extractFloat32Exp( a ); 1881 bSig = extractFloat32Frac( b ); 1882 bExp = extractFloat32Exp( b ); 1883 expDiff = aExp - bExp; 1884 aSig <<= 6; 1885 bSig <<= 6; 1886 if ( 0 < expDiff ) { 1887 if ( aExp == 0xFF ) { 1888 if ( aSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 1889 return a; 1890 } 1891 if ( bExp == 0 ) { 1892 --expDiff; 1893 } 1894 else { 1895 bSig |= 0x20000000; 1896 } 1897 shift32RightJamming( bSig, expDiff, &bSig ); 1898 zExp = aExp; 1899 } 1900 else if ( expDiff < 0 ) { 1901 if ( bExp == 0xFF ) { 1902 if ( bSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 1903 return packFloat32( zSign, 0xFF, 0 ); 1904 } 1905 if ( aExp == 0 ) { 1906 ++expDiff; 1907 } 1908 else { 1909 aSig |= 0x20000000; 1910 } 1911 shift32RightJamming( aSig, - expDiff, &aSig ); 1912 zExp = bExp; 1913 } 1914 else { 1915 if ( aExp == 0xFF ) { 1916 if ( aSig | bSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 1917 return a; 1918 } 1919 if ( aExp == 0 ) { 1920 if (STATUS(flush_to_zero)) { 1921 if (aSig | bSig) { 1922 float_raise(float_flag_output_denormal STATUS_VAR); 1923 } 1924 return packFloat32(zSign, 0, 0); 1925 } 1926 return packFloat32( zSign, 0, ( aSig + bSig )>>6 ); 1927 } 1928 zSig = 0x40000000 + aSig + bSig; 1929 zExp = aExp; 1930 goto roundAndPack; 1931 } 1932 aSig |= 0x20000000; 1933 zSig = ( aSig + bSig )<<1; 1934 --zExp; 1935 if ( (int32_t) zSig < 0 ) { 1936 zSig = aSig + bSig; 1937 ++zExp; 1938 } 1939 roundAndPack: 1940 return roundAndPackFloat32( zSign, zExp, zSig STATUS_VAR ); 1941 1942 } 1943 1944 /*---------------------------------------------------------------------------- 1945 | Returns the result of subtracting the absolute values of the single- 1946 | precision floating-point values `a' and `b'. If `zSign' is 1, the 1947 | difference is negated before being returned. `zSign' is ignored if the 1948 | result is a NaN. The subtraction is performed according to the IEC/IEEE 1949 | Standard for Binary Floating-Point Arithmetic. 1950 *----------------------------------------------------------------------------*/ 1951 1952 static float32 subFloat32Sigs( float32 a, float32 b, flag zSign STATUS_PARAM) 1953 { 1954 int_fast16_t aExp, bExp, zExp; 1955 uint32_t aSig, bSig, zSig; 1956 int_fast16_t expDiff; 1957 1958 aSig = extractFloat32Frac( a ); 1959 aExp = extractFloat32Exp( a ); 1960 bSig = extractFloat32Frac( b ); 1961 bExp = extractFloat32Exp( b ); 1962 expDiff = aExp - bExp; 1963 aSig <<= 7; 1964 bSig <<= 7; 1965 if ( 0 < expDiff ) goto aExpBigger; 1966 if ( expDiff < 0 ) goto bExpBigger; 1967 if ( aExp == 0xFF ) { 1968 if ( aSig | bSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 1969 float_raise( float_flag_invalid STATUS_VAR); 1970 return float32_default_nan; 1971 } 1972 if ( aExp == 0 ) { 1973 aExp = 1; 1974 bExp = 1; 1975 } 1976 if ( bSig < aSig ) goto aBigger; 1977 if ( aSig < bSig ) goto bBigger; 1978 return packFloat32( STATUS(float_rounding_mode) == float_round_down, 0, 0 ); 1979 bExpBigger: 1980 if ( bExp == 0xFF ) { 1981 if ( bSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 1982 return packFloat32( zSign ^ 1, 0xFF, 0 ); 1983 } 1984 if ( aExp == 0 ) { 1985 ++expDiff; 1986 } 1987 else { 1988 aSig |= 0x40000000; 1989 } 1990 shift32RightJamming( aSig, - expDiff, &aSig ); 1991 bSig |= 0x40000000; 1992 bBigger: 1993 zSig = bSig - aSig; 1994 zExp = bExp; 1995 zSign ^= 1; 1996 goto normalizeRoundAndPack; 1997 aExpBigger: 1998 if ( aExp == 0xFF ) { 1999 if ( aSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 2000 return a; 2001 } 2002 if ( bExp == 0 ) { 2003 --expDiff; 2004 } 2005 else { 2006 bSig |= 0x40000000; 2007 } 2008 shift32RightJamming( bSig, expDiff, &bSig ); 2009 aSig |= 0x40000000; 2010 aBigger: 2011 zSig = aSig - bSig; 2012 zExp = aExp; 2013 normalizeRoundAndPack: 2014 --zExp; 2015 return normalizeRoundAndPackFloat32( zSign, zExp, zSig STATUS_VAR ); 2016 2017 } 2018 2019 /*---------------------------------------------------------------------------- 2020 | Returns the result of adding the single-precision floating-point values `a' 2021 | and `b'. The operation is performed according to the IEC/IEEE Standard for 2022 | Binary Floating-Point Arithmetic. 2023 *----------------------------------------------------------------------------*/ 2024 2025 float32 float32_add( float32 a, float32 b STATUS_PARAM ) 2026 { 2027 flag aSign, bSign; 2028 a = float32_squash_input_denormal(a STATUS_VAR); 2029 b = float32_squash_input_denormal(b STATUS_VAR); 2030 2031 aSign = extractFloat32Sign( a ); 2032 bSign = extractFloat32Sign( b ); 2033 if ( aSign == bSign ) { 2034 return addFloat32Sigs( a, b, aSign STATUS_VAR); 2035 } 2036 else { 2037 return subFloat32Sigs( a, b, aSign STATUS_VAR ); 2038 } 2039 2040 } 2041 2042 /*---------------------------------------------------------------------------- 2043 | Returns the result of subtracting the single-precision floating-point values 2044 | `a' and `b'. The operation is performed according to the IEC/IEEE Standard 2045 | for Binary Floating-Point Arithmetic. 2046 *----------------------------------------------------------------------------*/ 2047 2048 float32 float32_sub( float32 a, float32 b STATUS_PARAM ) 2049 { 2050 flag aSign, bSign; 2051 a = float32_squash_input_denormal(a STATUS_VAR); 2052 b = float32_squash_input_denormal(b STATUS_VAR); 2053 2054 aSign = extractFloat32Sign( a ); 2055 bSign = extractFloat32Sign( b ); 2056 if ( aSign == bSign ) { 2057 return subFloat32Sigs( a, b, aSign STATUS_VAR ); 2058 } 2059 else { 2060 return addFloat32Sigs( a, b, aSign STATUS_VAR ); 2061 } 2062 2063 } 2064 2065 /*---------------------------------------------------------------------------- 2066 | Returns the result of multiplying the single-precision floating-point values 2067 | `a' and `b'. The operation is performed according to the IEC/IEEE Standard 2068 | for Binary Floating-Point Arithmetic. 2069 *----------------------------------------------------------------------------*/ 2070 2071 float32 float32_mul( float32 a, float32 b STATUS_PARAM ) 2072 { 2073 flag aSign, bSign, zSign; 2074 int_fast16_t aExp, bExp, zExp; 2075 uint32_t aSig, bSig; 2076 uint64_t zSig64; 2077 uint32_t zSig; 2078 2079 a = float32_squash_input_denormal(a STATUS_VAR); 2080 b = float32_squash_input_denormal(b STATUS_VAR); 2081 2082 aSig = extractFloat32Frac( a ); 2083 aExp = extractFloat32Exp( a ); 2084 aSign = extractFloat32Sign( a ); 2085 bSig = extractFloat32Frac( b ); 2086 bExp = extractFloat32Exp( b ); 2087 bSign = extractFloat32Sign( b ); 2088 zSign = aSign ^ bSign; 2089 if ( aExp == 0xFF ) { 2090 if ( aSig || ( ( bExp == 0xFF ) && bSig ) ) { 2091 return propagateFloat32NaN( a, b STATUS_VAR ); 2092 } 2093 if ( ( bExp | bSig ) == 0 ) { 2094 float_raise( float_flag_invalid STATUS_VAR); 2095 return float32_default_nan; 2096 } 2097 return packFloat32( zSign, 0xFF, 0 ); 2098 } 2099 if ( bExp == 0xFF ) { 2100 if ( bSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 2101 if ( ( aExp | aSig ) == 0 ) { 2102 float_raise( float_flag_invalid STATUS_VAR); 2103 return float32_default_nan; 2104 } 2105 return packFloat32( zSign, 0xFF, 0 ); 2106 } 2107 if ( aExp == 0 ) { 2108 if ( aSig == 0 ) return packFloat32( zSign, 0, 0 ); 2109 normalizeFloat32Subnormal( aSig, &aExp, &aSig ); 2110 } 2111 if ( bExp == 0 ) { 2112 if ( bSig == 0 ) return packFloat32( zSign, 0, 0 ); 2113 normalizeFloat32Subnormal( bSig, &bExp, &bSig ); 2114 } 2115 zExp = aExp + bExp - 0x7F; 2116 aSig = ( aSig | 0x00800000 )<<7; 2117 bSig = ( bSig | 0x00800000 )<<8; 2118 shift64RightJamming( ( (uint64_t) aSig ) * bSig, 32, &zSig64 ); 2119 zSig = zSig64; 2120 if ( 0 <= (int32_t) ( zSig<<1 ) ) { 2121 zSig <<= 1; 2122 --zExp; 2123 } 2124 return roundAndPackFloat32( zSign, zExp, zSig STATUS_VAR ); 2125 2126 } 2127 2128 /*---------------------------------------------------------------------------- 2129 | Returns the result of dividing the single-precision floating-point value `a' 2130 | by the corresponding value `b'. The operation is performed according to the 2131 | IEC/IEEE Standard for Binary Floating-Point Arithmetic. 2132 *----------------------------------------------------------------------------*/ 2133 2134 float32 float32_div( float32 a, float32 b STATUS_PARAM ) 2135 { 2136 flag aSign, bSign, zSign; 2137 int_fast16_t aExp, bExp, zExp; 2138 uint32_t aSig, bSig, zSig; 2139 a = float32_squash_input_denormal(a STATUS_VAR); 2140 b = float32_squash_input_denormal(b STATUS_VAR); 2141 2142 aSig = extractFloat32Frac( a ); 2143 aExp = extractFloat32Exp( a ); 2144 aSign = extractFloat32Sign( a ); 2145 bSig = extractFloat32Frac( b ); 2146 bExp = extractFloat32Exp( b ); 2147 bSign = extractFloat32Sign( b ); 2148 zSign = aSign ^ bSign; 2149 if ( aExp == 0xFF ) { 2150 if ( aSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 2151 if ( bExp == 0xFF ) { 2152 if ( bSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 2153 float_raise( float_flag_invalid STATUS_VAR); 2154 return float32_default_nan; 2155 } 2156 return packFloat32( zSign, 0xFF, 0 ); 2157 } 2158 if ( bExp == 0xFF ) { 2159 if ( bSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 2160 return packFloat32( zSign, 0, 0 ); 2161 } 2162 if ( bExp == 0 ) { 2163 if ( bSig == 0 ) { 2164 if ( ( aExp | aSig ) == 0 ) { 2165 float_raise( float_flag_invalid STATUS_VAR); 2166 return float32_default_nan; 2167 } 2168 float_raise( float_flag_divbyzero STATUS_VAR); 2169 return packFloat32( zSign, 0xFF, 0 ); 2170 } 2171 normalizeFloat32Subnormal( bSig, &bExp, &bSig ); 2172 } 2173 if ( aExp == 0 ) { 2174 if ( aSig == 0 ) return packFloat32( zSign, 0, 0 ); 2175 normalizeFloat32Subnormal( aSig, &aExp, &aSig ); 2176 } 2177 zExp = aExp - bExp + 0x7D; 2178 aSig = ( aSig | 0x00800000 )<<7; 2179 bSig = ( bSig | 0x00800000 )<<8; 2180 if ( bSig <= ( aSig + aSig ) ) { 2181 aSig >>= 1; 2182 ++zExp; 2183 } 2184 zSig = ( ( (uint64_t) aSig )<<32 ) / bSig; 2185 if ( ( zSig & 0x3F ) == 0 ) { 2186 zSig |= ( (uint64_t) bSig * zSig != ( (uint64_t) aSig )<<32 ); 2187 } 2188 return roundAndPackFloat32( zSign, zExp, zSig STATUS_VAR ); 2189 2190 } 2191 2192 /*---------------------------------------------------------------------------- 2193 | Returns the remainder of the single-precision floating-point value `a' 2194 | with respect to the corresponding value `b'. The operation is performed 2195 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 2196 *----------------------------------------------------------------------------*/ 2197 2198 float32 float32_rem( float32 a, float32 b STATUS_PARAM ) 2199 { 2200 flag aSign, zSign; 2201 int_fast16_t aExp, bExp, expDiff; 2202 uint32_t aSig, bSig; 2203 uint32_t q; 2204 uint64_t aSig64, bSig64, q64; 2205 uint32_t alternateASig; 2206 int32_t sigMean; 2207 a = float32_squash_input_denormal(a STATUS_VAR); 2208 b = float32_squash_input_denormal(b STATUS_VAR); 2209 2210 aSig = extractFloat32Frac( a ); 2211 aExp = extractFloat32Exp( a ); 2212 aSign = extractFloat32Sign( a ); 2213 bSig = extractFloat32Frac( b ); 2214 bExp = extractFloat32Exp( b ); 2215 if ( aExp == 0xFF ) { 2216 if ( aSig || ( ( bExp == 0xFF ) && bSig ) ) { 2217 return propagateFloat32NaN( a, b STATUS_VAR ); 2218 } 2219 float_raise( float_flag_invalid STATUS_VAR); 2220 return float32_default_nan; 2221 } 2222 if ( bExp == 0xFF ) { 2223 if ( bSig ) return propagateFloat32NaN( a, b STATUS_VAR ); 2224 return a; 2225 } 2226 if ( bExp == 0 ) { 2227 if ( bSig == 0 ) { 2228 float_raise( float_flag_invalid STATUS_VAR); 2229 return float32_default_nan; 2230 } 2231 normalizeFloat32Subnormal( bSig, &bExp, &bSig ); 2232 } 2233 if ( aExp == 0 ) { 2234 if ( aSig == 0 ) return a; 2235 normalizeFloat32Subnormal( aSig, &aExp, &aSig ); 2236 } 2237 expDiff = aExp - bExp; 2238 aSig |= 0x00800000; 2239 bSig |= 0x00800000; 2240 if ( expDiff < 32 ) { 2241 aSig <<= 8; 2242 bSig <<= 8; 2243 if ( expDiff < 0 ) { 2244 if ( expDiff < -1 ) return a; 2245 aSig >>= 1; 2246 } 2247 q = ( bSig <= aSig ); 2248 if ( q ) aSig -= bSig; 2249 if ( 0 < expDiff ) { 2250 q = ( ( (uint64_t) aSig )<<32 ) / bSig; 2251 q >>= 32 - expDiff; 2252 bSig >>= 2; 2253 aSig = ( ( aSig>>1 )<<( expDiff - 1 ) ) - bSig * q; 2254 } 2255 else { 2256 aSig >>= 2; 2257 bSig >>= 2; 2258 } 2259 } 2260 else { 2261 if ( bSig <= aSig ) aSig -= bSig; 2262 aSig64 = ( (uint64_t) aSig )<<40; 2263 bSig64 = ( (uint64_t) bSig )<<40; 2264 expDiff -= 64; 2265 while ( 0 < expDiff ) { 2266 q64 = estimateDiv128To64( aSig64, 0, bSig64 ); 2267 q64 = ( 2 < q64 ) ? q64 - 2 : 0; 2268 aSig64 = - ( ( bSig * q64 )<<38 ); 2269 expDiff -= 62; 2270 } 2271 expDiff += 64; 2272 q64 = estimateDiv128To64( aSig64, 0, bSig64 ); 2273 q64 = ( 2 < q64 ) ? q64 - 2 : 0; 2274 q = q64>>( 64 - expDiff ); 2275 bSig <<= 6; 2276 aSig = ( ( aSig64>>33 )<<( expDiff - 1 ) ) - bSig * q; 2277 } 2278 do { 2279 alternateASig = aSig; 2280 ++q; 2281 aSig -= bSig; 2282 } while ( 0 <= (int32_t) aSig ); 2283 sigMean = aSig + alternateASig; 2284 if ( ( sigMean < 0 ) || ( ( sigMean == 0 ) && ( q & 1 ) ) ) { 2285 aSig = alternateASig; 2286 } 2287 zSign = ( (int32_t) aSig < 0 ); 2288 if ( zSign ) aSig = - aSig; 2289 return normalizeRoundAndPackFloat32( aSign ^ zSign, bExp, aSig STATUS_VAR ); 2290 2291 } 2292 2293 /*---------------------------------------------------------------------------- 2294 | Returns the result of multiplying the single-precision floating-point values 2295 | `a' and `b' then adding 'c', with no intermediate rounding step after the 2296 | multiplication. The operation is performed according to the IEC/IEEE 2297 | Standard for Binary Floating-Point Arithmetic 754-2008. 2298 | The flags argument allows the caller to select negation of the 2299 | addend, the intermediate product, or the final result. (The difference 2300 | between this and having the caller do a separate negation is that negating 2301 | externally will flip the sign bit on NaNs.) 2302 *----------------------------------------------------------------------------*/ 2303 2304 float32 float32_muladd(float32 a, float32 b, float32 c, int flags STATUS_PARAM) 2305 { 2306 flag aSign, bSign, cSign, zSign; 2307 int_fast16_t aExp, bExp, cExp, pExp, zExp, expDiff; 2308 uint32_t aSig, bSig, cSig; 2309 flag pInf, pZero, pSign; 2310 uint64_t pSig64, cSig64, zSig64; 2311 uint32_t pSig; 2312 int shiftcount; 2313 flag signflip, infzero; 2314 2315 a = float32_squash_input_denormal(a STATUS_VAR); 2316 b = float32_squash_input_denormal(b STATUS_VAR); 2317 c = float32_squash_input_denormal(c STATUS_VAR); 2318 aSig = extractFloat32Frac(a); 2319 aExp = extractFloat32Exp(a); 2320 aSign = extractFloat32Sign(a); 2321 bSig = extractFloat32Frac(b); 2322 bExp = extractFloat32Exp(b); 2323 bSign = extractFloat32Sign(b); 2324 cSig = extractFloat32Frac(c); 2325 cExp = extractFloat32Exp(c); 2326 cSign = extractFloat32Sign(c); 2327 2328 infzero = ((aExp == 0 && aSig == 0 && bExp == 0xff && bSig == 0) || 2329 (aExp == 0xff && aSig == 0 && bExp == 0 && bSig == 0)); 2330 2331 /* It is implementation-defined whether the cases of (0,inf,qnan) 2332 * and (inf,0,qnan) raise InvalidOperation or not (and what QNaN 2333 * they return if they do), so we have to hand this information 2334 * off to the target-specific pick-a-NaN routine. 2335 */ 2336 if (((aExp == 0xff) && aSig) || 2337 ((bExp == 0xff) && bSig) || 2338 ((cExp == 0xff) && cSig)) { 2339 return propagateFloat32MulAddNaN(a, b, c, infzero STATUS_VAR); 2340 } 2341 2342 if (infzero) { 2343 float_raise(float_flag_invalid STATUS_VAR); 2344 return float32_default_nan; 2345 } 2346 2347 if (flags & float_muladd_negate_c) { 2348 cSign ^= 1; 2349 } 2350 2351 signflip = (flags & float_muladd_negate_result) ? 1 : 0; 2352 2353 /* Work out the sign and type of the product */ 2354 pSign = aSign ^ bSign; 2355 if (flags & float_muladd_negate_product) { 2356 pSign ^= 1; 2357 } 2358 pInf = (aExp == 0xff) || (bExp == 0xff); 2359 pZero = ((aExp | aSig) == 0) || ((bExp | bSig) == 0); 2360 2361 if (cExp == 0xff) { 2362 if (pInf && (pSign ^ cSign)) { 2363 /* addition of opposite-signed infinities => InvalidOperation */ 2364 float_raise(float_flag_invalid STATUS_VAR); 2365 return float32_default_nan; 2366 } 2367 /* Otherwise generate an infinity of the same sign */ 2368 return packFloat32(cSign ^ signflip, 0xff, 0); 2369 } 2370 2371 if (pInf) { 2372 return packFloat32(pSign ^ signflip, 0xff, 0); 2373 } 2374 2375 if (pZero) { 2376 if (cExp == 0) { 2377 if (cSig == 0) { 2378 /* Adding two exact zeroes */ 2379 if (pSign == cSign) { 2380 zSign = pSign; 2381 } else if (STATUS(float_rounding_mode) == float_round_down) { 2382 zSign = 1; 2383 } else { 2384 zSign = 0; 2385 } 2386 return packFloat32(zSign ^ signflip, 0, 0); 2387 } 2388 /* Exact zero plus a denorm */ 2389 if (STATUS(flush_to_zero)) { 2390 float_raise(float_flag_output_denormal STATUS_VAR); 2391 return packFloat32(cSign ^ signflip, 0, 0); 2392 } 2393 } 2394 /* Zero plus something non-zero : just return the something */ 2395 if (flags & float_muladd_halve_result) { 2396 if (cExp == 0) { 2397 normalizeFloat32Subnormal(cSig, &cExp, &cSig); 2398 } 2399 /* Subtract one to halve, and one again because roundAndPackFloat32 2400 * wants one less than the true exponent. 2401 */ 2402 cExp -= 2; 2403 cSig = (cSig | 0x00800000) << 7; 2404 return roundAndPackFloat32(cSign ^ signflip, cExp, cSig STATUS_VAR); 2405 } 2406 return packFloat32(cSign ^ signflip, cExp, cSig); 2407 } 2408 2409 if (aExp == 0) { 2410 normalizeFloat32Subnormal(aSig, &aExp, &aSig); 2411 } 2412 if (bExp == 0) { 2413 normalizeFloat32Subnormal(bSig, &bExp, &bSig); 2414 } 2415 2416 /* Calculate the actual result a * b + c */ 2417 2418 /* Multiply first; this is easy. */ 2419 /* NB: we subtract 0x7e where float32_mul() subtracts 0x7f 2420 * because we want the true exponent, not the "one-less-than" 2421 * flavour that roundAndPackFloat32() takes. 2422 */ 2423 pExp = aExp + bExp - 0x7e; 2424 aSig = (aSig | 0x00800000) << 7; 2425 bSig = (bSig | 0x00800000) << 8; 2426 pSig64 = (uint64_t)aSig * bSig; 2427 if ((int64_t)(pSig64 << 1) >= 0) { 2428 pSig64 <<= 1; 2429 pExp--; 2430 } 2431 2432 zSign = pSign ^ signflip; 2433 2434 /* Now pSig64 is the significand of the multiply, with the explicit bit in 2435 * position 62. 2436 */ 2437 if (cExp == 0) { 2438 if (!cSig) { 2439 /* Throw out the special case of c being an exact zero now */ 2440 shift64RightJamming(pSig64, 32, &pSig64); 2441 pSig = pSig64; 2442 if (flags & float_muladd_halve_result) { 2443 pExp--; 2444 } 2445 return roundAndPackFloat32(zSign, pExp - 1, 2446 pSig STATUS_VAR); 2447 } 2448 normalizeFloat32Subnormal(cSig, &cExp, &cSig); 2449 } 2450 2451 cSig64 = (uint64_t)cSig << (62 - 23); 2452 cSig64 |= LIT64(0x4000000000000000); 2453 expDiff = pExp - cExp; 2454 2455 if (pSign == cSign) { 2456 /* Addition */ 2457 if (expDiff > 0) { 2458 /* scale c to match p */ 2459 shift64RightJamming(cSig64, expDiff, &cSig64); 2460 zExp = pExp; 2461 } else if (expDiff < 0) { 2462 /* scale p to match c */ 2463 shift64RightJamming(pSig64, -expDiff, &pSig64); 2464 zExp = cExp; 2465 } else { 2466 /* no scaling needed */ 2467 zExp = cExp; 2468 } 2469 /* Add significands and make sure explicit bit ends up in posn 62 */ 2470 zSig64 = pSig64 + cSig64; 2471 if ((int64_t)zSig64 < 0) { 2472 shift64RightJamming(zSig64, 1, &zSig64); 2473 } else { 2474 zExp--; 2475 } 2476 } else { 2477 /* Subtraction */ 2478 if (expDiff > 0) { 2479 shift64RightJamming(cSig64, expDiff, &cSig64); 2480 zSig64 = pSig64 - cSig64; 2481 zExp = pExp; 2482 } else if (expDiff < 0) { 2483 shift64RightJamming(pSig64, -expDiff, &pSig64); 2484 zSig64 = cSig64 - pSig64; 2485 zExp = cExp; 2486 zSign ^= 1; 2487 } else { 2488 zExp = pExp; 2489 if (cSig64 < pSig64) { 2490 zSig64 = pSig64 - cSig64; 2491 } else if (pSig64 < cSig64) { 2492 zSig64 = cSig64 - pSig64; 2493 zSign ^= 1; 2494 } else { 2495 /* Exact zero */ 2496 zSign = signflip; 2497 if (STATUS(float_rounding_mode) == float_round_down) { 2498 zSign ^= 1; 2499 } 2500 return packFloat32(zSign, 0, 0); 2501 } 2502 } 2503 --zExp; 2504 /* Normalize to put the explicit bit back into bit 62. */ 2505 shiftcount = countLeadingZeros64(zSig64) - 1; 2506 zSig64 <<= shiftcount; 2507 zExp -= shiftcount; 2508 } 2509 if (flags & float_muladd_halve_result) { 2510 zExp--; 2511 } 2512 2513 shift64RightJamming(zSig64, 32, &zSig64); 2514 return roundAndPackFloat32(zSign, zExp, zSig64 STATUS_VAR); 2515 } 2516 2517 2518 /*---------------------------------------------------------------------------- 2519 | Returns the square root of the single-precision floating-point value `a'. 2520 | The operation is performed according to the IEC/IEEE Standard for Binary 2521 | Floating-Point Arithmetic. 2522 *----------------------------------------------------------------------------*/ 2523 2524 float32 float32_sqrt( float32 a STATUS_PARAM ) 2525 { 2526 flag aSign; 2527 int_fast16_t aExp, zExp; 2528 uint32_t aSig, zSig; 2529 uint64_t rem, term; 2530 a = float32_squash_input_denormal(a STATUS_VAR); 2531 2532 aSig = extractFloat32Frac( a ); 2533 aExp = extractFloat32Exp( a ); 2534 aSign = extractFloat32Sign( a ); 2535 if ( aExp == 0xFF ) { 2536 if ( aSig ) return propagateFloat32NaN( a, float32_zero STATUS_VAR ); 2537 if ( ! aSign ) return a; 2538 float_raise( float_flag_invalid STATUS_VAR); 2539 return float32_default_nan; 2540 } 2541 if ( aSign ) { 2542 if ( ( aExp | aSig ) == 0 ) return a; 2543 float_raise( float_flag_invalid STATUS_VAR); 2544 return float32_default_nan; 2545 } 2546 if ( aExp == 0 ) { 2547 if ( aSig == 0 ) return float32_zero; 2548 normalizeFloat32Subnormal( aSig, &aExp, &aSig ); 2549 } 2550 zExp = ( ( aExp - 0x7F )>>1 ) + 0x7E; 2551 aSig = ( aSig | 0x00800000 )<<8; 2552 zSig = estimateSqrt32( aExp, aSig ) + 2; 2553 if ( ( zSig & 0x7F ) <= 5 ) { 2554 if ( zSig < 2 ) { 2555 zSig = 0x7FFFFFFF; 2556 goto roundAndPack; 2557 } 2558 aSig >>= aExp & 1; 2559 term = ( (uint64_t) zSig ) * zSig; 2560 rem = ( ( (uint64_t) aSig )<<32 ) - term; 2561 while ( (int64_t) rem < 0 ) { 2562 --zSig; 2563 rem += ( ( (uint64_t) zSig )<<1 ) | 1; 2564 } 2565 zSig |= ( rem != 0 ); 2566 } 2567 shift32RightJamming( zSig, 1, &zSig ); 2568 roundAndPack: 2569 return roundAndPackFloat32( 0, zExp, zSig STATUS_VAR ); 2570 2571 } 2572 2573 /*---------------------------------------------------------------------------- 2574 | Returns the binary exponential of the single-precision floating-point value 2575 | `a'. The operation is performed according to the IEC/IEEE Standard for 2576 | Binary Floating-Point Arithmetic. 2577 | 2578 | Uses the following identities: 2579 | 2580 | 1. ------------------------------------------------------------------------- 2581 | x x*ln(2) 2582 | 2 = e 2583 | 2584 | 2. ------------------------------------------------------------------------- 2585 | 2 3 4 5 n 2586 | x x x x x x x 2587 | e = 1 + --- + --- + --- + --- + --- + ... + --- + ... 2588 | 1! 2! 3! 4! 5! n! 2589 *----------------------------------------------------------------------------*/ 2590 2591 static const float64 float32_exp2_coefficients[15] = 2592 { 2593 const_float64( 0x3ff0000000000000ll ), /* 1 */ 2594 const_float64( 0x3fe0000000000000ll ), /* 2 */ 2595 const_float64( 0x3fc5555555555555ll ), /* 3 */ 2596 const_float64( 0x3fa5555555555555ll ), /* 4 */ 2597 const_float64( 0x3f81111111111111ll ), /* 5 */ 2598 const_float64( 0x3f56c16c16c16c17ll ), /* 6 */ 2599 const_float64( 0x3f2a01a01a01a01all ), /* 7 */ 2600 const_float64( 0x3efa01a01a01a01all ), /* 8 */ 2601 const_float64( 0x3ec71de3a556c734ll ), /* 9 */ 2602 const_float64( 0x3e927e4fb7789f5cll ), /* 10 */ 2603 const_float64( 0x3e5ae64567f544e4ll ), /* 11 */ 2604 const_float64( 0x3e21eed8eff8d898ll ), /* 12 */ 2605 const_float64( 0x3de6124613a86d09ll ), /* 13 */ 2606 const_float64( 0x3da93974a8c07c9dll ), /* 14 */ 2607 const_float64( 0x3d6ae7f3e733b81fll ), /* 15 */ 2608 }; 2609 2610 float32 float32_exp2( float32 a STATUS_PARAM ) 2611 { 2612 flag aSign; 2613 int_fast16_t aExp; 2614 uint32_t aSig; 2615 float64 r, x, xn; 2616 int i; 2617 a = float32_squash_input_denormal(a STATUS_VAR); 2618 2619 aSig = extractFloat32Frac( a ); 2620 aExp = extractFloat32Exp( a ); 2621 aSign = extractFloat32Sign( a ); 2622 2623 if ( aExp == 0xFF) { 2624 if ( aSig ) return propagateFloat32NaN( a, float32_zero STATUS_VAR ); 2625 return (aSign) ? float32_zero : a; 2626 } 2627 if (aExp == 0) { 2628 if (aSig == 0) return float32_one; 2629 } 2630 2631 float_raise( float_flag_inexact STATUS_VAR); 2632 2633 /* ******************************* */ 2634 /* using float64 for approximation */ 2635 /* ******************************* */ 2636 x = float32_to_float64(a STATUS_VAR); 2637 x = float64_mul(x, float64_ln2 STATUS_VAR); 2638 2639 xn = x; 2640 r = float64_one; 2641 for (i = 0 ; i < 15 ; i++) { 2642 float64 f; 2643 2644 f = float64_mul(xn, float32_exp2_coefficients[i] STATUS_VAR); 2645 r = float64_add(r, f STATUS_VAR); 2646 2647 xn = float64_mul(xn, x STATUS_VAR); 2648 } 2649 2650 return float64_to_float32(r, status); 2651 } 2652 2653 /*---------------------------------------------------------------------------- 2654 | Returns the binary log of the single-precision floating-point value `a'. 2655 | The operation is performed according to the IEC/IEEE Standard for Binary 2656 | Floating-Point Arithmetic. 2657 *----------------------------------------------------------------------------*/ 2658 float32 float32_log2( float32 a STATUS_PARAM ) 2659 { 2660 flag aSign, zSign; 2661 int_fast16_t aExp; 2662 uint32_t aSig, zSig, i; 2663 2664 a = float32_squash_input_denormal(a STATUS_VAR); 2665 aSig = extractFloat32Frac( a ); 2666 aExp = extractFloat32Exp( a ); 2667 aSign = extractFloat32Sign( a ); 2668 2669 if ( aExp == 0 ) { 2670 if ( aSig == 0 ) return packFloat32( 1, 0xFF, 0 ); 2671 normalizeFloat32Subnormal( aSig, &aExp, &aSig ); 2672 } 2673 if ( aSign ) { 2674 float_raise( float_flag_invalid STATUS_VAR); 2675 return float32_default_nan; 2676 } 2677 if ( aExp == 0xFF ) { 2678 if ( aSig ) return propagateFloat32NaN( a, float32_zero STATUS_VAR ); 2679 return a; 2680 } 2681 2682 aExp -= 0x7F; 2683 aSig |= 0x00800000; 2684 zSign = aExp < 0; 2685 zSig = aExp << 23; 2686 2687 for (i = 1 << 22; i > 0; i >>= 1) { 2688 aSig = ( (uint64_t)aSig * aSig ) >> 23; 2689 if ( aSig & 0x01000000 ) { 2690 aSig >>= 1; 2691 zSig |= i; 2692 } 2693 } 2694 2695 if ( zSign ) 2696 zSig = -zSig; 2697 2698 return normalizeRoundAndPackFloat32( zSign, 0x85, zSig STATUS_VAR ); 2699 } 2700 2701 /*---------------------------------------------------------------------------- 2702 | Returns 1 if the single-precision floating-point value `a' is equal to 2703 | the corresponding value `b', and 0 otherwise. The invalid exception is 2704 | raised if either operand is a NaN. Otherwise, the comparison is performed 2705 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 2706 *----------------------------------------------------------------------------*/ 2707 2708 int float32_eq( float32 a, float32 b STATUS_PARAM ) 2709 { 2710 uint32_t av, bv; 2711 a = float32_squash_input_denormal(a STATUS_VAR); 2712 b = float32_squash_input_denormal(b STATUS_VAR); 2713 2714 if ( ( ( extractFloat32Exp( a ) == 0xFF ) && extractFloat32Frac( a ) ) 2715 || ( ( extractFloat32Exp( b ) == 0xFF ) && extractFloat32Frac( b ) ) 2716 ) { 2717 float_raise( float_flag_invalid STATUS_VAR); 2718 return 0; 2719 } 2720 av = float32_val(a); 2721 bv = float32_val(b); 2722 return ( av == bv ) || ( (uint32_t) ( ( av | bv )<<1 ) == 0 ); 2723 } 2724 2725 /*---------------------------------------------------------------------------- 2726 | Returns 1 if the single-precision floating-point value `a' is less than 2727 | or equal to the corresponding value `b', and 0 otherwise. The invalid 2728 | exception is raised if either operand is a NaN. The comparison is performed 2729 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 2730 *----------------------------------------------------------------------------*/ 2731 2732 int float32_le( float32 a, float32 b STATUS_PARAM ) 2733 { 2734 flag aSign, bSign; 2735 uint32_t av, bv; 2736 a = float32_squash_input_denormal(a STATUS_VAR); 2737 b = float32_squash_input_denormal(b STATUS_VAR); 2738 2739 if ( ( ( extractFloat32Exp( a ) == 0xFF ) && extractFloat32Frac( a ) ) 2740 || ( ( extractFloat32Exp( b ) == 0xFF ) && extractFloat32Frac( b ) ) 2741 ) { 2742 float_raise( float_flag_invalid STATUS_VAR); 2743 return 0; 2744 } 2745 aSign = extractFloat32Sign( a ); 2746 bSign = extractFloat32Sign( b ); 2747 av = float32_val(a); 2748 bv = float32_val(b); 2749 if ( aSign != bSign ) return aSign || ( (uint32_t) ( ( av | bv )<<1 ) == 0 ); 2750 return ( av == bv ) || ( aSign ^ ( av < bv ) ); 2751 2752 } 2753 2754 /*---------------------------------------------------------------------------- 2755 | Returns 1 if the single-precision floating-point value `a' is less than 2756 | the corresponding value `b', and 0 otherwise. The invalid exception is 2757 | raised if either operand is a NaN. The comparison is performed according 2758 | to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 2759 *----------------------------------------------------------------------------*/ 2760 2761 int float32_lt( float32 a, float32 b STATUS_PARAM ) 2762 { 2763 flag aSign, bSign; 2764 uint32_t av, bv; 2765 a = float32_squash_input_denormal(a STATUS_VAR); 2766 b = float32_squash_input_denormal(b STATUS_VAR); 2767 2768 if ( ( ( extractFloat32Exp( a ) == 0xFF ) && extractFloat32Frac( a ) ) 2769 || ( ( extractFloat32Exp( b ) == 0xFF ) && extractFloat32Frac( b ) ) 2770 ) { 2771 float_raise( float_flag_invalid STATUS_VAR); 2772 return 0; 2773 } 2774 aSign = extractFloat32Sign( a ); 2775 bSign = extractFloat32Sign( b ); 2776 av = float32_val(a); 2777 bv = float32_val(b); 2778 if ( aSign != bSign ) return aSign && ( (uint32_t) ( ( av | bv )<<1 ) != 0 ); 2779 return ( av != bv ) && ( aSign ^ ( av < bv ) ); 2780 2781 } 2782 2783 /*---------------------------------------------------------------------------- 2784 | Returns 1 if the single-precision floating-point values `a' and `b' cannot 2785 | be compared, and 0 otherwise. The invalid exception is raised if either 2786 | operand is a NaN. The comparison is performed according to the IEC/IEEE 2787 | Standard for Binary Floating-Point Arithmetic. 2788 *----------------------------------------------------------------------------*/ 2789 2790 int float32_unordered( float32 a, float32 b STATUS_PARAM ) 2791 { 2792 a = float32_squash_input_denormal(a STATUS_VAR); 2793 b = float32_squash_input_denormal(b STATUS_VAR); 2794 2795 if ( ( ( extractFloat32Exp( a ) == 0xFF ) && extractFloat32Frac( a ) ) 2796 || ( ( extractFloat32Exp( b ) == 0xFF ) && extractFloat32Frac( b ) ) 2797 ) { 2798 float_raise( float_flag_invalid STATUS_VAR); 2799 return 1; 2800 } 2801 return 0; 2802 } 2803 2804 /*---------------------------------------------------------------------------- 2805 | Returns 1 if the single-precision floating-point value `a' is equal to 2806 | the corresponding value `b', and 0 otherwise. Quiet NaNs do not cause an 2807 | exception. The comparison is performed according to the IEC/IEEE Standard 2808 | for Binary Floating-Point Arithmetic. 2809 *----------------------------------------------------------------------------*/ 2810 2811 int float32_eq_quiet( float32 a, float32 b STATUS_PARAM ) 2812 { 2813 a = float32_squash_input_denormal(a STATUS_VAR); 2814 b = float32_squash_input_denormal(b STATUS_VAR); 2815 2816 if ( ( ( extractFloat32Exp( a ) == 0xFF ) && extractFloat32Frac( a ) ) 2817 || ( ( extractFloat32Exp( b ) == 0xFF ) && extractFloat32Frac( b ) ) 2818 ) { 2819 if ( float32_is_signaling_nan( a ) || float32_is_signaling_nan( b ) ) { 2820 float_raise( float_flag_invalid STATUS_VAR); 2821 } 2822 return 0; 2823 } 2824 return ( float32_val(a) == float32_val(b) ) || 2825 ( (uint32_t) ( ( float32_val(a) | float32_val(b) )<<1 ) == 0 ); 2826 } 2827 2828 /*---------------------------------------------------------------------------- 2829 | Returns 1 if the single-precision floating-point value `a' is less than or 2830 | equal to the corresponding value `b', and 0 otherwise. Quiet NaNs do not 2831 | cause an exception. Otherwise, the comparison is performed according to the 2832 | IEC/IEEE Standard for Binary Floating-Point Arithmetic. 2833 *----------------------------------------------------------------------------*/ 2834 2835 int float32_le_quiet( float32 a, float32 b STATUS_PARAM ) 2836 { 2837 flag aSign, bSign; 2838 uint32_t av, bv; 2839 a = float32_squash_input_denormal(a STATUS_VAR); 2840 b = float32_squash_input_denormal(b STATUS_VAR); 2841 2842 if ( ( ( extractFloat32Exp( a ) == 0xFF ) && extractFloat32Frac( a ) ) 2843 || ( ( extractFloat32Exp( b ) == 0xFF ) && extractFloat32Frac( b ) ) 2844 ) { 2845 if ( float32_is_signaling_nan( a ) || float32_is_signaling_nan( b ) ) { 2846 float_raise( float_flag_invalid STATUS_VAR); 2847 } 2848 return 0; 2849 } 2850 aSign = extractFloat32Sign( a ); 2851 bSign = extractFloat32Sign( b ); 2852 av = float32_val(a); 2853 bv = float32_val(b); 2854 if ( aSign != bSign ) return aSign || ( (uint32_t) ( ( av | bv )<<1 ) == 0 ); 2855 return ( av == bv ) || ( aSign ^ ( av < bv ) ); 2856 2857 } 2858 2859 /*---------------------------------------------------------------------------- 2860 | Returns 1 if the single-precision floating-point value `a' is less than 2861 | the corresponding value `b', and 0 otherwise. Quiet NaNs do not cause an 2862 | exception. Otherwise, the comparison is performed according to the IEC/IEEE 2863 | Standard for Binary Floating-Point Arithmetic. 2864 *----------------------------------------------------------------------------*/ 2865 2866 int float32_lt_quiet( float32 a, float32 b STATUS_PARAM ) 2867 { 2868 flag aSign, bSign; 2869 uint32_t av, bv; 2870 a = float32_squash_input_denormal(a STATUS_VAR); 2871 b = float32_squash_input_denormal(b STATUS_VAR); 2872 2873 if ( ( ( extractFloat32Exp( a ) == 0xFF ) && extractFloat32Frac( a ) ) 2874 || ( ( extractFloat32Exp( b ) == 0xFF ) && extractFloat32Frac( b ) ) 2875 ) { 2876 if ( float32_is_signaling_nan( a ) || float32_is_signaling_nan( b ) ) { 2877 float_raise( float_flag_invalid STATUS_VAR); 2878 } 2879 return 0; 2880 } 2881 aSign = extractFloat32Sign( a ); 2882 bSign = extractFloat32Sign( b ); 2883 av = float32_val(a); 2884 bv = float32_val(b); 2885 if ( aSign != bSign ) return aSign && ( (uint32_t) ( ( av | bv )<<1 ) != 0 ); 2886 return ( av != bv ) && ( aSign ^ ( av < bv ) ); 2887 2888 } 2889 2890 /*---------------------------------------------------------------------------- 2891 | Returns 1 if the single-precision floating-point values `a' and `b' cannot 2892 | be compared, and 0 otherwise. Quiet NaNs do not cause an exception. The 2893 | comparison is performed according to the IEC/IEEE Standard for Binary 2894 | Floating-Point Arithmetic. 2895 *----------------------------------------------------------------------------*/ 2896 2897 int float32_unordered_quiet( float32 a, float32 b STATUS_PARAM ) 2898 { 2899 a = float32_squash_input_denormal(a STATUS_VAR); 2900 b = float32_squash_input_denormal(b STATUS_VAR); 2901 2902 if ( ( ( extractFloat32Exp( a ) == 0xFF ) && extractFloat32Frac( a ) ) 2903 || ( ( extractFloat32Exp( b ) == 0xFF ) && extractFloat32Frac( b ) ) 2904 ) { 2905 if ( float32_is_signaling_nan( a ) || float32_is_signaling_nan( b ) ) { 2906 float_raise( float_flag_invalid STATUS_VAR); 2907 } 2908 return 1; 2909 } 2910 return 0; 2911 } 2912 2913 /*---------------------------------------------------------------------------- 2914 | Returns the result of converting the double-precision floating-point value 2915 | `a' to the 32-bit two's complement integer format. The conversion is 2916 | performed according to the IEC/IEEE Standard for Binary Floating-Point 2917 | Arithmetic---which means in particular that the conversion is rounded 2918 | according to the current rounding mode. If `a' is a NaN, the largest 2919 | positive integer is returned. Otherwise, if the conversion overflows, the 2920 | largest integer with the same sign as `a' is returned. 2921 *----------------------------------------------------------------------------*/ 2922 2923 int32 float64_to_int32( float64 a STATUS_PARAM ) 2924 { 2925 flag aSign; 2926 int_fast16_t aExp, shiftCount; 2927 uint64_t aSig; 2928 a = float64_squash_input_denormal(a STATUS_VAR); 2929 2930 aSig = extractFloat64Frac( a ); 2931 aExp = extractFloat64Exp( a ); 2932 aSign = extractFloat64Sign( a ); 2933 if ( ( aExp == 0x7FF ) && aSig ) aSign = 0; 2934 if ( aExp ) aSig |= LIT64( 0x0010000000000000 ); 2935 shiftCount = 0x42C - aExp; 2936 if ( 0 < shiftCount ) shift64RightJamming( aSig, shiftCount, &aSig ); 2937 return roundAndPackInt32( aSign, aSig STATUS_VAR ); 2938 2939 } 2940 2941 /*---------------------------------------------------------------------------- 2942 | Returns the result of converting the double-precision floating-point value 2943 | `a' to the 32-bit two's complement integer format. The conversion is 2944 | performed according to the IEC/IEEE Standard for Binary Floating-Point 2945 | Arithmetic, except that the conversion is always rounded toward zero. 2946 | If `a' is a NaN, the largest positive integer is returned. Otherwise, if 2947 | the conversion overflows, the largest integer with the same sign as `a' is 2948 | returned. 2949 *----------------------------------------------------------------------------*/ 2950 2951 int32 float64_to_int32_round_to_zero( float64 a STATUS_PARAM ) 2952 { 2953 flag aSign; 2954 int_fast16_t aExp, shiftCount; 2955 uint64_t aSig, savedASig; 2956 int32_t z; 2957 a = float64_squash_input_denormal(a STATUS_VAR); 2958 2959 aSig = extractFloat64Frac( a ); 2960 aExp = extractFloat64Exp( a ); 2961 aSign = extractFloat64Sign( a ); 2962 if ( 0x41E < aExp ) { 2963 if ( ( aExp == 0x7FF ) && aSig ) aSign = 0; 2964 goto invalid; 2965 } 2966 else if ( aExp < 0x3FF ) { 2967 if ( aExp || aSig ) STATUS(float_exception_flags) |= float_flag_inexact; 2968 return 0; 2969 } 2970 aSig |= LIT64( 0x0010000000000000 ); 2971 shiftCount = 0x433 - aExp; 2972 savedASig = aSig; 2973 aSig >>= shiftCount; 2974 z = aSig; 2975 if ( aSign ) z = - z; 2976 if ( ( z < 0 ) ^ aSign ) { 2977 invalid: 2978 float_raise( float_flag_invalid STATUS_VAR); 2979 return aSign ? (int32_t) 0x80000000 : 0x7FFFFFFF; 2980 } 2981 if ( ( aSig<<shiftCount ) != savedASig ) { 2982 STATUS(float_exception_flags) |= float_flag_inexact; 2983 } 2984 return z; 2985 2986 } 2987 2988 /*---------------------------------------------------------------------------- 2989 | Returns the result of converting the double-precision floating-point value 2990 | `a' to the 16-bit two's complement integer format. The conversion is 2991 | performed according to the IEC/IEEE Standard for Binary Floating-Point 2992 | Arithmetic, except that the conversion is always rounded toward zero. 2993 | If `a' is a NaN, the largest positive integer is returned. Otherwise, if 2994 | the conversion overflows, the largest integer with the same sign as `a' is 2995 | returned. 2996 *----------------------------------------------------------------------------*/ 2997 2998 int_fast16_t float64_to_int16_round_to_zero(float64 a STATUS_PARAM) 2999 { 3000 flag aSign; 3001 int_fast16_t aExp, shiftCount; 3002 uint64_t aSig, savedASig; 3003 int32 z; 3004 3005 aSig = extractFloat64Frac( a ); 3006 aExp = extractFloat64Exp( a ); 3007 aSign = extractFloat64Sign( a ); 3008 if ( 0x40E < aExp ) { 3009 if ( ( aExp == 0x7FF ) && aSig ) { 3010 aSign = 0; 3011 } 3012 goto invalid; 3013 } 3014 else if ( aExp < 0x3FF ) { 3015 if ( aExp || aSig ) { 3016 STATUS(float_exception_flags) |= float_flag_inexact; 3017 } 3018 return 0; 3019 } 3020 aSig |= LIT64( 0x0010000000000000 ); 3021 shiftCount = 0x433 - aExp; 3022 savedASig = aSig; 3023 aSig >>= shiftCount; 3024 z = aSig; 3025 if ( aSign ) { 3026 z = - z; 3027 } 3028 if ( ( (int16_t)z < 0 ) ^ aSign ) { 3029 invalid: 3030 float_raise( float_flag_invalid STATUS_VAR); 3031 return aSign ? (int32_t) 0xffff8000 : 0x7FFF; 3032 } 3033 if ( ( aSig<<shiftCount ) != savedASig ) { 3034 STATUS(float_exception_flags) |= float_flag_inexact; 3035 } 3036 return z; 3037 } 3038 3039 /*---------------------------------------------------------------------------- 3040 | Returns the result of converting the double-precision floating-point value 3041 | `a' to the 64-bit two's complement integer format. The conversion is 3042 | performed according to the IEC/IEEE Standard for Binary Floating-Point 3043 | Arithmetic---which means in particular that the conversion is rounded 3044 | according to the current rounding mode. If `a' is a NaN, the largest 3045 | positive integer is returned. Otherwise, if the conversion overflows, the 3046 | largest integer with the same sign as `a' is returned. 3047 *----------------------------------------------------------------------------*/ 3048 3049 int64 float64_to_int64( float64 a STATUS_PARAM ) 3050 { 3051 flag aSign; 3052 int_fast16_t aExp, shiftCount; 3053 uint64_t aSig, aSigExtra; 3054 a = float64_squash_input_denormal(a STATUS_VAR); 3055 3056 aSig = extractFloat64Frac( a ); 3057 aExp = extractFloat64Exp( a ); 3058 aSign = extractFloat64Sign( a ); 3059 if ( aExp ) aSig |= LIT64( 0x0010000000000000 ); 3060 shiftCount = 0x433 - aExp; 3061 if ( shiftCount <= 0 ) { 3062 if ( 0x43E < aExp ) { 3063 float_raise( float_flag_invalid STATUS_VAR); 3064 if ( ! aSign 3065 || ( ( aExp == 0x7FF ) 3066 && ( aSig != LIT64( 0x0010000000000000 ) ) ) 3067 ) { 3068 return LIT64( 0x7FFFFFFFFFFFFFFF ); 3069 } 3070 return (int64_t) LIT64( 0x8000000000000000 ); 3071 } 3072 aSigExtra = 0; 3073 aSig <<= - shiftCount; 3074 } 3075 else { 3076 shift64ExtraRightJamming( aSig, 0, shiftCount, &aSig, &aSigExtra ); 3077 } 3078 return roundAndPackInt64( aSign, aSig, aSigExtra STATUS_VAR ); 3079 3080 } 3081 3082 /*---------------------------------------------------------------------------- 3083 | Returns the result of converting the double-precision floating-point value 3084 | `a' to the 64-bit two's complement integer format. The conversion is 3085 | performed according to the IEC/IEEE Standard for Binary Floating-Point 3086 | Arithmetic, except that the conversion is always rounded toward zero. 3087 | If `a' is a NaN, the largest positive integer is returned. Otherwise, if 3088 | the conversion overflows, the largest integer with the same sign as `a' is 3089 | returned. 3090 *----------------------------------------------------------------------------*/ 3091 3092 int64 float64_to_int64_round_to_zero( float64 a STATUS_PARAM ) 3093 { 3094 flag aSign; 3095 int_fast16_t aExp, shiftCount; 3096 uint64_t aSig; 3097 int64 z; 3098 a = float64_squash_input_denormal(a STATUS_VAR); 3099 3100 aSig = extractFloat64Frac( a ); 3101 aExp = extractFloat64Exp( a ); 3102 aSign = extractFloat64Sign( a ); 3103 if ( aExp ) aSig |= LIT64( 0x0010000000000000 ); 3104 shiftCount = aExp - 0x433; 3105 if ( 0 <= shiftCount ) { 3106 if ( 0x43E <= aExp ) { 3107 if ( float64_val(a) != LIT64( 0xC3E0000000000000 ) ) { 3108 float_raise( float_flag_invalid STATUS_VAR); 3109 if ( ! aSign 3110 || ( ( aExp == 0x7FF ) 3111 && ( aSig != LIT64( 0x0010000000000000 ) ) ) 3112 ) { 3113 return LIT64( 0x7FFFFFFFFFFFFFFF ); 3114 } 3115 } 3116 return (int64_t) LIT64( 0x8000000000000000 ); 3117 } 3118 z = aSig<<shiftCount; 3119 } 3120 else { 3121 if ( aExp < 0x3FE ) { 3122 if ( aExp | aSig ) STATUS(float_exception_flags) |= float_flag_inexact; 3123 return 0; 3124 } 3125 z = aSig>>( - shiftCount ); 3126 if ( (uint64_t) ( aSig<<( shiftCount & 63 ) ) ) { 3127 STATUS(float_exception_flags) |= float_flag_inexact; 3128 } 3129 } 3130 if ( aSign ) z = - z; 3131 return z; 3132 3133 } 3134 3135 /*---------------------------------------------------------------------------- 3136 | Returns the result of converting the double-precision floating-point value 3137 | `a' to the single-precision floating-point format. The conversion is 3138 | performed according to the IEC/IEEE Standard for Binary Floating-Point 3139 | Arithmetic. 3140 *----------------------------------------------------------------------------*/ 3141 3142 float32 float64_to_float32( float64 a STATUS_PARAM ) 3143 { 3144 flag aSign; 3145 int_fast16_t aExp; 3146 uint64_t aSig; 3147 uint32_t zSig; 3148 a = float64_squash_input_denormal(a STATUS_VAR); 3149 3150 aSig = extractFloat64Frac( a ); 3151 aExp = extractFloat64Exp( a ); 3152 aSign = extractFloat64Sign( a ); 3153 if ( aExp == 0x7FF ) { 3154 if ( aSig ) return commonNaNToFloat32( float64ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 3155 return packFloat32( aSign, 0xFF, 0 ); 3156 } 3157 shift64RightJamming( aSig, 22, &aSig ); 3158 zSig = aSig; 3159 if ( aExp || zSig ) { 3160 zSig |= 0x40000000; 3161 aExp -= 0x381; 3162 } 3163 return roundAndPackFloat32( aSign, aExp, zSig STATUS_VAR ); 3164 3165 } 3166 3167 3168 /*---------------------------------------------------------------------------- 3169 | Packs the sign `zSign', exponent `zExp', and significand `zSig' into a 3170 | half-precision floating-point value, returning the result. After being 3171 | shifted into the proper positions, the three fields are simply added 3172 | together to form the result. This means that any integer portion of `zSig' 3173 | will be added into the exponent. Since a properly normalized significand 3174 | will have an integer portion equal to 1, the `zExp' input should be 1 less 3175 | than the desired result exponent whenever `zSig' is a complete, normalized 3176 | significand. 3177 *----------------------------------------------------------------------------*/ 3178 static float16 packFloat16(flag zSign, int_fast16_t zExp, uint16_t zSig) 3179 { 3180 return make_float16( 3181 (((uint32_t)zSign) << 15) + (((uint32_t)zExp) << 10) + zSig); 3182 } 3183 3184 /*---------------------------------------------------------------------------- 3185 | Takes an abstract floating-point value having sign `zSign', exponent `zExp', 3186 | and significand `zSig', and returns the proper half-precision floating- 3187 | point value corresponding to the abstract input. Ordinarily, the abstract 3188 | value is simply rounded and packed into the half-precision format, with 3189 | the inexact exception raised if the abstract input cannot be represented 3190 | exactly. However, if the abstract value is too large, the overflow and 3191 | inexact exceptions are raised and an infinity or maximal finite value is 3192 | returned. If the abstract value is too small, the input value is rounded to 3193 | a subnormal number, and the underflow and inexact exceptions are raised if 3194 | the abstract input cannot be represented exactly as a subnormal half- 3195 | precision floating-point number. 3196 | The `ieee' flag indicates whether to use IEEE standard half precision, or 3197 | ARM-style "alternative representation", which omits the NaN and Inf 3198 | encodings in order to raise the maximum representable exponent by one. 3199 | The input significand `zSig' has its binary point between bits 22 3200 | and 23, which is 13 bits to the left of the usual location. This shifted 3201 | significand must be normalized or smaller. If `zSig' is not normalized, 3202 | `zExp' must be 0; in that case, the result returned is a subnormal number, 3203 | and it must not require rounding. In the usual case that `zSig' is 3204 | normalized, `zExp' must be 1 less than the ``true'' floating-point exponent. 3205 | Note the slightly odd position of the binary point in zSig compared with the 3206 | other roundAndPackFloat functions. This should probably be fixed if we 3207 | need to implement more float16 routines than just conversion. 3208 | The handling of underflow and overflow follows the IEC/IEEE Standard for 3209 | Binary Floating-Point Arithmetic. 3210 *----------------------------------------------------------------------------*/ 3211 3212 static float32 roundAndPackFloat16(flag zSign, int_fast16_t zExp, 3213 uint32_t zSig, flag ieee STATUS_PARAM) 3214 { 3215 int maxexp = ieee ? 29 : 30; 3216 uint32_t mask; 3217 uint32_t increment; 3218 bool rounding_bumps_exp; 3219 bool is_tiny = false; 3220 3221 /* Calculate the mask of bits of the mantissa which are not 3222 * representable in half-precision and will be lost. 3223 */ 3224 if (zExp < 1) { 3225 /* Will be denormal in halfprec */ 3226 mask = 0x00ffffff; 3227 if (zExp >= -11) { 3228 mask >>= 11 + zExp; 3229 } 3230 } else { 3231 /* Normal number in halfprec */ 3232 mask = 0x00001fff; 3233 } 3234 3235 switch (STATUS(float_rounding_mode)) { 3236 case float_round_nearest_even: 3237 increment = (mask + 1) >> 1; 3238 if ((zSig & mask) == increment) { 3239 increment = zSig & (increment << 1); 3240 } 3241 break; 3242 case float_round_ties_away: 3243 increment = (mask + 1) >> 1; 3244 break; 3245 case float_round_up: 3246 increment = zSign ? 0 : mask; 3247 break; 3248 case float_round_down: 3249 increment = zSign ? mask : 0; 3250 break; 3251 default: /* round_to_zero */ 3252 increment = 0; 3253 break; 3254 } 3255 3256 rounding_bumps_exp = (zSig + increment >= 0x01000000); 3257 3258 if (zExp > maxexp || (zExp == maxexp && rounding_bumps_exp)) { 3259 if (ieee) { 3260 float_raise(float_flag_overflow | float_flag_inexact STATUS_VAR); 3261 return packFloat16(zSign, 0x1f, 0); 3262 } else { 3263 float_raise(float_flag_invalid STATUS_VAR); 3264 return packFloat16(zSign, 0x1f, 0x3ff); 3265 } 3266 } 3267 3268 if (zExp < 0) { 3269 /* Note that flush-to-zero does not affect half-precision results */ 3270 is_tiny = 3271 (STATUS(float_detect_tininess) == float_tininess_before_rounding) 3272 || (zExp < -1) 3273 || (!rounding_bumps_exp); 3274 } 3275 if (zSig & mask) { 3276 float_raise(float_flag_inexact STATUS_VAR); 3277 if (is_tiny) { 3278 float_raise(float_flag_underflow STATUS_VAR); 3279 } 3280 } 3281 3282 zSig += increment; 3283 if (rounding_bumps_exp) { 3284 zSig >>= 1; 3285 zExp++; 3286 } 3287 3288 if (zExp < -10) { 3289 return packFloat16(zSign, 0, 0); 3290 } 3291 if (zExp < 0) { 3292 zSig >>= -zExp; 3293 zExp = 0; 3294 } 3295 return packFloat16(zSign, zExp, zSig >> 13); 3296 } 3297 3298 static void normalizeFloat16Subnormal(uint32_t aSig, int_fast16_t *zExpPtr, 3299 uint32_t *zSigPtr) 3300 { 3301 int8_t shiftCount = countLeadingZeros32(aSig) - 21; 3302 *zSigPtr = aSig << shiftCount; 3303 *zExpPtr = 1 - shiftCount; 3304 } 3305 3306 /* Half precision floats come in two formats: standard IEEE and "ARM" format. 3307 The latter gains extra exponent range by omitting the NaN/Inf encodings. */ 3308 3309 float32 float16_to_float32(float16 a, flag ieee STATUS_PARAM) 3310 { 3311 flag aSign; 3312 int_fast16_t aExp; 3313 uint32_t aSig; 3314 3315 aSign = extractFloat16Sign(a); 3316 aExp = extractFloat16Exp(a); 3317 aSig = extractFloat16Frac(a); 3318 3319 if (aExp == 0x1f && ieee) { 3320 if (aSig) { 3321 return commonNaNToFloat32(float16ToCommonNaN(a STATUS_VAR) STATUS_VAR); 3322 } 3323 return packFloat32(aSign, 0xff, 0); 3324 } 3325 if (aExp == 0) { 3326 if (aSig == 0) { 3327 return packFloat32(aSign, 0, 0); 3328 } 3329 3330 normalizeFloat16Subnormal(aSig, &aExp, &aSig); 3331 aExp--; 3332 } 3333 return packFloat32( aSign, aExp + 0x70, aSig << 13); 3334 } 3335 3336 float16 float32_to_float16(float32 a, flag ieee STATUS_PARAM) 3337 { 3338 flag aSign; 3339 int_fast16_t aExp; 3340 uint32_t aSig; 3341 3342 a = float32_squash_input_denormal(a STATUS_VAR); 3343 3344 aSig = extractFloat32Frac( a ); 3345 aExp = extractFloat32Exp( a ); 3346 aSign = extractFloat32Sign( a ); 3347 if ( aExp == 0xFF ) { 3348 if (aSig) { 3349 /* Input is a NaN */ 3350 if (!ieee) { 3351 float_raise(float_flag_invalid STATUS_VAR); 3352 return packFloat16(aSign, 0, 0); 3353 } 3354 return commonNaNToFloat16( 3355 float32ToCommonNaN(a STATUS_VAR) STATUS_VAR); 3356 } 3357 /* Infinity */ 3358 if (!ieee) { 3359 float_raise(float_flag_invalid STATUS_VAR); 3360 return packFloat16(aSign, 0x1f, 0x3ff); 3361 } 3362 return packFloat16(aSign, 0x1f, 0); 3363 } 3364 if (aExp == 0 && aSig == 0) { 3365 return packFloat16(aSign, 0, 0); 3366 } 3367 /* Decimal point between bits 22 and 23. Note that we add the 1 bit 3368 * even if the input is denormal; however this is harmless because 3369 * the largest possible single-precision denormal is still smaller 3370 * than the smallest representable half-precision denormal, and so we 3371 * will end up ignoring aSig and returning via the "always return zero" 3372 * codepath. 3373 */ 3374 aSig |= 0x00800000; 3375 aExp -= 0x71; 3376 3377 return roundAndPackFloat16(aSign, aExp, aSig, ieee STATUS_VAR); 3378 } 3379 3380 float64 float16_to_float64(float16 a, flag ieee STATUS_PARAM) 3381 { 3382 flag aSign; 3383 int_fast16_t aExp; 3384 uint32_t aSig; 3385 3386 aSign = extractFloat16Sign(a); 3387 aExp = extractFloat16Exp(a); 3388 aSig = extractFloat16Frac(a); 3389 3390 if (aExp == 0x1f && ieee) { 3391 if (aSig) { 3392 return commonNaNToFloat64( 3393 float16ToCommonNaN(a STATUS_VAR) STATUS_VAR); 3394 } 3395 return packFloat64(aSign, 0x7ff, 0); 3396 } 3397 if (aExp == 0) { 3398 if (aSig == 0) { 3399 return packFloat64(aSign, 0, 0); 3400 } 3401 3402 normalizeFloat16Subnormal(aSig, &aExp, &aSig); 3403 aExp--; 3404 } 3405 return packFloat64(aSign, aExp + 0x3f0, ((uint64_t)aSig) << 42); 3406 } 3407 3408 float16 float64_to_float16(float64 a, flag ieee STATUS_PARAM) 3409 { 3410 flag aSign; 3411 int_fast16_t aExp; 3412 uint64_t aSig; 3413 uint32_t zSig; 3414 3415 a = float64_squash_input_denormal(a STATUS_VAR); 3416 3417 aSig = extractFloat64Frac(a); 3418 aExp = extractFloat64Exp(a); 3419 aSign = extractFloat64Sign(a); 3420 if (aExp == 0x7FF) { 3421 if (aSig) { 3422 /* Input is a NaN */ 3423 if (!ieee) { 3424 float_raise(float_flag_invalid STATUS_VAR); 3425 return packFloat16(aSign, 0, 0); 3426 } 3427 return commonNaNToFloat16( 3428 float64ToCommonNaN(a STATUS_VAR) STATUS_VAR); 3429 } 3430 /* Infinity */ 3431 if (!ieee) { 3432 float_raise(float_flag_invalid STATUS_VAR); 3433 return packFloat16(aSign, 0x1f, 0x3ff); 3434 } 3435 return packFloat16(aSign, 0x1f, 0); 3436 } 3437 shift64RightJamming(aSig, 29, &aSig); 3438 zSig = aSig; 3439 if (aExp == 0 && zSig == 0) { 3440 return packFloat16(aSign, 0, 0); 3441 } 3442 /* Decimal point between bits 22 and 23. Note that we add the 1 bit 3443 * even if the input is denormal; however this is harmless because 3444 * the largest possible single-precision denormal is still smaller 3445 * than the smallest representable half-precision denormal, and so we 3446 * will end up ignoring aSig and returning via the "always return zero" 3447 * codepath. 3448 */ 3449 zSig |= 0x00800000; 3450 aExp -= 0x3F1; 3451 3452 return roundAndPackFloat16(aSign, aExp, zSig, ieee STATUS_VAR); 3453 } 3454 3455 /*---------------------------------------------------------------------------- 3456 | Returns the result of converting the double-precision floating-point value 3457 | `a' to the extended double-precision floating-point format. The conversion 3458 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 3459 | Arithmetic. 3460 *----------------------------------------------------------------------------*/ 3461 3462 floatx80 float64_to_floatx80( float64 a STATUS_PARAM ) 3463 { 3464 flag aSign; 3465 int_fast16_t aExp; 3466 uint64_t aSig; 3467 3468 a = float64_squash_input_denormal(a STATUS_VAR); 3469 aSig = extractFloat64Frac( a ); 3470 aExp = extractFloat64Exp( a ); 3471 aSign = extractFloat64Sign( a ); 3472 if ( aExp == 0x7FF ) { 3473 if ( aSig ) return commonNaNToFloatx80( float64ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 3474 return packFloatx80( aSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 3475 } 3476 if ( aExp == 0 ) { 3477 if ( aSig == 0 ) return packFloatx80( aSign, 0, 0 ); 3478 normalizeFloat64Subnormal( aSig, &aExp, &aSig ); 3479 } 3480 return 3481 packFloatx80( 3482 aSign, aExp + 0x3C00, ( aSig | LIT64( 0x0010000000000000 ) )<<11 ); 3483 3484 } 3485 3486 /*---------------------------------------------------------------------------- 3487 | Returns the result of converting the double-precision floating-point value 3488 | `a' to the quadruple-precision floating-point format. The conversion is 3489 | performed according to the IEC/IEEE Standard for Binary Floating-Point 3490 | Arithmetic. 3491 *----------------------------------------------------------------------------*/ 3492 3493 float128 float64_to_float128( float64 a STATUS_PARAM ) 3494 { 3495 flag aSign; 3496 int_fast16_t aExp; 3497 uint64_t aSig, zSig0, zSig1; 3498 3499 a = float64_squash_input_denormal(a STATUS_VAR); 3500 aSig = extractFloat64Frac( a ); 3501 aExp = extractFloat64Exp( a ); 3502 aSign = extractFloat64Sign( a ); 3503 if ( aExp == 0x7FF ) { 3504 if ( aSig ) return commonNaNToFloat128( float64ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 3505 return packFloat128( aSign, 0x7FFF, 0, 0 ); 3506 } 3507 if ( aExp == 0 ) { 3508 if ( aSig == 0 ) return packFloat128( aSign, 0, 0, 0 ); 3509 normalizeFloat64Subnormal( aSig, &aExp, &aSig ); 3510 --aExp; 3511 } 3512 shift128Right( aSig, 0, 4, &zSig0, &zSig1 ); 3513 return packFloat128( aSign, aExp + 0x3C00, zSig0, zSig1 ); 3514 3515 } 3516 3517 /*---------------------------------------------------------------------------- 3518 | Rounds the double-precision floating-point value `a' to an integer, and 3519 | returns the result as a double-precision floating-point value. The 3520 | operation is performed according to the IEC/IEEE Standard for Binary 3521 | Floating-Point Arithmetic. 3522 *----------------------------------------------------------------------------*/ 3523 3524 float64 float64_round_to_int( float64 a STATUS_PARAM ) 3525 { 3526 flag aSign; 3527 int_fast16_t aExp; 3528 uint64_t lastBitMask, roundBitsMask; 3529 uint64_t z; 3530 a = float64_squash_input_denormal(a STATUS_VAR); 3531 3532 aExp = extractFloat64Exp( a ); 3533 if ( 0x433 <= aExp ) { 3534 if ( ( aExp == 0x7FF ) && extractFloat64Frac( a ) ) { 3535 return propagateFloat64NaN( a, a STATUS_VAR ); 3536 } 3537 return a; 3538 } 3539 if ( aExp < 0x3FF ) { 3540 if ( (uint64_t) ( float64_val(a)<<1 ) == 0 ) return a; 3541 STATUS(float_exception_flags) |= float_flag_inexact; 3542 aSign = extractFloat64Sign( a ); 3543 switch ( STATUS(float_rounding_mode) ) { 3544 case float_round_nearest_even: 3545 if ( ( aExp == 0x3FE ) && extractFloat64Frac( a ) ) { 3546 return packFloat64( aSign, 0x3FF, 0 ); 3547 } 3548 break; 3549 case float_round_ties_away: 3550 if (aExp == 0x3FE) { 3551 return packFloat64(aSign, 0x3ff, 0); 3552 } 3553 break; 3554 case float_round_down: 3555 return make_float64(aSign ? LIT64( 0xBFF0000000000000 ) : 0); 3556 case float_round_up: 3557 return make_float64( 3558 aSign ? LIT64( 0x8000000000000000 ) : LIT64( 0x3FF0000000000000 )); 3559 } 3560 return packFloat64( aSign, 0, 0 ); 3561 } 3562 lastBitMask = 1; 3563 lastBitMask <<= 0x433 - aExp; 3564 roundBitsMask = lastBitMask - 1; 3565 z = float64_val(a); 3566 switch (STATUS(float_rounding_mode)) { 3567 case float_round_nearest_even: 3568 z += lastBitMask >> 1; 3569 if ((z & roundBitsMask) == 0) { 3570 z &= ~lastBitMask; 3571 } 3572 break; 3573 case float_round_ties_away: 3574 z += lastBitMask >> 1; 3575 break; 3576 case float_round_to_zero: 3577 break; 3578 case float_round_up: 3579 if (!extractFloat64Sign(make_float64(z))) { 3580 z += roundBitsMask; 3581 } 3582 break; 3583 case float_round_down: 3584 if (extractFloat64Sign(make_float64(z))) { 3585 z += roundBitsMask; 3586 } 3587 break; 3588 default: 3589 abort(); 3590 } 3591 z &= ~ roundBitsMask; 3592 if ( z != float64_val(a) ) 3593 STATUS(float_exception_flags) |= float_flag_inexact; 3594 return make_float64(z); 3595 3596 } 3597 3598 float64 float64_trunc_to_int( float64 a STATUS_PARAM) 3599 { 3600 int oldmode; 3601 float64 res; 3602 oldmode = STATUS(float_rounding_mode); 3603 STATUS(float_rounding_mode) = float_round_to_zero; 3604 res = float64_round_to_int(a STATUS_VAR); 3605 STATUS(float_rounding_mode) = oldmode; 3606 return res; 3607 } 3608 3609 /*---------------------------------------------------------------------------- 3610 | Returns the result of adding the absolute values of the double-precision 3611 | floating-point values `a' and `b'. If `zSign' is 1, the sum is negated 3612 | before being returned. `zSign' is ignored if the result is a NaN. 3613 | The addition is performed according to the IEC/IEEE Standard for Binary 3614 | Floating-Point Arithmetic. 3615 *----------------------------------------------------------------------------*/ 3616 3617 static float64 addFloat64Sigs( float64 a, float64 b, flag zSign STATUS_PARAM ) 3618 { 3619 int_fast16_t aExp, bExp, zExp; 3620 uint64_t aSig, bSig, zSig; 3621 int_fast16_t expDiff; 3622 3623 aSig = extractFloat64Frac( a ); 3624 aExp = extractFloat64Exp( a ); 3625 bSig = extractFloat64Frac( b ); 3626 bExp = extractFloat64Exp( b ); 3627 expDiff = aExp - bExp; 3628 aSig <<= 9; 3629 bSig <<= 9; 3630 if ( 0 < expDiff ) { 3631 if ( aExp == 0x7FF ) { 3632 if ( aSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3633 return a; 3634 } 3635 if ( bExp == 0 ) { 3636 --expDiff; 3637 } 3638 else { 3639 bSig |= LIT64( 0x2000000000000000 ); 3640 } 3641 shift64RightJamming( bSig, expDiff, &bSig ); 3642 zExp = aExp; 3643 } 3644 else if ( expDiff < 0 ) { 3645 if ( bExp == 0x7FF ) { 3646 if ( bSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3647 return packFloat64( zSign, 0x7FF, 0 ); 3648 } 3649 if ( aExp == 0 ) { 3650 ++expDiff; 3651 } 3652 else { 3653 aSig |= LIT64( 0x2000000000000000 ); 3654 } 3655 shift64RightJamming( aSig, - expDiff, &aSig ); 3656 zExp = bExp; 3657 } 3658 else { 3659 if ( aExp == 0x7FF ) { 3660 if ( aSig | bSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3661 return a; 3662 } 3663 if ( aExp == 0 ) { 3664 if (STATUS(flush_to_zero)) { 3665 if (aSig | bSig) { 3666 float_raise(float_flag_output_denormal STATUS_VAR); 3667 } 3668 return packFloat64(zSign, 0, 0); 3669 } 3670 return packFloat64( zSign, 0, ( aSig + bSig )>>9 ); 3671 } 3672 zSig = LIT64( 0x4000000000000000 ) + aSig + bSig; 3673 zExp = aExp; 3674 goto roundAndPack; 3675 } 3676 aSig |= LIT64( 0x2000000000000000 ); 3677 zSig = ( aSig + bSig )<<1; 3678 --zExp; 3679 if ( (int64_t) zSig < 0 ) { 3680 zSig = aSig + bSig; 3681 ++zExp; 3682 } 3683 roundAndPack: 3684 return roundAndPackFloat64( zSign, zExp, zSig STATUS_VAR ); 3685 3686 } 3687 3688 /*---------------------------------------------------------------------------- 3689 | Returns the result of subtracting the absolute values of the double- 3690 | precision floating-point values `a' and `b'. If `zSign' is 1, the 3691 | difference is negated before being returned. `zSign' is ignored if the 3692 | result is a NaN. The subtraction is performed according to the IEC/IEEE 3693 | Standard for Binary Floating-Point Arithmetic. 3694 *----------------------------------------------------------------------------*/ 3695 3696 static float64 subFloat64Sigs( float64 a, float64 b, flag zSign STATUS_PARAM ) 3697 { 3698 int_fast16_t aExp, bExp, zExp; 3699 uint64_t aSig, bSig, zSig; 3700 int_fast16_t expDiff; 3701 3702 aSig = extractFloat64Frac( a ); 3703 aExp = extractFloat64Exp( a ); 3704 bSig = extractFloat64Frac( b ); 3705 bExp = extractFloat64Exp( b ); 3706 expDiff = aExp - bExp; 3707 aSig <<= 10; 3708 bSig <<= 10; 3709 if ( 0 < expDiff ) goto aExpBigger; 3710 if ( expDiff < 0 ) goto bExpBigger; 3711 if ( aExp == 0x7FF ) { 3712 if ( aSig | bSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3713 float_raise( float_flag_invalid STATUS_VAR); 3714 return float64_default_nan; 3715 } 3716 if ( aExp == 0 ) { 3717 aExp = 1; 3718 bExp = 1; 3719 } 3720 if ( bSig < aSig ) goto aBigger; 3721 if ( aSig < bSig ) goto bBigger; 3722 return packFloat64( STATUS(float_rounding_mode) == float_round_down, 0, 0 ); 3723 bExpBigger: 3724 if ( bExp == 0x7FF ) { 3725 if ( bSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3726 return packFloat64( zSign ^ 1, 0x7FF, 0 ); 3727 } 3728 if ( aExp == 0 ) { 3729 ++expDiff; 3730 } 3731 else { 3732 aSig |= LIT64( 0x4000000000000000 ); 3733 } 3734 shift64RightJamming( aSig, - expDiff, &aSig ); 3735 bSig |= LIT64( 0x4000000000000000 ); 3736 bBigger: 3737 zSig = bSig - aSig; 3738 zExp = bExp; 3739 zSign ^= 1; 3740 goto normalizeRoundAndPack; 3741 aExpBigger: 3742 if ( aExp == 0x7FF ) { 3743 if ( aSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3744 return a; 3745 } 3746 if ( bExp == 0 ) { 3747 --expDiff; 3748 } 3749 else { 3750 bSig |= LIT64( 0x4000000000000000 ); 3751 } 3752 shift64RightJamming( bSig, expDiff, &bSig ); 3753 aSig |= LIT64( 0x4000000000000000 ); 3754 aBigger: 3755 zSig = aSig - bSig; 3756 zExp = aExp; 3757 normalizeRoundAndPack: 3758 --zExp; 3759 return normalizeRoundAndPackFloat64( zSign, zExp, zSig STATUS_VAR ); 3760 3761 } 3762 3763 /*---------------------------------------------------------------------------- 3764 | Returns the result of adding the double-precision floating-point values `a' 3765 | and `b'. The operation is performed according to the IEC/IEEE Standard for 3766 | Binary Floating-Point Arithmetic. 3767 *----------------------------------------------------------------------------*/ 3768 3769 float64 float64_add( float64 a, float64 b STATUS_PARAM ) 3770 { 3771 flag aSign, bSign; 3772 a = float64_squash_input_denormal(a STATUS_VAR); 3773 b = float64_squash_input_denormal(b STATUS_VAR); 3774 3775 aSign = extractFloat64Sign( a ); 3776 bSign = extractFloat64Sign( b ); 3777 if ( aSign == bSign ) { 3778 return addFloat64Sigs( a, b, aSign STATUS_VAR ); 3779 } 3780 else { 3781 return subFloat64Sigs( a, b, aSign STATUS_VAR ); 3782 } 3783 3784 } 3785 3786 /*---------------------------------------------------------------------------- 3787 | Returns the result of subtracting the double-precision floating-point values 3788 | `a' and `b'. The operation is performed according to the IEC/IEEE Standard 3789 | for Binary Floating-Point Arithmetic. 3790 *----------------------------------------------------------------------------*/ 3791 3792 float64 float64_sub( float64 a, float64 b STATUS_PARAM ) 3793 { 3794 flag aSign, bSign; 3795 a = float64_squash_input_denormal(a STATUS_VAR); 3796 b = float64_squash_input_denormal(b STATUS_VAR); 3797 3798 aSign = extractFloat64Sign( a ); 3799 bSign = extractFloat64Sign( b ); 3800 if ( aSign == bSign ) { 3801 return subFloat64Sigs( a, b, aSign STATUS_VAR ); 3802 } 3803 else { 3804 return addFloat64Sigs( a, b, aSign STATUS_VAR ); 3805 } 3806 3807 } 3808 3809 /*---------------------------------------------------------------------------- 3810 | Returns the result of multiplying the double-precision floating-point values 3811 | `a' and `b'. The operation is performed according to the IEC/IEEE Standard 3812 | for Binary Floating-Point Arithmetic. 3813 *----------------------------------------------------------------------------*/ 3814 3815 float64 float64_mul( float64 a, float64 b STATUS_PARAM ) 3816 { 3817 flag aSign, bSign, zSign; 3818 int_fast16_t aExp, bExp, zExp; 3819 uint64_t aSig, bSig, zSig0, zSig1; 3820 3821 a = float64_squash_input_denormal(a STATUS_VAR); 3822 b = float64_squash_input_denormal(b STATUS_VAR); 3823 3824 aSig = extractFloat64Frac( a ); 3825 aExp = extractFloat64Exp( a ); 3826 aSign = extractFloat64Sign( a ); 3827 bSig = extractFloat64Frac( b ); 3828 bExp = extractFloat64Exp( b ); 3829 bSign = extractFloat64Sign( b ); 3830 zSign = aSign ^ bSign; 3831 if ( aExp == 0x7FF ) { 3832 if ( aSig || ( ( bExp == 0x7FF ) && bSig ) ) { 3833 return propagateFloat64NaN( a, b STATUS_VAR ); 3834 } 3835 if ( ( bExp | bSig ) == 0 ) { 3836 float_raise( float_flag_invalid STATUS_VAR); 3837 return float64_default_nan; 3838 } 3839 return packFloat64( zSign, 0x7FF, 0 ); 3840 } 3841 if ( bExp == 0x7FF ) { 3842 if ( bSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3843 if ( ( aExp | aSig ) == 0 ) { 3844 float_raise( float_flag_invalid STATUS_VAR); 3845 return float64_default_nan; 3846 } 3847 return packFloat64( zSign, 0x7FF, 0 ); 3848 } 3849 if ( aExp == 0 ) { 3850 if ( aSig == 0 ) return packFloat64( zSign, 0, 0 ); 3851 normalizeFloat64Subnormal( aSig, &aExp, &aSig ); 3852 } 3853 if ( bExp == 0 ) { 3854 if ( bSig == 0 ) return packFloat64( zSign, 0, 0 ); 3855 normalizeFloat64Subnormal( bSig, &bExp, &bSig ); 3856 } 3857 zExp = aExp + bExp - 0x3FF; 3858 aSig = ( aSig | LIT64( 0x0010000000000000 ) )<<10; 3859 bSig = ( bSig | LIT64( 0x0010000000000000 ) )<<11; 3860 mul64To128( aSig, bSig, &zSig0, &zSig1 ); 3861 zSig0 |= ( zSig1 != 0 ); 3862 if ( 0 <= (int64_t) ( zSig0<<1 ) ) { 3863 zSig0 <<= 1; 3864 --zExp; 3865 } 3866 return roundAndPackFloat64( zSign, zExp, zSig0 STATUS_VAR ); 3867 3868 } 3869 3870 /*---------------------------------------------------------------------------- 3871 | Returns the result of dividing the double-precision floating-point value `a' 3872 | by the corresponding value `b'. The operation is performed according to 3873 | the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 3874 *----------------------------------------------------------------------------*/ 3875 3876 float64 float64_div( float64 a, float64 b STATUS_PARAM ) 3877 { 3878 flag aSign, bSign, zSign; 3879 int_fast16_t aExp, bExp, zExp; 3880 uint64_t aSig, bSig, zSig; 3881 uint64_t rem0, rem1; 3882 uint64_t term0, term1; 3883 a = float64_squash_input_denormal(a STATUS_VAR); 3884 b = float64_squash_input_denormal(b STATUS_VAR); 3885 3886 aSig = extractFloat64Frac( a ); 3887 aExp = extractFloat64Exp( a ); 3888 aSign = extractFloat64Sign( a ); 3889 bSig = extractFloat64Frac( b ); 3890 bExp = extractFloat64Exp( b ); 3891 bSign = extractFloat64Sign( b ); 3892 zSign = aSign ^ bSign; 3893 if ( aExp == 0x7FF ) { 3894 if ( aSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3895 if ( bExp == 0x7FF ) { 3896 if ( bSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3897 float_raise( float_flag_invalid STATUS_VAR); 3898 return float64_default_nan; 3899 } 3900 return packFloat64( zSign, 0x7FF, 0 ); 3901 } 3902 if ( bExp == 0x7FF ) { 3903 if ( bSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3904 return packFloat64( zSign, 0, 0 ); 3905 } 3906 if ( bExp == 0 ) { 3907 if ( bSig == 0 ) { 3908 if ( ( aExp | aSig ) == 0 ) { 3909 float_raise( float_flag_invalid STATUS_VAR); 3910 return float64_default_nan; 3911 } 3912 float_raise( float_flag_divbyzero STATUS_VAR); 3913 return packFloat64( zSign, 0x7FF, 0 ); 3914 } 3915 normalizeFloat64Subnormal( bSig, &bExp, &bSig ); 3916 } 3917 if ( aExp == 0 ) { 3918 if ( aSig == 0 ) return packFloat64( zSign, 0, 0 ); 3919 normalizeFloat64Subnormal( aSig, &aExp, &aSig ); 3920 } 3921 zExp = aExp - bExp + 0x3FD; 3922 aSig = ( aSig | LIT64( 0x0010000000000000 ) )<<10; 3923 bSig = ( bSig | LIT64( 0x0010000000000000 ) )<<11; 3924 if ( bSig <= ( aSig + aSig ) ) { 3925 aSig >>= 1; 3926 ++zExp; 3927 } 3928 zSig = estimateDiv128To64( aSig, 0, bSig ); 3929 if ( ( zSig & 0x1FF ) <= 2 ) { 3930 mul64To128( bSig, zSig, &term0, &term1 ); 3931 sub128( aSig, 0, term0, term1, &rem0, &rem1 ); 3932 while ( (int64_t) rem0 < 0 ) { 3933 --zSig; 3934 add128( rem0, rem1, 0, bSig, &rem0, &rem1 ); 3935 } 3936 zSig |= ( rem1 != 0 ); 3937 } 3938 return roundAndPackFloat64( zSign, zExp, zSig STATUS_VAR ); 3939 3940 } 3941 3942 /*---------------------------------------------------------------------------- 3943 | Returns the remainder of the double-precision floating-point value `a' 3944 | with respect to the corresponding value `b'. The operation is performed 3945 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 3946 *----------------------------------------------------------------------------*/ 3947 3948 float64 float64_rem( float64 a, float64 b STATUS_PARAM ) 3949 { 3950 flag aSign, zSign; 3951 int_fast16_t aExp, bExp, expDiff; 3952 uint64_t aSig, bSig; 3953 uint64_t q, alternateASig; 3954 int64_t sigMean; 3955 3956 a = float64_squash_input_denormal(a STATUS_VAR); 3957 b = float64_squash_input_denormal(b STATUS_VAR); 3958 aSig = extractFloat64Frac( a ); 3959 aExp = extractFloat64Exp( a ); 3960 aSign = extractFloat64Sign( a ); 3961 bSig = extractFloat64Frac( b ); 3962 bExp = extractFloat64Exp( b ); 3963 if ( aExp == 0x7FF ) { 3964 if ( aSig || ( ( bExp == 0x7FF ) && bSig ) ) { 3965 return propagateFloat64NaN( a, b STATUS_VAR ); 3966 } 3967 float_raise( float_flag_invalid STATUS_VAR); 3968 return float64_default_nan; 3969 } 3970 if ( bExp == 0x7FF ) { 3971 if ( bSig ) return propagateFloat64NaN( a, b STATUS_VAR ); 3972 return a; 3973 } 3974 if ( bExp == 0 ) { 3975 if ( bSig == 0 ) { 3976 float_raise( float_flag_invalid STATUS_VAR); 3977 return float64_default_nan; 3978 } 3979 normalizeFloat64Subnormal( bSig, &bExp, &bSig ); 3980 } 3981 if ( aExp == 0 ) { 3982 if ( aSig == 0 ) return a; 3983 normalizeFloat64Subnormal( aSig, &aExp, &aSig ); 3984 } 3985 expDiff = aExp - bExp; 3986 aSig = ( aSig | LIT64( 0x0010000000000000 ) )<<11; 3987 bSig = ( bSig | LIT64( 0x0010000000000000 ) )<<11; 3988 if ( expDiff < 0 ) { 3989 if ( expDiff < -1 ) return a; 3990 aSig >>= 1; 3991 } 3992 q = ( bSig <= aSig ); 3993 if ( q ) aSig -= bSig; 3994 expDiff -= 64; 3995 while ( 0 < expDiff ) { 3996 q = estimateDiv128To64( aSig, 0, bSig ); 3997 q = ( 2 < q ) ? q - 2 : 0; 3998 aSig = - ( ( bSig>>2 ) * q ); 3999 expDiff -= 62; 4000 } 4001 expDiff += 64; 4002 if ( 0 < expDiff ) { 4003 q = estimateDiv128To64( aSig, 0, bSig ); 4004 q = ( 2 < q ) ? q - 2 : 0; 4005 q >>= 64 - expDiff; 4006 bSig >>= 2; 4007 aSig = ( ( aSig>>1 )<<( expDiff - 1 ) ) - bSig * q; 4008 } 4009 else { 4010 aSig >>= 2; 4011 bSig >>= 2; 4012 } 4013 do { 4014 alternateASig = aSig; 4015 ++q; 4016 aSig -= bSig; 4017 } while ( 0 <= (int64_t) aSig ); 4018 sigMean = aSig + alternateASig; 4019 if ( ( sigMean < 0 ) || ( ( sigMean == 0 ) && ( q & 1 ) ) ) { 4020 aSig = alternateASig; 4021 } 4022 zSign = ( (int64_t) aSig < 0 ); 4023 if ( zSign ) aSig = - aSig; 4024 return normalizeRoundAndPackFloat64( aSign ^ zSign, bExp, aSig STATUS_VAR ); 4025 4026 } 4027 4028 /*---------------------------------------------------------------------------- 4029 | Returns the result of multiplying the double-precision floating-point values 4030 | `a' and `b' then adding 'c', with no intermediate rounding step after the 4031 | multiplication. The operation is performed according to the IEC/IEEE 4032 | Standard for Binary Floating-Point Arithmetic 754-2008. 4033 | The flags argument allows the caller to select negation of the 4034 | addend, the intermediate product, or the final result. (The difference 4035 | between this and having the caller do a separate negation is that negating 4036 | externally will flip the sign bit on NaNs.) 4037 *----------------------------------------------------------------------------*/ 4038 4039 float64 float64_muladd(float64 a, float64 b, float64 c, int flags STATUS_PARAM) 4040 { 4041 flag aSign, bSign, cSign, zSign; 4042 int_fast16_t aExp, bExp, cExp, pExp, zExp, expDiff; 4043 uint64_t aSig, bSig, cSig; 4044 flag pInf, pZero, pSign; 4045 uint64_t pSig0, pSig1, cSig0, cSig1, zSig0, zSig1; 4046 int shiftcount; 4047 flag signflip, infzero; 4048 4049 a = float64_squash_input_denormal(a STATUS_VAR); 4050 b = float64_squash_input_denormal(b STATUS_VAR); 4051 c = float64_squash_input_denormal(c STATUS_VAR); 4052 aSig = extractFloat64Frac(a); 4053 aExp = extractFloat64Exp(a); 4054 aSign = extractFloat64Sign(a); 4055 bSig = extractFloat64Frac(b); 4056 bExp = extractFloat64Exp(b); 4057 bSign = extractFloat64Sign(b); 4058 cSig = extractFloat64Frac(c); 4059 cExp = extractFloat64Exp(c); 4060 cSign = extractFloat64Sign(c); 4061 4062 infzero = ((aExp == 0 && aSig == 0 && bExp == 0x7ff && bSig == 0) || 4063 (aExp == 0x7ff && aSig == 0 && bExp == 0 && bSig == 0)); 4064 4065 /* It is implementation-defined whether the cases of (0,inf,qnan) 4066 * and (inf,0,qnan) raise InvalidOperation or not (and what QNaN 4067 * they return if they do), so we have to hand this information 4068 * off to the target-specific pick-a-NaN routine. 4069 */ 4070 if (((aExp == 0x7ff) && aSig) || 4071 ((bExp == 0x7ff) && bSig) || 4072 ((cExp == 0x7ff) && cSig)) { 4073 return propagateFloat64MulAddNaN(a, b, c, infzero STATUS_VAR); 4074 } 4075 4076 if (infzero) { 4077 float_raise(float_flag_invalid STATUS_VAR); 4078 return float64_default_nan; 4079 } 4080 4081 if (flags & float_muladd_negate_c) { 4082 cSign ^= 1; 4083 } 4084 4085 signflip = (flags & float_muladd_negate_result) ? 1 : 0; 4086 4087 /* Work out the sign and type of the product */ 4088 pSign = aSign ^ bSign; 4089 if (flags & float_muladd_negate_product) { 4090 pSign ^= 1; 4091 } 4092 pInf = (aExp == 0x7ff) || (bExp == 0x7ff); 4093 pZero = ((aExp | aSig) == 0) || ((bExp | bSig) == 0); 4094 4095 if (cExp == 0x7ff) { 4096 if (pInf && (pSign ^ cSign)) { 4097 /* addition of opposite-signed infinities => InvalidOperation */ 4098 float_raise(float_flag_invalid STATUS_VAR); 4099 return float64_default_nan; 4100 } 4101 /* Otherwise generate an infinity of the same sign */ 4102 return packFloat64(cSign ^ signflip, 0x7ff, 0); 4103 } 4104 4105 if (pInf) { 4106 return packFloat64(pSign ^ signflip, 0x7ff, 0); 4107 } 4108 4109 if (pZero) { 4110 if (cExp == 0) { 4111 if (cSig == 0) { 4112 /* Adding two exact zeroes */ 4113 if (pSign == cSign) { 4114 zSign = pSign; 4115 } else if (STATUS(float_rounding_mode) == float_round_down) { 4116 zSign = 1; 4117 } else { 4118 zSign = 0; 4119 } 4120 return packFloat64(zSign ^ signflip, 0, 0); 4121 } 4122 /* Exact zero plus a denorm */ 4123 if (STATUS(flush_to_zero)) { 4124 float_raise(float_flag_output_denormal STATUS_VAR); 4125 return packFloat64(cSign ^ signflip, 0, 0); 4126 } 4127 } 4128 /* Zero plus something non-zero : just return the something */ 4129 if (flags & float_muladd_halve_result) { 4130 if (cExp == 0) { 4131 normalizeFloat64Subnormal(cSig, &cExp, &cSig); 4132 } 4133 /* Subtract one to halve, and one again because roundAndPackFloat64 4134 * wants one less than the true exponent. 4135 */ 4136 cExp -= 2; 4137 cSig = (cSig | 0x0010000000000000ULL) << 10; 4138 return roundAndPackFloat64(cSign ^ signflip, cExp, cSig STATUS_VAR); 4139 } 4140 return packFloat64(cSign ^ signflip, cExp, cSig); 4141 } 4142 4143 if (aExp == 0) { 4144 normalizeFloat64Subnormal(aSig, &aExp, &aSig); 4145 } 4146 if (bExp == 0) { 4147 normalizeFloat64Subnormal(bSig, &bExp, &bSig); 4148 } 4149 4150 /* Calculate the actual result a * b + c */ 4151 4152 /* Multiply first; this is easy. */ 4153 /* NB: we subtract 0x3fe where float64_mul() subtracts 0x3ff 4154 * because we want the true exponent, not the "one-less-than" 4155 * flavour that roundAndPackFloat64() takes. 4156 */ 4157 pExp = aExp + bExp - 0x3fe; 4158 aSig = (aSig | LIT64(0x0010000000000000))<<10; 4159 bSig = (bSig | LIT64(0x0010000000000000))<<11; 4160 mul64To128(aSig, bSig, &pSig0, &pSig1); 4161 if ((int64_t)(pSig0 << 1) >= 0) { 4162 shortShift128Left(pSig0, pSig1, 1, &pSig0, &pSig1); 4163 pExp--; 4164 } 4165 4166 zSign = pSign ^ signflip; 4167 4168 /* Now [pSig0:pSig1] is the significand of the multiply, with the explicit 4169 * bit in position 126. 4170 */ 4171 if (cExp == 0) { 4172 if (!cSig) { 4173 /* Throw out the special case of c being an exact zero now */ 4174 shift128RightJamming(pSig0, pSig1, 64, &pSig0, &pSig1); 4175 if (flags & float_muladd_halve_result) { 4176 pExp--; 4177 } 4178 return roundAndPackFloat64(zSign, pExp - 1, 4179 pSig1 STATUS_VAR); 4180 } 4181 normalizeFloat64Subnormal(cSig, &cExp, &cSig); 4182 } 4183 4184 /* Shift cSig and add the explicit bit so [cSig0:cSig1] is the 4185 * significand of the addend, with the explicit bit in position 126. 4186 */ 4187 cSig0 = cSig << (126 - 64 - 52); 4188 cSig1 = 0; 4189 cSig0 |= LIT64(0x4000000000000000); 4190 expDiff = pExp - cExp; 4191 4192 if (pSign == cSign) { 4193 /* Addition */ 4194 if (expDiff > 0) { 4195 /* scale c to match p */ 4196 shift128RightJamming(cSig0, cSig1, expDiff, &cSig0, &cSig1); 4197 zExp = pExp; 4198 } else if (expDiff < 0) { 4199 /* scale p to match c */ 4200 shift128RightJamming(pSig0, pSig1, -expDiff, &pSig0, &pSig1); 4201 zExp = cExp; 4202 } else { 4203 /* no scaling needed */ 4204 zExp = cExp; 4205 } 4206 /* Add significands and make sure explicit bit ends up in posn 126 */ 4207 add128(pSig0, pSig1, cSig0, cSig1, &zSig0, &zSig1); 4208 if ((int64_t)zSig0 < 0) { 4209 shift128RightJamming(zSig0, zSig1, 1, &zSig0, &zSig1); 4210 } else { 4211 zExp--; 4212 } 4213 shift128RightJamming(zSig0, zSig1, 64, &zSig0, &zSig1); 4214 if (flags & float_muladd_halve_result) { 4215 zExp--; 4216 } 4217 return roundAndPackFloat64(zSign, zExp, zSig1 STATUS_VAR); 4218 } else { 4219 /* Subtraction */ 4220 if (expDiff > 0) { 4221 shift128RightJamming(cSig0, cSig1, expDiff, &cSig0, &cSig1); 4222 sub128(pSig0, pSig1, cSig0, cSig1, &zSig0, &zSig1); 4223 zExp = pExp; 4224 } else if (expDiff < 0) { 4225 shift128RightJamming(pSig0, pSig1, -expDiff, &pSig0, &pSig1); 4226 sub128(cSig0, cSig1, pSig0, pSig1, &zSig0, &zSig1); 4227 zExp = cExp; 4228 zSign ^= 1; 4229 } else { 4230 zExp = pExp; 4231 if (lt128(cSig0, cSig1, pSig0, pSig1)) { 4232 sub128(pSig0, pSig1, cSig0, cSig1, &zSig0, &zSig1); 4233 } else if (lt128(pSig0, pSig1, cSig0, cSig1)) { 4234 sub128(cSig0, cSig1, pSig0, pSig1, &zSig0, &zSig1); 4235 zSign ^= 1; 4236 } else { 4237 /* Exact zero */ 4238 zSign = signflip; 4239 if (STATUS(float_rounding_mode) == float_round_down) { 4240 zSign ^= 1; 4241 } 4242 return packFloat64(zSign, 0, 0); 4243 } 4244 } 4245 --zExp; 4246 /* Do the equivalent of normalizeRoundAndPackFloat64() but 4247 * starting with the significand in a pair of uint64_t. 4248 */ 4249 if (zSig0) { 4250 shiftcount = countLeadingZeros64(zSig0) - 1; 4251 shortShift128Left(zSig0, zSig1, shiftcount, &zSig0, &zSig1); 4252 if (zSig1) { 4253 zSig0 |= 1; 4254 } 4255 zExp -= shiftcount; 4256 } else { 4257 shiftcount = countLeadingZeros64(zSig1); 4258 if (shiftcount == 0) { 4259 zSig0 = (zSig1 >> 1) | (zSig1 & 1); 4260 zExp -= 63; 4261 } else { 4262 shiftcount--; 4263 zSig0 = zSig1 << shiftcount; 4264 zExp -= (shiftcount + 64); 4265 } 4266 } 4267 if (flags & float_muladd_halve_result) { 4268 zExp--; 4269 } 4270 return roundAndPackFloat64(zSign, zExp, zSig0 STATUS_VAR); 4271 } 4272 } 4273 4274 /*---------------------------------------------------------------------------- 4275 | Returns the square root of the double-precision floating-point value `a'. 4276 | The operation is performed according to the IEC/IEEE Standard for Binary 4277 | Floating-Point Arithmetic. 4278 *----------------------------------------------------------------------------*/ 4279 4280 float64 float64_sqrt( float64 a STATUS_PARAM ) 4281 { 4282 flag aSign; 4283 int_fast16_t aExp, zExp; 4284 uint64_t aSig, zSig, doubleZSig; 4285 uint64_t rem0, rem1, term0, term1; 4286 a = float64_squash_input_denormal(a STATUS_VAR); 4287 4288 aSig = extractFloat64Frac( a ); 4289 aExp = extractFloat64Exp( a ); 4290 aSign = extractFloat64Sign( a ); 4291 if ( aExp == 0x7FF ) { 4292 if ( aSig ) return propagateFloat64NaN( a, a STATUS_VAR ); 4293 if ( ! aSign ) return a; 4294 float_raise( float_flag_invalid STATUS_VAR); 4295 return float64_default_nan; 4296 } 4297 if ( aSign ) { 4298 if ( ( aExp | aSig ) == 0 ) return a; 4299 float_raise( float_flag_invalid STATUS_VAR); 4300 return float64_default_nan; 4301 } 4302 if ( aExp == 0 ) { 4303 if ( aSig == 0 ) return float64_zero; 4304 normalizeFloat64Subnormal( aSig, &aExp, &aSig ); 4305 } 4306 zExp = ( ( aExp - 0x3FF )>>1 ) + 0x3FE; 4307 aSig |= LIT64( 0x0010000000000000 ); 4308 zSig = estimateSqrt32( aExp, aSig>>21 ); 4309 aSig <<= 9 - ( aExp & 1 ); 4310 zSig = estimateDiv128To64( aSig, 0, zSig<<32 ) + ( zSig<<30 ); 4311 if ( ( zSig & 0x1FF ) <= 5 ) { 4312 doubleZSig = zSig<<1; 4313 mul64To128( zSig, zSig, &term0, &term1 ); 4314 sub128( aSig, 0, term0, term1, &rem0, &rem1 ); 4315 while ( (int64_t) rem0 < 0 ) { 4316 --zSig; 4317 doubleZSig -= 2; 4318 add128( rem0, rem1, zSig>>63, doubleZSig | 1, &rem0, &rem1 ); 4319 } 4320 zSig |= ( ( rem0 | rem1 ) != 0 ); 4321 } 4322 return roundAndPackFloat64( 0, zExp, zSig STATUS_VAR ); 4323 4324 } 4325 4326 /*---------------------------------------------------------------------------- 4327 | Returns the binary log of the double-precision floating-point value `a'. 4328 | The operation is performed according to the IEC/IEEE Standard for Binary 4329 | Floating-Point Arithmetic. 4330 *----------------------------------------------------------------------------*/ 4331 float64 float64_log2( float64 a STATUS_PARAM ) 4332 { 4333 flag aSign, zSign; 4334 int_fast16_t aExp; 4335 uint64_t aSig, aSig0, aSig1, zSig, i; 4336 a = float64_squash_input_denormal(a STATUS_VAR); 4337 4338 aSig = extractFloat64Frac( a ); 4339 aExp = extractFloat64Exp( a ); 4340 aSign = extractFloat64Sign( a ); 4341 4342 if ( aExp == 0 ) { 4343 if ( aSig == 0 ) return packFloat64( 1, 0x7FF, 0 ); 4344 normalizeFloat64Subnormal( aSig, &aExp, &aSig ); 4345 } 4346 if ( aSign ) { 4347 float_raise( float_flag_invalid STATUS_VAR); 4348 return float64_default_nan; 4349 } 4350 if ( aExp == 0x7FF ) { 4351 if ( aSig ) return propagateFloat64NaN( a, float64_zero STATUS_VAR ); 4352 return a; 4353 } 4354 4355 aExp -= 0x3FF; 4356 aSig |= LIT64( 0x0010000000000000 ); 4357 zSign = aExp < 0; 4358 zSig = (uint64_t)aExp << 52; 4359 for (i = 1LL << 51; i > 0; i >>= 1) { 4360 mul64To128( aSig, aSig, &aSig0, &aSig1 ); 4361 aSig = ( aSig0 << 12 ) | ( aSig1 >> 52 ); 4362 if ( aSig & LIT64( 0x0020000000000000 ) ) { 4363 aSig >>= 1; 4364 zSig |= i; 4365 } 4366 } 4367 4368 if ( zSign ) 4369 zSig = -zSig; 4370 return normalizeRoundAndPackFloat64( zSign, 0x408, zSig STATUS_VAR ); 4371 } 4372 4373 /*---------------------------------------------------------------------------- 4374 | Returns 1 if the double-precision floating-point value `a' is equal to the 4375 | corresponding value `b', and 0 otherwise. The invalid exception is raised 4376 | if either operand is a NaN. Otherwise, the comparison is performed 4377 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 4378 *----------------------------------------------------------------------------*/ 4379 4380 int float64_eq( float64 a, float64 b STATUS_PARAM ) 4381 { 4382 uint64_t av, bv; 4383 a = float64_squash_input_denormal(a STATUS_VAR); 4384 b = float64_squash_input_denormal(b STATUS_VAR); 4385 4386 if ( ( ( extractFloat64Exp( a ) == 0x7FF ) && extractFloat64Frac( a ) ) 4387 || ( ( extractFloat64Exp( b ) == 0x7FF ) && extractFloat64Frac( b ) ) 4388 ) { 4389 float_raise( float_flag_invalid STATUS_VAR); 4390 return 0; 4391 } 4392 av = float64_val(a); 4393 bv = float64_val(b); 4394 return ( av == bv ) || ( (uint64_t) ( ( av | bv )<<1 ) == 0 ); 4395 4396 } 4397 4398 /*---------------------------------------------------------------------------- 4399 | Returns 1 if the double-precision floating-point value `a' is less than or 4400 | equal to the corresponding value `b', and 0 otherwise. The invalid 4401 | exception is raised if either operand is a NaN. The comparison is performed 4402 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 4403 *----------------------------------------------------------------------------*/ 4404 4405 int float64_le( float64 a, float64 b STATUS_PARAM ) 4406 { 4407 flag aSign, bSign; 4408 uint64_t av, bv; 4409 a = float64_squash_input_denormal(a STATUS_VAR); 4410 b = float64_squash_input_denormal(b STATUS_VAR); 4411 4412 if ( ( ( extractFloat64Exp( a ) == 0x7FF ) && extractFloat64Frac( a ) ) 4413 || ( ( extractFloat64Exp( b ) == 0x7FF ) && extractFloat64Frac( b ) ) 4414 ) { 4415 float_raise( float_flag_invalid STATUS_VAR); 4416 return 0; 4417 } 4418 aSign = extractFloat64Sign( a ); 4419 bSign = extractFloat64Sign( b ); 4420 av = float64_val(a); 4421 bv = float64_val(b); 4422 if ( aSign != bSign ) return aSign || ( (uint64_t) ( ( av | bv )<<1 ) == 0 ); 4423 return ( av == bv ) || ( aSign ^ ( av < bv ) ); 4424 4425 } 4426 4427 /*---------------------------------------------------------------------------- 4428 | Returns 1 if the double-precision floating-point value `a' is less than 4429 | the corresponding value `b', and 0 otherwise. The invalid exception is 4430 | raised if either operand is a NaN. The comparison is performed according 4431 | to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 4432 *----------------------------------------------------------------------------*/ 4433 4434 int float64_lt( float64 a, float64 b STATUS_PARAM ) 4435 { 4436 flag aSign, bSign; 4437 uint64_t av, bv; 4438 4439 a = float64_squash_input_denormal(a STATUS_VAR); 4440 b = float64_squash_input_denormal(b STATUS_VAR); 4441 if ( ( ( extractFloat64Exp( a ) == 0x7FF ) && extractFloat64Frac( a ) ) 4442 || ( ( extractFloat64Exp( b ) == 0x7FF ) && extractFloat64Frac( b ) ) 4443 ) { 4444 float_raise( float_flag_invalid STATUS_VAR); 4445 return 0; 4446 } 4447 aSign = extractFloat64Sign( a ); 4448 bSign = extractFloat64Sign( b ); 4449 av = float64_val(a); 4450 bv = float64_val(b); 4451 if ( aSign != bSign ) return aSign && ( (uint64_t) ( ( av | bv )<<1 ) != 0 ); 4452 return ( av != bv ) && ( aSign ^ ( av < bv ) ); 4453 4454 } 4455 4456 /*---------------------------------------------------------------------------- 4457 | Returns 1 if the double-precision floating-point values `a' and `b' cannot 4458 | be compared, and 0 otherwise. The invalid exception is raised if either 4459 | operand is a NaN. The comparison is performed according to the IEC/IEEE 4460 | Standard for Binary Floating-Point Arithmetic. 4461 *----------------------------------------------------------------------------*/ 4462 4463 int float64_unordered( float64 a, float64 b STATUS_PARAM ) 4464 { 4465 a = float64_squash_input_denormal(a STATUS_VAR); 4466 b = float64_squash_input_denormal(b STATUS_VAR); 4467 4468 if ( ( ( extractFloat64Exp( a ) == 0x7FF ) && extractFloat64Frac( a ) ) 4469 || ( ( extractFloat64Exp( b ) == 0x7FF ) && extractFloat64Frac( b ) ) 4470 ) { 4471 float_raise( float_flag_invalid STATUS_VAR); 4472 return 1; 4473 } 4474 return 0; 4475 } 4476 4477 /*---------------------------------------------------------------------------- 4478 | Returns 1 if the double-precision floating-point value `a' is equal to the 4479 | corresponding value `b', and 0 otherwise. Quiet NaNs do not cause an 4480 | exception.The comparison is performed according to the IEC/IEEE Standard 4481 | for Binary Floating-Point Arithmetic. 4482 *----------------------------------------------------------------------------*/ 4483 4484 int float64_eq_quiet( float64 a, float64 b STATUS_PARAM ) 4485 { 4486 uint64_t av, bv; 4487 a = float64_squash_input_denormal(a STATUS_VAR); 4488 b = float64_squash_input_denormal(b STATUS_VAR); 4489 4490 if ( ( ( extractFloat64Exp( a ) == 0x7FF ) && extractFloat64Frac( a ) ) 4491 || ( ( extractFloat64Exp( b ) == 0x7FF ) && extractFloat64Frac( b ) ) 4492 ) { 4493 if ( float64_is_signaling_nan( a ) || float64_is_signaling_nan( b ) ) { 4494 float_raise( float_flag_invalid STATUS_VAR); 4495 } 4496 return 0; 4497 } 4498 av = float64_val(a); 4499 bv = float64_val(b); 4500 return ( av == bv ) || ( (uint64_t) ( ( av | bv )<<1 ) == 0 ); 4501 4502 } 4503 4504 /*---------------------------------------------------------------------------- 4505 | Returns 1 if the double-precision floating-point value `a' is less than or 4506 | equal to the corresponding value `b', and 0 otherwise. Quiet NaNs do not 4507 | cause an exception. Otherwise, the comparison is performed according to the 4508 | IEC/IEEE Standard for Binary Floating-Point Arithmetic. 4509 *----------------------------------------------------------------------------*/ 4510 4511 int float64_le_quiet( float64 a, float64 b STATUS_PARAM ) 4512 { 4513 flag aSign, bSign; 4514 uint64_t av, bv; 4515 a = float64_squash_input_denormal(a STATUS_VAR); 4516 b = float64_squash_input_denormal(b STATUS_VAR); 4517 4518 if ( ( ( extractFloat64Exp( a ) == 0x7FF ) && extractFloat64Frac( a ) ) 4519 || ( ( extractFloat64Exp( b ) == 0x7FF ) && extractFloat64Frac( b ) ) 4520 ) { 4521 if ( float64_is_signaling_nan( a ) || float64_is_signaling_nan( b ) ) { 4522 float_raise( float_flag_invalid STATUS_VAR); 4523 } 4524 return 0; 4525 } 4526 aSign = extractFloat64Sign( a ); 4527 bSign = extractFloat64Sign( b ); 4528 av = float64_val(a); 4529 bv = float64_val(b); 4530 if ( aSign != bSign ) return aSign || ( (uint64_t) ( ( av | bv )<<1 ) == 0 ); 4531 return ( av == bv ) || ( aSign ^ ( av < bv ) ); 4532 4533 } 4534 4535 /*---------------------------------------------------------------------------- 4536 | Returns 1 if the double-precision floating-point value `a' is less than 4537 | the corresponding value `b', and 0 otherwise. Quiet NaNs do not cause an 4538 | exception. Otherwise, the comparison is performed according to the IEC/IEEE 4539 | Standard for Binary Floating-Point Arithmetic. 4540 *----------------------------------------------------------------------------*/ 4541 4542 int float64_lt_quiet( float64 a, float64 b STATUS_PARAM ) 4543 { 4544 flag aSign, bSign; 4545 uint64_t av, bv; 4546 a = float64_squash_input_denormal(a STATUS_VAR); 4547 b = float64_squash_input_denormal(b STATUS_VAR); 4548 4549 if ( ( ( extractFloat64Exp( a ) == 0x7FF ) && extractFloat64Frac( a ) ) 4550 || ( ( extractFloat64Exp( b ) == 0x7FF ) && extractFloat64Frac( b ) ) 4551 ) { 4552 if ( float64_is_signaling_nan( a ) || float64_is_signaling_nan( b ) ) { 4553 float_raise( float_flag_invalid STATUS_VAR); 4554 } 4555 return 0; 4556 } 4557 aSign = extractFloat64Sign( a ); 4558 bSign = extractFloat64Sign( b ); 4559 av = float64_val(a); 4560 bv = float64_val(b); 4561 if ( aSign != bSign ) return aSign && ( (uint64_t) ( ( av | bv )<<1 ) != 0 ); 4562 return ( av != bv ) && ( aSign ^ ( av < bv ) ); 4563 4564 } 4565 4566 /*---------------------------------------------------------------------------- 4567 | Returns 1 if the double-precision floating-point values `a' and `b' cannot 4568 | be compared, and 0 otherwise. Quiet NaNs do not cause an exception. The 4569 | comparison is performed according to the IEC/IEEE Standard for Binary 4570 | Floating-Point Arithmetic. 4571 *----------------------------------------------------------------------------*/ 4572 4573 int float64_unordered_quiet( float64 a, float64 b STATUS_PARAM ) 4574 { 4575 a = float64_squash_input_denormal(a STATUS_VAR); 4576 b = float64_squash_input_denormal(b STATUS_VAR); 4577 4578 if ( ( ( extractFloat64Exp( a ) == 0x7FF ) && extractFloat64Frac( a ) ) 4579 || ( ( extractFloat64Exp( b ) == 0x7FF ) && extractFloat64Frac( b ) ) 4580 ) { 4581 if ( float64_is_signaling_nan( a ) || float64_is_signaling_nan( b ) ) { 4582 float_raise( float_flag_invalid STATUS_VAR); 4583 } 4584 return 1; 4585 } 4586 return 0; 4587 } 4588 4589 /*---------------------------------------------------------------------------- 4590 | Returns the result of converting the extended double-precision floating- 4591 | point value `a' to the 32-bit two's complement integer format. The 4592 | conversion is performed according to the IEC/IEEE Standard for Binary 4593 | Floating-Point Arithmetic---which means in particular that the conversion 4594 | is rounded according to the current rounding mode. If `a' is a NaN, the 4595 | largest positive integer is returned. Otherwise, if the conversion 4596 | overflows, the largest integer with the same sign as `a' is returned. 4597 *----------------------------------------------------------------------------*/ 4598 4599 int32 floatx80_to_int32( floatx80 a STATUS_PARAM ) 4600 { 4601 flag aSign; 4602 int32 aExp, shiftCount; 4603 uint64_t aSig; 4604 4605 aSig = extractFloatx80Frac( a ); 4606 aExp = extractFloatx80Exp( a ); 4607 aSign = extractFloatx80Sign( a ); 4608 if ( ( aExp == 0x7FFF ) && (uint64_t) ( aSig<<1 ) ) aSign = 0; 4609 shiftCount = 0x4037 - aExp; 4610 if ( shiftCount <= 0 ) shiftCount = 1; 4611 shift64RightJamming( aSig, shiftCount, &aSig ); 4612 return roundAndPackInt32( aSign, aSig STATUS_VAR ); 4613 4614 } 4615 4616 /*---------------------------------------------------------------------------- 4617 | Returns the result of converting the extended double-precision floating- 4618 | point value `a' to the 32-bit two's complement integer format. The 4619 | conversion is performed according to the IEC/IEEE Standard for Binary 4620 | Floating-Point Arithmetic, except that the conversion is always rounded 4621 | toward zero. If `a' is a NaN, the largest positive integer is returned. 4622 | Otherwise, if the conversion overflows, the largest integer with the same 4623 | sign as `a' is returned. 4624 *----------------------------------------------------------------------------*/ 4625 4626 int32 floatx80_to_int32_round_to_zero( floatx80 a STATUS_PARAM ) 4627 { 4628 flag aSign; 4629 int32 aExp, shiftCount; 4630 uint64_t aSig, savedASig; 4631 int32_t z; 4632 4633 aSig = extractFloatx80Frac( a ); 4634 aExp = extractFloatx80Exp( a ); 4635 aSign = extractFloatx80Sign( a ); 4636 if ( 0x401E < aExp ) { 4637 if ( ( aExp == 0x7FFF ) && (uint64_t) ( aSig<<1 ) ) aSign = 0; 4638 goto invalid; 4639 } 4640 else if ( aExp < 0x3FFF ) { 4641 if ( aExp || aSig ) STATUS(float_exception_flags) |= float_flag_inexact; 4642 return 0; 4643 } 4644 shiftCount = 0x403E - aExp; 4645 savedASig = aSig; 4646 aSig >>= shiftCount; 4647 z = aSig; 4648 if ( aSign ) z = - z; 4649 if ( ( z < 0 ) ^ aSign ) { 4650 invalid: 4651 float_raise( float_flag_invalid STATUS_VAR); 4652 return aSign ? (int32_t) 0x80000000 : 0x7FFFFFFF; 4653 } 4654 if ( ( aSig<<shiftCount ) != savedASig ) { 4655 STATUS(float_exception_flags) |= float_flag_inexact; 4656 } 4657 return z; 4658 4659 } 4660 4661 /*---------------------------------------------------------------------------- 4662 | Returns the result of converting the extended double-precision floating- 4663 | point value `a' to the 64-bit two's complement integer format. The 4664 | conversion is performed according to the IEC/IEEE Standard for Binary 4665 | Floating-Point Arithmetic---which means in particular that the conversion 4666 | is rounded according to the current rounding mode. If `a' is a NaN, 4667 | the largest positive integer is returned. Otherwise, if the conversion 4668 | overflows, the largest integer with the same sign as `a' is returned. 4669 *----------------------------------------------------------------------------*/ 4670 4671 int64 floatx80_to_int64( floatx80 a STATUS_PARAM ) 4672 { 4673 flag aSign; 4674 int32 aExp, shiftCount; 4675 uint64_t aSig, aSigExtra; 4676 4677 aSig = extractFloatx80Frac( a ); 4678 aExp = extractFloatx80Exp( a ); 4679 aSign = extractFloatx80Sign( a ); 4680 shiftCount = 0x403E - aExp; 4681 if ( shiftCount <= 0 ) { 4682 if ( shiftCount ) { 4683 float_raise( float_flag_invalid STATUS_VAR); 4684 if ( ! aSign 4685 || ( ( aExp == 0x7FFF ) 4686 && ( aSig != LIT64( 0x8000000000000000 ) ) ) 4687 ) { 4688 return LIT64( 0x7FFFFFFFFFFFFFFF ); 4689 } 4690 return (int64_t) LIT64( 0x8000000000000000 ); 4691 } 4692 aSigExtra = 0; 4693 } 4694 else { 4695 shift64ExtraRightJamming( aSig, 0, shiftCount, &aSig, &aSigExtra ); 4696 } 4697 return roundAndPackInt64( aSign, aSig, aSigExtra STATUS_VAR ); 4698 4699 } 4700 4701 /*---------------------------------------------------------------------------- 4702 | Returns the result of converting the extended double-precision floating- 4703 | point value `a' to the 64-bit two's complement integer format. The 4704 | conversion is performed according to the IEC/IEEE Standard for Binary 4705 | Floating-Point Arithmetic, except that the conversion is always rounded 4706 | toward zero. If `a' is a NaN, the largest positive integer is returned. 4707 | Otherwise, if the conversion overflows, the largest integer with the same 4708 | sign as `a' is returned. 4709 *----------------------------------------------------------------------------*/ 4710 4711 int64 floatx80_to_int64_round_to_zero( floatx80 a STATUS_PARAM ) 4712 { 4713 flag aSign; 4714 int32 aExp, shiftCount; 4715 uint64_t aSig; 4716 int64 z; 4717 4718 aSig = extractFloatx80Frac( a ); 4719 aExp = extractFloatx80Exp( a ); 4720 aSign = extractFloatx80Sign( a ); 4721 shiftCount = aExp - 0x403E; 4722 if ( 0 <= shiftCount ) { 4723 aSig &= LIT64( 0x7FFFFFFFFFFFFFFF ); 4724 if ( ( a.high != 0xC03E ) || aSig ) { 4725 float_raise( float_flag_invalid STATUS_VAR); 4726 if ( ! aSign || ( ( aExp == 0x7FFF ) && aSig ) ) { 4727 return LIT64( 0x7FFFFFFFFFFFFFFF ); 4728 } 4729 } 4730 return (int64_t) LIT64( 0x8000000000000000 ); 4731 } 4732 else if ( aExp < 0x3FFF ) { 4733 if ( aExp | aSig ) STATUS(float_exception_flags) |= float_flag_inexact; 4734 return 0; 4735 } 4736 z = aSig>>( - shiftCount ); 4737 if ( (uint64_t) ( aSig<<( shiftCount & 63 ) ) ) { 4738 STATUS(float_exception_flags) |= float_flag_inexact; 4739 } 4740 if ( aSign ) z = - z; 4741 return z; 4742 4743 } 4744 4745 /*---------------------------------------------------------------------------- 4746 | Returns the result of converting the extended double-precision floating- 4747 | point value `a' to the single-precision floating-point format. The 4748 | conversion is performed according to the IEC/IEEE Standard for Binary 4749 | Floating-Point Arithmetic. 4750 *----------------------------------------------------------------------------*/ 4751 4752 float32 floatx80_to_float32( floatx80 a STATUS_PARAM ) 4753 { 4754 flag aSign; 4755 int32 aExp; 4756 uint64_t aSig; 4757 4758 aSig = extractFloatx80Frac( a ); 4759 aExp = extractFloatx80Exp( a ); 4760 aSign = extractFloatx80Sign( a ); 4761 if ( aExp == 0x7FFF ) { 4762 if ( (uint64_t) ( aSig<<1 ) ) { 4763 return commonNaNToFloat32( floatx80ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 4764 } 4765 return packFloat32( aSign, 0xFF, 0 ); 4766 } 4767 shift64RightJamming( aSig, 33, &aSig ); 4768 if ( aExp || aSig ) aExp -= 0x3F81; 4769 return roundAndPackFloat32( aSign, aExp, aSig STATUS_VAR ); 4770 4771 } 4772 4773 /*---------------------------------------------------------------------------- 4774 | Returns the result of converting the extended double-precision floating- 4775 | point value `a' to the double-precision floating-point format. The 4776 | conversion is performed according to the IEC/IEEE Standard for Binary 4777 | Floating-Point Arithmetic. 4778 *----------------------------------------------------------------------------*/ 4779 4780 float64 floatx80_to_float64( floatx80 a STATUS_PARAM ) 4781 { 4782 flag aSign; 4783 int32 aExp; 4784 uint64_t aSig, zSig; 4785 4786 aSig = extractFloatx80Frac( a ); 4787 aExp = extractFloatx80Exp( a ); 4788 aSign = extractFloatx80Sign( a ); 4789 if ( aExp == 0x7FFF ) { 4790 if ( (uint64_t) ( aSig<<1 ) ) { 4791 return commonNaNToFloat64( floatx80ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 4792 } 4793 return packFloat64( aSign, 0x7FF, 0 ); 4794 } 4795 shift64RightJamming( aSig, 1, &zSig ); 4796 if ( aExp || aSig ) aExp -= 0x3C01; 4797 return roundAndPackFloat64( aSign, aExp, zSig STATUS_VAR ); 4798 4799 } 4800 4801 /*---------------------------------------------------------------------------- 4802 | Returns the result of converting the extended double-precision floating- 4803 | point value `a' to the quadruple-precision floating-point format. The 4804 | conversion is performed according to the IEC/IEEE Standard for Binary 4805 | Floating-Point Arithmetic. 4806 *----------------------------------------------------------------------------*/ 4807 4808 float128 floatx80_to_float128( floatx80 a STATUS_PARAM ) 4809 { 4810 flag aSign; 4811 int_fast16_t aExp; 4812 uint64_t aSig, zSig0, zSig1; 4813 4814 aSig = extractFloatx80Frac( a ); 4815 aExp = extractFloatx80Exp( a ); 4816 aSign = extractFloatx80Sign( a ); 4817 if ( ( aExp == 0x7FFF ) && (uint64_t) ( aSig<<1 ) ) { 4818 return commonNaNToFloat128( floatx80ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 4819 } 4820 shift128Right( aSig<<1, 0, 16, &zSig0, &zSig1 ); 4821 return packFloat128( aSign, aExp, zSig0, zSig1 ); 4822 4823 } 4824 4825 /*---------------------------------------------------------------------------- 4826 | Rounds the extended double-precision floating-point value `a' to an integer, 4827 | and returns the result as an extended quadruple-precision floating-point 4828 | value. The operation is performed according to the IEC/IEEE Standard for 4829 | Binary Floating-Point Arithmetic. 4830 *----------------------------------------------------------------------------*/ 4831 4832 floatx80 floatx80_round_to_int( floatx80 a STATUS_PARAM ) 4833 { 4834 flag aSign; 4835 int32 aExp; 4836 uint64_t lastBitMask, roundBitsMask; 4837 floatx80 z; 4838 4839 aExp = extractFloatx80Exp( a ); 4840 if ( 0x403E <= aExp ) { 4841 if ( ( aExp == 0x7FFF ) && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) { 4842 return propagateFloatx80NaN( a, a STATUS_VAR ); 4843 } 4844 return a; 4845 } 4846 if ( aExp < 0x3FFF ) { 4847 if ( ( aExp == 0 ) 4848 && ( (uint64_t) ( extractFloatx80Frac( a )<<1 ) == 0 ) ) { 4849 return a; 4850 } 4851 STATUS(float_exception_flags) |= float_flag_inexact; 4852 aSign = extractFloatx80Sign( a ); 4853 switch ( STATUS(float_rounding_mode) ) { 4854 case float_round_nearest_even: 4855 if ( ( aExp == 0x3FFE ) && (uint64_t) ( extractFloatx80Frac( a )<<1 ) 4856 ) { 4857 return 4858 packFloatx80( aSign, 0x3FFF, LIT64( 0x8000000000000000 ) ); 4859 } 4860 break; 4861 case float_round_ties_away: 4862 if (aExp == 0x3FFE) { 4863 return packFloatx80(aSign, 0x3FFF, LIT64(0x8000000000000000)); 4864 } 4865 break; 4866 case float_round_down: 4867 return 4868 aSign ? 4869 packFloatx80( 1, 0x3FFF, LIT64( 0x8000000000000000 ) ) 4870 : packFloatx80( 0, 0, 0 ); 4871 case float_round_up: 4872 return 4873 aSign ? packFloatx80( 1, 0, 0 ) 4874 : packFloatx80( 0, 0x3FFF, LIT64( 0x8000000000000000 ) ); 4875 } 4876 return packFloatx80( aSign, 0, 0 ); 4877 } 4878 lastBitMask = 1; 4879 lastBitMask <<= 0x403E - aExp; 4880 roundBitsMask = lastBitMask - 1; 4881 z = a; 4882 switch (STATUS(float_rounding_mode)) { 4883 case float_round_nearest_even: 4884 z.low += lastBitMask>>1; 4885 if ((z.low & roundBitsMask) == 0) { 4886 z.low &= ~lastBitMask; 4887 } 4888 break; 4889 case float_round_ties_away: 4890 z.low += lastBitMask >> 1; 4891 break; 4892 case float_round_to_zero: 4893 break; 4894 case float_round_up: 4895 if (!extractFloatx80Sign(z)) { 4896 z.low += roundBitsMask; 4897 } 4898 break; 4899 case float_round_down: 4900 if (extractFloatx80Sign(z)) { 4901 z.low += roundBitsMask; 4902 } 4903 break; 4904 default: 4905 abort(); 4906 } 4907 z.low &= ~ roundBitsMask; 4908 if ( z.low == 0 ) { 4909 ++z.high; 4910 z.low = LIT64( 0x8000000000000000 ); 4911 } 4912 if ( z.low != a.low ) STATUS(float_exception_flags) |= float_flag_inexact; 4913 return z; 4914 4915 } 4916 4917 /*---------------------------------------------------------------------------- 4918 | Returns the result of adding the absolute values of the extended double- 4919 | precision floating-point values `a' and `b'. If `zSign' is 1, the sum is 4920 | negated before being returned. `zSign' is ignored if the result is a NaN. 4921 | The addition is performed according to the IEC/IEEE Standard for Binary 4922 | Floating-Point Arithmetic. 4923 *----------------------------------------------------------------------------*/ 4924 4925 static floatx80 addFloatx80Sigs( floatx80 a, floatx80 b, flag zSign STATUS_PARAM) 4926 { 4927 int32 aExp, bExp, zExp; 4928 uint64_t aSig, bSig, zSig0, zSig1; 4929 int32 expDiff; 4930 4931 aSig = extractFloatx80Frac( a ); 4932 aExp = extractFloatx80Exp( a ); 4933 bSig = extractFloatx80Frac( b ); 4934 bExp = extractFloatx80Exp( b ); 4935 expDiff = aExp - bExp; 4936 if ( 0 < expDiff ) { 4937 if ( aExp == 0x7FFF ) { 4938 if ( (uint64_t) ( aSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 4939 return a; 4940 } 4941 if ( bExp == 0 ) --expDiff; 4942 shift64ExtraRightJamming( bSig, 0, expDiff, &bSig, &zSig1 ); 4943 zExp = aExp; 4944 } 4945 else if ( expDiff < 0 ) { 4946 if ( bExp == 0x7FFF ) { 4947 if ( (uint64_t) ( bSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 4948 return packFloatx80( zSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 4949 } 4950 if ( aExp == 0 ) ++expDiff; 4951 shift64ExtraRightJamming( aSig, 0, - expDiff, &aSig, &zSig1 ); 4952 zExp = bExp; 4953 } 4954 else { 4955 if ( aExp == 0x7FFF ) { 4956 if ( (uint64_t) ( ( aSig | bSig )<<1 ) ) { 4957 return propagateFloatx80NaN( a, b STATUS_VAR ); 4958 } 4959 return a; 4960 } 4961 zSig1 = 0; 4962 zSig0 = aSig + bSig; 4963 if ( aExp == 0 ) { 4964 normalizeFloatx80Subnormal( zSig0, &zExp, &zSig0 ); 4965 goto roundAndPack; 4966 } 4967 zExp = aExp; 4968 goto shiftRight1; 4969 } 4970 zSig0 = aSig + bSig; 4971 if ( (int64_t) zSig0 < 0 ) goto roundAndPack; 4972 shiftRight1: 4973 shift64ExtraRightJamming( zSig0, zSig1, 1, &zSig0, &zSig1 ); 4974 zSig0 |= LIT64( 0x8000000000000000 ); 4975 ++zExp; 4976 roundAndPack: 4977 return 4978 roundAndPackFloatx80( 4979 STATUS(floatx80_rounding_precision), zSign, zExp, zSig0, zSig1 STATUS_VAR ); 4980 4981 } 4982 4983 /*---------------------------------------------------------------------------- 4984 | Returns the result of subtracting the absolute values of the extended 4985 | double-precision floating-point values `a' and `b'. If `zSign' is 1, the 4986 | difference is negated before being returned. `zSign' is ignored if the 4987 | result is a NaN. The subtraction is performed according to the IEC/IEEE 4988 | Standard for Binary Floating-Point Arithmetic. 4989 *----------------------------------------------------------------------------*/ 4990 4991 static floatx80 subFloatx80Sigs( floatx80 a, floatx80 b, flag zSign STATUS_PARAM ) 4992 { 4993 int32 aExp, bExp, zExp; 4994 uint64_t aSig, bSig, zSig0, zSig1; 4995 int32 expDiff; 4996 floatx80 z; 4997 4998 aSig = extractFloatx80Frac( a ); 4999 aExp = extractFloatx80Exp( a ); 5000 bSig = extractFloatx80Frac( b ); 5001 bExp = extractFloatx80Exp( b ); 5002 expDiff = aExp - bExp; 5003 if ( 0 < expDiff ) goto aExpBigger; 5004 if ( expDiff < 0 ) goto bExpBigger; 5005 if ( aExp == 0x7FFF ) { 5006 if ( (uint64_t) ( ( aSig | bSig )<<1 ) ) { 5007 return propagateFloatx80NaN( a, b STATUS_VAR ); 5008 } 5009 float_raise( float_flag_invalid STATUS_VAR); 5010 z.low = floatx80_default_nan_low; 5011 z.high = floatx80_default_nan_high; 5012 return z; 5013 } 5014 if ( aExp == 0 ) { 5015 aExp = 1; 5016 bExp = 1; 5017 } 5018 zSig1 = 0; 5019 if ( bSig < aSig ) goto aBigger; 5020 if ( aSig < bSig ) goto bBigger; 5021 return packFloatx80( STATUS(float_rounding_mode) == float_round_down, 0, 0 ); 5022 bExpBigger: 5023 if ( bExp == 0x7FFF ) { 5024 if ( (uint64_t) ( bSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 5025 return packFloatx80( zSign ^ 1, 0x7FFF, LIT64( 0x8000000000000000 ) ); 5026 } 5027 if ( aExp == 0 ) ++expDiff; 5028 shift128RightJamming( aSig, 0, - expDiff, &aSig, &zSig1 ); 5029 bBigger: 5030 sub128( bSig, 0, aSig, zSig1, &zSig0, &zSig1 ); 5031 zExp = bExp; 5032 zSign ^= 1; 5033 goto normalizeRoundAndPack; 5034 aExpBigger: 5035 if ( aExp == 0x7FFF ) { 5036 if ( (uint64_t) ( aSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 5037 return a; 5038 } 5039 if ( bExp == 0 ) --expDiff; 5040 shift128RightJamming( bSig, 0, expDiff, &bSig, &zSig1 ); 5041 aBigger: 5042 sub128( aSig, 0, bSig, zSig1, &zSig0, &zSig1 ); 5043 zExp = aExp; 5044 normalizeRoundAndPack: 5045 return 5046 normalizeRoundAndPackFloatx80( 5047 STATUS(floatx80_rounding_precision), zSign, zExp, zSig0, zSig1 STATUS_VAR ); 5048 5049 } 5050 5051 /*---------------------------------------------------------------------------- 5052 | Returns the result of adding the extended double-precision floating-point 5053 | values `a' and `b'. The operation is performed according to the IEC/IEEE 5054 | Standard for Binary Floating-Point Arithmetic. 5055 *----------------------------------------------------------------------------*/ 5056 5057 floatx80 floatx80_add( floatx80 a, floatx80 b STATUS_PARAM ) 5058 { 5059 flag aSign, bSign; 5060 5061 aSign = extractFloatx80Sign( a ); 5062 bSign = extractFloatx80Sign( b ); 5063 if ( aSign == bSign ) { 5064 return addFloatx80Sigs( a, b, aSign STATUS_VAR ); 5065 } 5066 else { 5067 return subFloatx80Sigs( a, b, aSign STATUS_VAR ); 5068 } 5069 5070 } 5071 5072 /*---------------------------------------------------------------------------- 5073 | Returns the result of subtracting the extended double-precision floating- 5074 | point values `a' and `b'. The operation is performed according to the 5075 | IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5076 *----------------------------------------------------------------------------*/ 5077 5078 floatx80 floatx80_sub( floatx80 a, floatx80 b STATUS_PARAM ) 5079 { 5080 flag aSign, bSign; 5081 5082 aSign = extractFloatx80Sign( a ); 5083 bSign = extractFloatx80Sign( b ); 5084 if ( aSign == bSign ) { 5085 return subFloatx80Sigs( a, b, aSign STATUS_VAR ); 5086 } 5087 else { 5088 return addFloatx80Sigs( a, b, aSign STATUS_VAR ); 5089 } 5090 5091 } 5092 5093 /*---------------------------------------------------------------------------- 5094 | Returns the result of multiplying the extended double-precision floating- 5095 | point values `a' and `b'. The operation is performed according to the 5096 | IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5097 *----------------------------------------------------------------------------*/ 5098 5099 floatx80 floatx80_mul( floatx80 a, floatx80 b STATUS_PARAM ) 5100 { 5101 flag aSign, bSign, zSign; 5102 int32 aExp, bExp, zExp; 5103 uint64_t aSig, bSig, zSig0, zSig1; 5104 floatx80 z; 5105 5106 aSig = extractFloatx80Frac( a ); 5107 aExp = extractFloatx80Exp( a ); 5108 aSign = extractFloatx80Sign( a ); 5109 bSig = extractFloatx80Frac( b ); 5110 bExp = extractFloatx80Exp( b ); 5111 bSign = extractFloatx80Sign( b ); 5112 zSign = aSign ^ bSign; 5113 if ( aExp == 0x7FFF ) { 5114 if ( (uint64_t) ( aSig<<1 ) 5115 || ( ( bExp == 0x7FFF ) && (uint64_t) ( bSig<<1 ) ) ) { 5116 return propagateFloatx80NaN( a, b STATUS_VAR ); 5117 } 5118 if ( ( bExp | bSig ) == 0 ) goto invalid; 5119 return packFloatx80( zSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 5120 } 5121 if ( bExp == 0x7FFF ) { 5122 if ( (uint64_t) ( bSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 5123 if ( ( aExp | aSig ) == 0 ) { 5124 invalid: 5125 float_raise( float_flag_invalid STATUS_VAR); 5126 z.low = floatx80_default_nan_low; 5127 z.high = floatx80_default_nan_high; 5128 return z; 5129 } 5130 return packFloatx80( zSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 5131 } 5132 if ( aExp == 0 ) { 5133 if ( aSig == 0 ) return packFloatx80( zSign, 0, 0 ); 5134 normalizeFloatx80Subnormal( aSig, &aExp, &aSig ); 5135 } 5136 if ( bExp == 0 ) { 5137 if ( bSig == 0 ) return packFloatx80( zSign, 0, 0 ); 5138 normalizeFloatx80Subnormal( bSig, &bExp, &bSig ); 5139 } 5140 zExp = aExp + bExp - 0x3FFE; 5141 mul64To128( aSig, bSig, &zSig0, &zSig1 ); 5142 if ( 0 < (int64_t) zSig0 ) { 5143 shortShift128Left( zSig0, zSig1, 1, &zSig0, &zSig1 ); 5144 --zExp; 5145 } 5146 return 5147 roundAndPackFloatx80( 5148 STATUS(floatx80_rounding_precision), zSign, zExp, zSig0, zSig1 STATUS_VAR ); 5149 5150 } 5151 5152 /*---------------------------------------------------------------------------- 5153 | Returns the result of dividing the extended double-precision floating-point 5154 | value `a' by the corresponding value `b'. The operation is performed 5155 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5156 *----------------------------------------------------------------------------*/ 5157 5158 floatx80 floatx80_div( floatx80 a, floatx80 b STATUS_PARAM ) 5159 { 5160 flag aSign, bSign, zSign; 5161 int32 aExp, bExp, zExp; 5162 uint64_t aSig, bSig, zSig0, zSig1; 5163 uint64_t rem0, rem1, rem2, term0, term1, term2; 5164 floatx80 z; 5165 5166 aSig = extractFloatx80Frac( a ); 5167 aExp = extractFloatx80Exp( a ); 5168 aSign = extractFloatx80Sign( a ); 5169 bSig = extractFloatx80Frac( b ); 5170 bExp = extractFloatx80Exp( b ); 5171 bSign = extractFloatx80Sign( b ); 5172 zSign = aSign ^ bSign; 5173 if ( aExp == 0x7FFF ) { 5174 if ( (uint64_t) ( aSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 5175 if ( bExp == 0x7FFF ) { 5176 if ( (uint64_t) ( bSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 5177 goto invalid; 5178 } 5179 return packFloatx80( zSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 5180 } 5181 if ( bExp == 0x7FFF ) { 5182 if ( (uint64_t) ( bSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 5183 return packFloatx80( zSign, 0, 0 ); 5184 } 5185 if ( bExp == 0 ) { 5186 if ( bSig == 0 ) { 5187 if ( ( aExp | aSig ) == 0 ) { 5188 invalid: 5189 float_raise( float_flag_invalid STATUS_VAR); 5190 z.low = floatx80_default_nan_low; 5191 z.high = floatx80_default_nan_high; 5192 return z; 5193 } 5194 float_raise( float_flag_divbyzero STATUS_VAR); 5195 return packFloatx80( zSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 5196 } 5197 normalizeFloatx80Subnormal( bSig, &bExp, &bSig ); 5198 } 5199 if ( aExp == 0 ) { 5200 if ( aSig == 0 ) return packFloatx80( zSign, 0, 0 ); 5201 normalizeFloatx80Subnormal( aSig, &aExp, &aSig ); 5202 } 5203 zExp = aExp - bExp + 0x3FFE; 5204 rem1 = 0; 5205 if ( bSig <= aSig ) { 5206 shift128Right( aSig, 0, 1, &aSig, &rem1 ); 5207 ++zExp; 5208 } 5209 zSig0 = estimateDiv128To64( aSig, rem1, bSig ); 5210 mul64To128( bSig, zSig0, &term0, &term1 ); 5211 sub128( aSig, rem1, term0, term1, &rem0, &rem1 ); 5212 while ( (int64_t) rem0 < 0 ) { 5213 --zSig0; 5214 add128( rem0, rem1, 0, bSig, &rem0, &rem1 ); 5215 } 5216 zSig1 = estimateDiv128To64( rem1, 0, bSig ); 5217 if ( (uint64_t) ( zSig1<<1 ) <= 8 ) { 5218 mul64To128( bSig, zSig1, &term1, &term2 ); 5219 sub128( rem1, 0, term1, term2, &rem1, &rem2 ); 5220 while ( (int64_t) rem1 < 0 ) { 5221 --zSig1; 5222 add128( rem1, rem2, 0, bSig, &rem1, &rem2 ); 5223 } 5224 zSig1 |= ( ( rem1 | rem2 ) != 0 ); 5225 } 5226 return 5227 roundAndPackFloatx80( 5228 STATUS(floatx80_rounding_precision), zSign, zExp, zSig0, zSig1 STATUS_VAR ); 5229 5230 } 5231 5232 /*---------------------------------------------------------------------------- 5233 | Returns the remainder of the extended double-precision floating-point value 5234 | `a' with respect to the corresponding value `b'. The operation is performed 5235 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5236 *----------------------------------------------------------------------------*/ 5237 5238 floatx80 floatx80_rem( floatx80 a, floatx80 b STATUS_PARAM ) 5239 { 5240 flag aSign, zSign; 5241 int32 aExp, bExp, expDiff; 5242 uint64_t aSig0, aSig1, bSig; 5243 uint64_t q, term0, term1, alternateASig0, alternateASig1; 5244 floatx80 z; 5245 5246 aSig0 = extractFloatx80Frac( a ); 5247 aExp = extractFloatx80Exp( a ); 5248 aSign = extractFloatx80Sign( a ); 5249 bSig = extractFloatx80Frac( b ); 5250 bExp = extractFloatx80Exp( b ); 5251 if ( aExp == 0x7FFF ) { 5252 if ( (uint64_t) ( aSig0<<1 ) 5253 || ( ( bExp == 0x7FFF ) && (uint64_t) ( bSig<<1 ) ) ) { 5254 return propagateFloatx80NaN( a, b STATUS_VAR ); 5255 } 5256 goto invalid; 5257 } 5258 if ( bExp == 0x7FFF ) { 5259 if ( (uint64_t) ( bSig<<1 ) ) return propagateFloatx80NaN( a, b STATUS_VAR ); 5260 return a; 5261 } 5262 if ( bExp == 0 ) { 5263 if ( bSig == 0 ) { 5264 invalid: 5265 float_raise( float_flag_invalid STATUS_VAR); 5266 z.low = floatx80_default_nan_low; 5267 z.high = floatx80_default_nan_high; 5268 return z; 5269 } 5270 normalizeFloatx80Subnormal( bSig, &bExp, &bSig ); 5271 } 5272 if ( aExp == 0 ) { 5273 if ( (uint64_t) ( aSig0<<1 ) == 0 ) return a; 5274 normalizeFloatx80Subnormal( aSig0, &aExp, &aSig0 ); 5275 } 5276 bSig |= LIT64( 0x8000000000000000 ); 5277 zSign = aSign; 5278 expDiff = aExp - bExp; 5279 aSig1 = 0; 5280 if ( expDiff < 0 ) { 5281 if ( expDiff < -1 ) return a; 5282 shift128Right( aSig0, 0, 1, &aSig0, &aSig1 ); 5283 expDiff = 0; 5284 } 5285 q = ( bSig <= aSig0 ); 5286 if ( q ) aSig0 -= bSig; 5287 expDiff -= 64; 5288 while ( 0 < expDiff ) { 5289 q = estimateDiv128To64( aSig0, aSig1, bSig ); 5290 q = ( 2 < q ) ? q - 2 : 0; 5291 mul64To128( bSig, q, &term0, &term1 ); 5292 sub128( aSig0, aSig1, term0, term1, &aSig0, &aSig1 ); 5293 shortShift128Left( aSig0, aSig1, 62, &aSig0, &aSig1 ); 5294 expDiff -= 62; 5295 } 5296 expDiff += 64; 5297 if ( 0 < expDiff ) { 5298 q = estimateDiv128To64( aSig0, aSig1, bSig ); 5299 q = ( 2 < q ) ? q - 2 : 0; 5300 q >>= 64 - expDiff; 5301 mul64To128( bSig, q<<( 64 - expDiff ), &term0, &term1 ); 5302 sub128( aSig0, aSig1, term0, term1, &aSig0, &aSig1 ); 5303 shortShift128Left( 0, bSig, 64 - expDiff, &term0, &term1 ); 5304 while ( le128( term0, term1, aSig0, aSig1 ) ) { 5305 ++q; 5306 sub128( aSig0, aSig1, term0, term1, &aSig0, &aSig1 ); 5307 } 5308 } 5309 else { 5310 term1 = 0; 5311 term0 = bSig; 5312 } 5313 sub128( term0, term1, aSig0, aSig1, &alternateASig0, &alternateASig1 ); 5314 if ( lt128( alternateASig0, alternateASig1, aSig0, aSig1 ) 5315 || ( eq128( alternateASig0, alternateASig1, aSig0, aSig1 ) 5316 && ( q & 1 ) ) 5317 ) { 5318 aSig0 = alternateASig0; 5319 aSig1 = alternateASig1; 5320 zSign = ! zSign; 5321 } 5322 return 5323 normalizeRoundAndPackFloatx80( 5324 80, zSign, bExp + expDiff, aSig0, aSig1 STATUS_VAR ); 5325 5326 } 5327 5328 /*---------------------------------------------------------------------------- 5329 | Returns the square root of the extended double-precision floating-point 5330 | value `a'. The operation is performed according to the IEC/IEEE Standard 5331 | for Binary Floating-Point Arithmetic. 5332 *----------------------------------------------------------------------------*/ 5333 5334 floatx80 floatx80_sqrt( floatx80 a STATUS_PARAM ) 5335 { 5336 flag aSign; 5337 int32 aExp, zExp; 5338 uint64_t aSig0, aSig1, zSig0, zSig1, doubleZSig0; 5339 uint64_t rem0, rem1, rem2, rem3, term0, term1, term2, term3; 5340 floatx80 z; 5341 5342 aSig0 = extractFloatx80Frac( a ); 5343 aExp = extractFloatx80Exp( a ); 5344 aSign = extractFloatx80Sign( a ); 5345 if ( aExp == 0x7FFF ) { 5346 if ( (uint64_t) ( aSig0<<1 ) ) return propagateFloatx80NaN( a, a STATUS_VAR ); 5347 if ( ! aSign ) return a; 5348 goto invalid; 5349 } 5350 if ( aSign ) { 5351 if ( ( aExp | aSig0 ) == 0 ) return a; 5352 invalid: 5353 float_raise( float_flag_invalid STATUS_VAR); 5354 z.low = floatx80_default_nan_low; 5355 z.high = floatx80_default_nan_high; 5356 return z; 5357 } 5358 if ( aExp == 0 ) { 5359 if ( aSig0 == 0 ) return packFloatx80( 0, 0, 0 ); 5360 normalizeFloatx80Subnormal( aSig0, &aExp, &aSig0 ); 5361 } 5362 zExp = ( ( aExp - 0x3FFF )>>1 ) + 0x3FFF; 5363 zSig0 = estimateSqrt32( aExp, aSig0>>32 ); 5364 shift128Right( aSig0, 0, 2 + ( aExp & 1 ), &aSig0, &aSig1 ); 5365 zSig0 = estimateDiv128To64( aSig0, aSig1, zSig0<<32 ) + ( zSig0<<30 ); 5366 doubleZSig0 = zSig0<<1; 5367 mul64To128( zSig0, zSig0, &term0, &term1 ); 5368 sub128( aSig0, aSig1, term0, term1, &rem0, &rem1 ); 5369 while ( (int64_t) rem0 < 0 ) { 5370 --zSig0; 5371 doubleZSig0 -= 2; 5372 add128( rem0, rem1, zSig0>>63, doubleZSig0 | 1, &rem0, &rem1 ); 5373 } 5374 zSig1 = estimateDiv128To64( rem1, 0, doubleZSig0 ); 5375 if ( ( zSig1 & LIT64( 0x3FFFFFFFFFFFFFFF ) ) <= 5 ) { 5376 if ( zSig1 == 0 ) zSig1 = 1; 5377 mul64To128( doubleZSig0, zSig1, &term1, &term2 ); 5378 sub128( rem1, 0, term1, term2, &rem1, &rem2 ); 5379 mul64To128( zSig1, zSig1, &term2, &term3 ); 5380 sub192( rem1, rem2, 0, 0, term2, term3, &rem1, &rem2, &rem3 ); 5381 while ( (int64_t) rem1 < 0 ) { 5382 --zSig1; 5383 shortShift128Left( 0, zSig1, 1, &term2, &term3 ); 5384 term3 |= 1; 5385 term2 |= doubleZSig0; 5386 add192( rem1, rem2, rem3, 0, term2, term3, &rem1, &rem2, &rem3 ); 5387 } 5388 zSig1 |= ( ( rem1 | rem2 | rem3 ) != 0 ); 5389 } 5390 shortShift128Left( 0, zSig1, 1, &zSig0, &zSig1 ); 5391 zSig0 |= doubleZSig0; 5392 return 5393 roundAndPackFloatx80( 5394 STATUS(floatx80_rounding_precision), 0, zExp, zSig0, zSig1 STATUS_VAR ); 5395 5396 } 5397 5398 /*---------------------------------------------------------------------------- 5399 | Returns 1 if the extended double-precision floating-point value `a' is equal 5400 | to the corresponding value `b', and 0 otherwise. The invalid exception is 5401 | raised if either operand is a NaN. Otherwise, the comparison is performed 5402 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5403 *----------------------------------------------------------------------------*/ 5404 5405 int floatx80_eq( floatx80 a, floatx80 b STATUS_PARAM ) 5406 { 5407 5408 if ( ( ( extractFloatx80Exp( a ) == 0x7FFF ) 5409 && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) 5410 || ( ( extractFloatx80Exp( b ) == 0x7FFF ) 5411 && (uint64_t) ( extractFloatx80Frac( b )<<1 ) ) 5412 ) { 5413 float_raise( float_flag_invalid STATUS_VAR); 5414 return 0; 5415 } 5416 return 5417 ( a.low == b.low ) 5418 && ( ( a.high == b.high ) 5419 || ( ( a.low == 0 ) 5420 && ( (uint16_t) ( ( a.high | b.high )<<1 ) == 0 ) ) 5421 ); 5422 5423 } 5424 5425 /*---------------------------------------------------------------------------- 5426 | Returns 1 if the extended double-precision floating-point value `a' is 5427 | less than or equal to the corresponding value `b', and 0 otherwise. The 5428 | invalid exception is raised if either operand is a NaN. The comparison is 5429 | performed according to the IEC/IEEE Standard for Binary Floating-Point 5430 | Arithmetic. 5431 *----------------------------------------------------------------------------*/ 5432 5433 int floatx80_le( floatx80 a, floatx80 b STATUS_PARAM ) 5434 { 5435 flag aSign, bSign; 5436 5437 if ( ( ( extractFloatx80Exp( a ) == 0x7FFF ) 5438 && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) 5439 || ( ( extractFloatx80Exp( b ) == 0x7FFF ) 5440 && (uint64_t) ( extractFloatx80Frac( b )<<1 ) ) 5441 ) { 5442 float_raise( float_flag_invalid STATUS_VAR); 5443 return 0; 5444 } 5445 aSign = extractFloatx80Sign( a ); 5446 bSign = extractFloatx80Sign( b ); 5447 if ( aSign != bSign ) { 5448 return 5449 aSign 5450 || ( ( ( (uint16_t) ( ( a.high | b.high )<<1 ) ) | a.low | b.low ) 5451 == 0 ); 5452 } 5453 return 5454 aSign ? le128( b.high, b.low, a.high, a.low ) 5455 : le128( a.high, a.low, b.high, b.low ); 5456 5457 } 5458 5459 /*---------------------------------------------------------------------------- 5460 | Returns 1 if the extended double-precision floating-point value `a' is 5461 | less than the corresponding value `b', and 0 otherwise. The invalid 5462 | exception is raised if either operand is a NaN. The comparison is performed 5463 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5464 *----------------------------------------------------------------------------*/ 5465 5466 int floatx80_lt( floatx80 a, floatx80 b STATUS_PARAM ) 5467 { 5468 flag aSign, bSign; 5469 5470 if ( ( ( extractFloatx80Exp( a ) == 0x7FFF ) 5471 && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) 5472 || ( ( extractFloatx80Exp( b ) == 0x7FFF ) 5473 && (uint64_t) ( extractFloatx80Frac( b )<<1 ) ) 5474 ) { 5475 float_raise( float_flag_invalid STATUS_VAR); 5476 return 0; 5477 } 5478 aSign = extractFloatx80Sign( a ); 5479 bSign = extractFloatx80Sign( b ); 5480 if ( aSign != bSign ) { 5481 return 5482 aSign 5483 && ( ( ( (uint16_t) ( ( a.high | b.high )<<1 ) ) | a.low | b.low ) 5484 != 0 ); 5485 } 5486 return 5487 aSign ? lt128( b.high, b.low, a.high, a.low ) 5488 : lt128( a.high, a.low, b.high, b.low ); 5489 5490 } 5491 5492 /*---------------------------------------------------------------------------- 5493 | Returns 1 if the extended double-precision floating-point values `a' and `b' 5494 | cannot be compared, and 0 otherwise. The invalid exception is raised if 5495 | either operand is a NaN. The comparison is performed according to the 5496 | IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5497 *----------------------------------------------------------------------------*/ 5498 int floatx80_unordered( floatx80 a, floatx80 b STATUS_PARAM ) 5499 { 5500 if ( ( ( extractFloatx80Exp( a ) == 0x7FFF ) 5501 && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) 5502 || ( ( extractFloatx80Exp( b ) == 0x7FFF ) 5503 && (uint64_t) ( extractFloatx80Frac( b )<<1 ) ) 5504 ) { 5505 float_raise( float_flag_invalid STATUS_VAR); 5506 return 1; 5507 } 5508 return 0; 5509 } 5510 5511 /*---------------------------------------------------------------------------- 5512 | Returns 1 if the extended double-precision floating-point value `a' is 5513 | equal to the corresponding value `b', and 0 otherwise. Quiet NaNs do not 5514 | cause an exception. The comparison is performed according to the IEC/IEEE 5515 | Standard for Binary Floating-Point Arithmetic. 5516 *----------------------------------------------------------------------------*/ 5517 5518 int floatx80_eq_quiet( floatx80 a, floatx80 b STATUS_PARAM ) 5519 { 5520 5521 if ( ( ( extractFloatx80Exp( a ) == 0x7FFF ) 5522 && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) 5523 || ( ( extractFloatx80Exp( b ) == 0x7FFF ) 5524 && (uint64_t) ( extractFloatx80Frac( b )<<1 ) ) 5525 ) { 5526 if ( floatx80_is_signaling_nan( a ) 5527 || floatx80_is_signaling_nan( b ) ) { 5528 float_raise( float_flag_invalid STATUS_VAR); 5529 } 5530 return 0; 5531 } 5532 return 5533 ( a.low == b.low ) 5534 && ( ( a.high == b.high ) 5535 || ( ( a.low == 0 ) 5536 && ( (uint16_t) ( ( a.high | b.high )<<1 ) == 0 ) ) 5537 ); 5538 5539 } 5540 5541 /*---------------------------------------------------------------------------- 5542 | Returns 1 if the extended double-precision floating-point value `a' is less 5543 | than or equal to the corresponding value `b', and 0 otherwise. Quiet NaNs 5544 | do not cause an exception. Otherwise, the comparison is performed according 5545 | to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5546 *----------------------------------------------------------------------------*/ 5547 5548 int floatx80_le_quiet( floatx80 a, floatx80 b STATUS_PARAM ) 5549 { 5550 flag aSign, bSign; 5551 5552 if ( ( ( extractFloatx80Exp( a ) == 0x7FFF ) 5553 && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) 5554 || ( ( extractFloatx80Exp( b ) == 0x7FFF ) 5555 && (uint64_t) ( extractFloatx80Frac( b )<<1 ) ) 5556 ) { 5557 if ( floatx80_is_signaling_nan( a ) 5558 || floatx80_is_signaling_nan( b ) ) { 5559 float_raise( float_flag_invalid STATUS_VAR); 5560 } 5561 return 0; 5562 } 5563 aSign = extractFloatx80Sign( a ); 5564 bSign = extractFloatx80Sign( b ); 5565 if ( aSign != bSign ) { 5566 return 5567 aSign 5568 || ( ( ( (uint16_t) ( ( a.high | b.high )<<1 ) ) | a.low | b.low ) 5569 == 0 ); 5570 } 5571 return 5572 aSign ? le128( b.high, b.low, a.high, a.low ) 5573 : le128( a.high, a.low, b.high, b.low ); 5574 5575 } 5576 5577 /*---------------------------------------------------------------------------- 5578 | Returns 1 if the extended double-precision floating-point value `a' is less 5579 | than the corresponding value `b', and 0 otherwise. Quiet NaNs do not cause 5580 | an exception. Otherwise, the comparison is performed according to the 5581 | IEC/IEEE Standard for Binary Floating-Point Arithmetic. 5582 *----------------------------------------------------------------------------*/ 5583 5584 int floatx80_lt_quiet( floatx80 a, floatx80 b STATUS_PARAM ) 5585 { 5586 flag aSign, bSign; 5587 5588 if ( ( ( extractFloatx80Exp( a ) == 0x7FFF ) 5589 && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) 5590 || ( ( extractFloatx80Exp( b ) == 0x7FFF ) 5591 && (uint64_t) ( extractFloatx80Frac( b )<<1 ) ) 5592 ) { 5593 if ( floatx80_is_signaling_nan( a ) 5594 || floatx80_is_signaling_nan( b ) ) { 5595 float_raise( float_flag_invalid STATUS_VAR); 5596 } 5597 return 0; 5598 } 5599 aSign = extractFloatx80Sign( a ); 5600 bSign = extractFloatx80Sign( b ); 5601 if ( aSign != bSign ) { 5602 return 5603 aSign 5604 && ( ( ( (uint16_t) ( ( a.high | b.high )<<1 ) ) | a.low | b.low ) 5605 != 0 ); 5606 } 5607 return 5608 aSign ? lt128( b.high, b.low, a.high, a.low ) 5609 : lt128( a.high, a.low, b.high, b.low ); 5610 5611 } 5612 5613 /*---------------------------------------------------------------------------- 5614 | Returns 1 if the extended double-precision floating-point values `a' and `b' 5615 | cannot be compared, and 0 otherwise. Quiet NaNs do not cause an exception. 5616 | The comparison is performed according to the IEC/IEEE Standard for Binary 5617 | Floating-Point Arithmetic. 5618 *----------------------------------------------------------------------------*/ 5619 int floatx80_unordered_quiet( floatx80 a, floatx80 b STATUS_PARAM ) 5620 { 5621 if ( ( ( extractFloatx80Exp( a ) == 0x7FFF ) 5622 && (uint64_t) ( extractFloatx80Frac( a )<<1 ) ) 5623 || ( ( extractFloatx80Exp( b ) == 0x7FFF ) 5624 && (uint64_t) ( extractFloatx80Frac( b )<<1 ) ) 5625 ) { 5626 if ( floatx80_is_signaling_nan( a ) 5627 || floatx80_is_signaling_nan( b ) ) { 5628 float_raise( float_flag_invalid STATUS_VAR); 5629 } 5630 return 1; 5631 } 5632 return 0; 5633 } 5634 5635 /*---------------------------------------------------------------------------- 5636 | Returns the result of converting the quadruple-precision floating-point 5637 | value `a' to the 32-bit two's complement integer format. The conversion 5638 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 5639 | Arithmetic---which means in particular that the conversion is rounded 5640 | according to the current rounding mode. If `a' is a NaN, the largest 5641 | positive integer is returned. Otherwise, if the conversion overflows, the 5642 | largest integer with the same sign as `a' is returned. 5643 *----------------------------------------------------------------------------*/ 5644 5645 int32 float128_to_int32( float128 a STATUS_PARAM ) 5646 { 5647 flag aSign; 5648 int32 aExp, shiftCount; 5649 uint64_t aSig0, aSig1; 5650 5651 aSig1 = extractFloat128Frac1( a ); 5652 aSig0 = extractFloat128Frac0( a ); 5653 aExp = extractFloat128Exp( a ); 5654 aSign = extractFloat128Sign( a ); 5655 if ( ( aExp == 0x7FFF ) && ( aSig0 | aSig1 ) ) aSign = 0; 5656 if ( aExp ) aSig0 |= LIT64( 0x0001000000000000 ); 5657 aSig0 |= ( aSig1 != 0 ); 5658 shiftCount = 0x4028 - aExp; 5659 if ( 0 < shiftCount ) shift64RightJamming( aSig0, shiftCount, &aSig0 ); 5660 return roundAndPackInt32( aSign, aSig0 STATUS_VAR ); 5661 5662 } 5663 5664 /*---------------------------------------------------------------------------- 5665 | Returns the result of converting the quadruple-precision floating-point 5666 | value `a' to the 32-bit two's complement integer format. The conversion 5667 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 5668 | Arithmetic, except that the conversion is always rounded toward zero. If 5669 | `a' is a NaN, the largest positive integer is returned. Otherwise, if the 5670 | conversion overflows, the largest integer with the same sign as `a' is 5671 | returned. 5672 *----------------------------------------------------------------------------*/ 5673 5674 int32 float128_to_int32_round_to_zero( float128 a STATUS_PARAM ) 5675 { 5676 flag aSign; 5677 int32 aExp, shiftCount; 5678 uint64_t aSig0, aSig1, savedASig; 5679 int32_t z; 5680 5681 aSig1 = extractFloat128Frac1( a ); 5682 aSig0 = extractFloat128Frac0( a ); 5683 aExp = extractFloat128Exp( a ); 5684 aSign = extractFloat128Sign( a ); 5685 aSig0 |= ( aSig1 != 0 ); 5686 if ( 0x401E < aExp ) { 5687 if ( ( aExp == 0x7FFF ) && aSig0 ) aSign = 0; 5688 goto invalid; 5689 } 5690 else if ( aExp < 0x3FFF ) { 5691 if ( aExp || aSig0 ) STATUS(float_exception_flags) |= float_flag_inexact; 5692 return 0; 5693 } 5694 aSig0 |= LIT64( 0x0001000000000000 ); 5695 shiftCount = 0x402F - aExp; 5696 savedASig = aSig0; 5697 aSig0 >>= shiftCount; 5698 z = aSig0; 5699 if ( aSign ) z = - z; 5700 if ( ( z < 0 ) ^ aSign ) { 5701 invalid: 5702 float_raise( float_flag_invalid STATUS_VAR); 5703 return aSign ? (int32_t) 0x80000000 : 0x7FFFFFFF; 5704 } 5705 if ( ( aSig0<<shiftCount ) != savedASig ) { 5706 STATUS(float_exception_flags) |= float_flag_inexact; 5707 } 5708 return z; 5709 5710 } 5711 5712 /*---------------------------------------------------------------------------- 5713 | Returns the result of converting the quadruple-precision floating-point 5714 | value `a' to the 64-bit two's complement integer format. The conversion 5715 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 5716 | Arithmetic---which means in particular that the conversion is rounded 5717 | according to the current rounding mode. If `a' is a NaN, the largest 5718 | positive integer is returned. Otherwise, if the conversion overflows, the 5719 | largest integer with the same sign as `a' is returned. 5720 *----------------------------------------------------------------------------*/ 5721 5722 int64 float128_to_int64( float128 a STATUS_PARAM ) 5723 { 5724 flag aSign; 5725 int32 aExp, shiftCount; 5726 uint64_t aSig0, aSig1; 5727 5728 aSig1 = extractFloat128Frac1( a ); 5729 aSig0 = extractFloat128Frac0( a ); 5730 aExp = extractFloat128Exp( a ); 5731 aSign = extractFloat128Sign( a ); 5732 if ( aExp ) aSig0 |= LIT64( 0x0001000000000000 ); 5733 shiftCount = 0x402F - aExp; 5734 if ( shiftCount <= 0 ) { 5735 if ( 0x403E < aExp ) { 5736 float_raise( float_flag_invalid STATUS_VAR); 5737 if ( ! aSign 5738 || ( ( aExp == 0x7FFF ) 5739 && ( aSig1 || ( aSig0 != LIT64( 0x0001000000000000 ) ) ) 5740 ) 5741 ) { 5742 return LIT64( 0x7FFFFFFFFFFFFFFF ); 5743 } 5744 return (int64_t) LIT64( 0x8000000000000000 ); 5745 } 5746 shortShift128Left( aSig0, aSig1, - shiftCount, &aSig0, &aSig1 ); 5747 } 5748 else { 5749 shift64ExtraRightJamming( aSig0, aSig1, shiftCount, &aSig0, &aSig1 ); 5750 } 5751 return roundAndPackInt64( aSign, aSig0, aSig1 STATUS_VAR ); 5752 5753 } 5754 5755 /*---------------------------------------------------------------------------- 5756 | Returns the result of converting the quadruple-precision floating-point 5757 | value `a' to the 64-bit two's complement integer format. The conversion 5758 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 5759 | Arithmetic, except that the conversion is always rounded toward zero. 5760 | If `a' is a NaN, the largest positive integer is returned. Otherwise, if 5761 | the conversion overflows, the largest integer with the same sign as `a' is 5762 | returned. 5763 *----------------------------------------------------------------------------*/ 5764 5765 int64 float128_to_int64_round_to_zero( float128 a STATUS_PARAM ) 5766 { 5767 flag aSign; 5768 int32 aExp, shiftCount; 5769 uint64_t aSig0, aSig1; 5770 int64 z; 5771 5772 aSig1 = extractFloat128Frac1( a ); 5773 aSig0 = extractFloat128Frac0( a ); 5774 aExp = extractFloat128Exp( a ); 5775 aSign = extractFloat128Sign( a ); 5776 if ( aExp ) aSig0 |= LIT64( 0x0001000000000000 ); 5777 shiftCount = aExp - 0x402F; 5778 if ( 0 < shiftCount ) { 5779 if ( 0x403E <= aExp ) { 5780 aSig0 &= LIT64( 0x0000FFFFFFFFFFFF ); 5781 if ( ( a.high == LIT64( 0xC03E000000000000 ) ) 5782 && ( aSig1 < LIT64( 0x0002000000000000 ) ) ) { 5783 if ( aSig1 ) STATUS(float_exception_flags) |= float_flag_inexact; 5784 } 5785 else { 5786 float_raise( float_flag_invalid STATUS_VAR); 5787 if ( ! aSign || ( ( aExp == 0x7FFF ) && ( aSig0 | aSig1 ) ) ) { 5788 return LIT64( 0x7FFFFFFFFFFFFFFF ); 5789 } 5790 } 5791 return (int64_t) LIT64( 0x8000000000000000 ); 5792 } 5793 z = ( aSig0<<shiftCount ) | ( aSig1>>( ( - shiftCount ) & 63 ) ); 5794 if ( (uint64_t) ( aSig1<<shiftCount ) ) { 5795 STATUS(float_exception_flags) |= float_flag_inexact; 5796 } 5797 } 5798 else { 5799 if ( aExp < 0x3FFF ) { 5800 if ( aExp | aSig0 | aSig1 ) { 5801 STATUS(float_exception_flags) |= float_flag_inexact; 5802 } 5803 return 0; 5804 } 5805 z = aSig0>>( - shiftCount ); 5806 if ( aSig1 5807 || ( shiftCount && (uint64_t) ( aSig0<<( shiftCount & 63 ) ) ) ) { 5808 STATUS(float_exception_flags) |= float_flag_inexact; 5809 } 5810 } 5811 if ( aSign ) z = - z; 5812 return z; 5813 5814 } 5815 5816 /*---------------------------------------------------------------------------- 5817 | Returns the result of converting the quadruple-precision floating-point 5818 | value `a' to the single-precision floating-point format. The conversion 5819 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 5820 | Arithmetic. 5821 *----------------------------------------------------------------------------*/ 5822 5823 float32 float128_to_float32( float128 a STATUS_PARAM ) 5824 { 5825 flag aSign; 5826 int32 aExp; 5827 uint64_t aSig0, aSig1; 5828 uint32_t zSig; 5829 5830 aSig1 = extractFloat128Frac1( a ); 5831 aSig0 = extractFloat128Frac0( a ); 5832 aExp = extractFloat128Exp( a ); 5833 aSign = extractFloat128Sign( a ); 5834 if ( aExp == 0x7FFF ) { 5835 if ( aSig0 | aSig1 ) { 5836 return commonNaNToFloat32( float128ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 5837 } 5838 return packFloat32( aSign, 0xFF, 0 ); 5839 } 5840 aSig0 |= ( aSig1 != 0 ); 5841 shift64RightJamming( aSig0, 18, &aSig0 ); 5842 zSig = aSig0; 5843 if ( aExp || zSig ) { 5844 zSig |= 0x40000000; 5845 aExp -= 0x3F81; 5846 } 5847 return roundAndPackFloat32( aSign, aExp, zSig STATUS_VAR ); 5848 5849 } 5850 5851 /*---------------------------------------------------------------------------- 5852 | Returns the result of converting the quadruple-precision floating-point 5853 | value `a' to the double-precision floating-point format. The conversion 5854 | is performed according to the IEC/IEEE Standard for Binary Floating-Point 5855 | Arithmetic. 5856 *----------------------------------------------------------------------------*/ 5857 5858 float64 float128_to_float64( float128 a STATUS_PARAM ) 5859 { 5860 flag aSign; 5861 int32 aExp; 5862 uint64_t aSig0, aSig1; 5863 5864 aSig1 = extractFloat128Frac1( a ); 5865 aSig0 = extractFloat128Frac0( a ); 5866 aExp = extractFloat128Exp( a ); 5867 aSign = extractFloat128Sign( a ); 5868 if ( aExp == 0x7FFF ) { 5869 if ( aSig0 | aSig1 ) { 5870 return commonNaNToFloat64( float128ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 5871 } 5872 return packFloat64( aSign, 0x7FF, 0 ); 5873 } 5874 shortShift128Left( aSig0, aSig1, 14, &aSig0, &aSig1 ); 5875 aSig0 |= ( aSig1 != 0 ); 5876 if ( aExp || aSig0 ) { 5877 aSig0 |= LIT64( 0x4000000000000000 ); 5878 aExp -= 0x3C01; 5879 } 5880 return roundAndPackFloat64( aSign, aExp, aSig0 STATUS_VAR ); 5881 5882 } 5883 5884 /*---------------------------------------------------------------------------- 5885 | Returns the result of converting the quadruple-precision floating-point 5886 | value `a' to the extended double-precision floating-point format. The 5887 | conversion is performed according to the IEC/IEEE Standard for Binary 5888 | Floating-Point Arithmetic. 5889 *----------------------------------------------------------------------------*/ 5890 5891 floatx80 float128_to_floatx80( float128 a STATUS_PARAM ) 5892 { 5893 flag aSign; 5894 int32 aExp; 5895 uint64_t aSig0, aSig1; 5896 5897 aSig1 = extractFloat128Frac1( a ); 5898 aSig0 = extractFloat128Frac0( a ); 5899 aExp = extractFloat128Exp( a ); 5900 aSign = extractFloat128Sign( a ); 5901 if ( aExp == 0x7FFF ) { 5902 if ( aSig0 | aSig1 ) { 5903 return commonNaNToFloatx80( float128ToCommonNaN( a STATUS_VAR ) STATUS_VAR ); 5904 } 5905 return packFloatx80( aSign, 0x7FFF, LIT64( 0x8000000000000000 ) ); 5906 } 5907 if ( aExp == 0 ) { 5908 if ( ( aSig0 | aSig1 ) == 0 ) return packFloatx80( aSign, 0, 0 ); 5909 normalizeFloat128Subnormal( aSig0, aSig1, &aExp, &aSig0, &aSig1 ); 5910 } 5911 else { 5912 aSig0 |= LIT64( 0x0001000000000000 ); 5913 } 5914 shortShift128Left( aSig0, aSig1, 15, &aSig0, &aSig1 ); 5915 return roundAndPackFloatx80( 80, aSign, aExp, aSig0, aSig1 STATUS_VAR ); 5916 5917 } 5918 5919 /*---------------------------------------------------------------------------- 5920 | Rounds the quadruple-precision floating-point value `a' to an integer, and 5921 | returns the result as a quadruple-precision floating-point value. The 5922 | operation is performed according to the IEC/IEEE Standard for Binary 5923 | Floating-Point Arithmetic. 5924 *----------------------------------------------------------------------------*/ 5925 5926 float128 float128_round_to_int( float128 a STATUS_PARAM ) 5927 { 5928 flag aSign; 5929 int32 aExp; 5930 uint64_t lastBitMask, roundBitsMask; 5931 float128 z; 5932 5933 aExp = extractFloat128Exp( a ); 5934 if ( 0x402F <= aExp ) { 5935 if ( 0x406F <= aExp ) { 5936 if ( ( aExp == 0x7FFF ) 5937 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) 5938 ) { 5939 return propagateFloat128NaN( a, a STATUS_VAR ); 5940 } 5941 return a; 5942 } 5943 lastBitMask = 1; 5944 lastBitMask = ( lastBitMask<<( 0x406E - aExp ) )<<1; 5945 roundBitsMask = lastBitMask - 1; 5946 z = a; 5947 switch (STATUS(float_rounding_mode)) { 5948 case float_round_nearest_even: 5949 if ( lastBitMask ) { 5950 add128( z.high, z.low, 0, lastBitMask>>1, &z.high, &z.low ); 5951 if ( ( z.low & roundBitsMask ) == 0 ) z.low &= ~ lastBitMask; 5952 } 5953 else { 5954 if ( (int64_t) z.low < 0 ) { 5955 ++z.high; 5956 if ( (uint64_t) ( z.low<<1 ) == 0 ) z.high &= ~1; 5957 } 5958 } 5959 break; 5960 case float_round_ties_away: 5961 if (lastBitMask) { 5962 add128(z.high, z.low, 0, lastBitMask >> 1, &z.high, &z.low); 5963 } else { 5964 if ((int64_t) z.low < 0) { 5965 ++z.high; 5966 } 5967 } 5968 break; 5969 case float_round_to_zero: 5970 break; 5971 case float_round_up: 5972 if (!extractFloat128Sign(z)) { 5973 add128(z.high, z.low, 0, roundBitsMask, &z.high, &z.low); 5974 } 5975 break; 5976 case float_round_down: 5977 if (extractFloat128Sign(z)) { 5978 add128(z.high, z.low, 0, roundBitsMask, &z.high, &z.low); 5979 } 5980 break; 5981 default: 5982 abort(); 5983 } 5984 z.low &= ~ roundBitsMask; 5985 } 5986 else { 5987 if ( aExp < 0x3FFF ) { 5988 if ( ( ( (uint64_t) ( a.high<<1 ) ) | a.low ) == 0 ) return a; 5989 STATUS(float_exception_flags) |= float_flag_inexact; 5990 aSign = extractFloat128Sign( a ); 5991 switch ( STATUS(float_rounding_mode) ) { 5992 case float_round_nearest_even: 5993 if ( ( aExp == 0x3FFE ) 5994 && ( extractFloat128Frac0( a ) 5995 | extractFloat128Frac1( a ) ) 5996 ) { 5997 return packFloat128( aSign, 0x3FFF, 0, 0 ); 5998 } 5999 break; 6000 case float_round_ties_away: 6001 if (aExp == 0x3FFE) { 6002 return packFloat128(aSign, 0x3FFF, 0, 0); 6003 } 6004 break; 6005 case float_round_down: 6006 return 6007 aSign ? packFloat128( 1, 0x3FFF, 0, 0 ) 6008 : packFloat128( 0, 0, 0, 0 ); 6009 case float_round_up: 6010 return 6011 aSign ? packFloat128( 1, 0, 0, 0 ) 6012 : packFloat128( 0, 0x3FFF, 0, 0 ); 6013 } 6014 return packFloat128( aSign, 0, 0, 0 ); 6015 } 6016 lastBitMask = 1; 6017 lastBitMask <<= 0x402F - aExp; 6018 roundBitsMask = lastBitMask - 1; 6019 z.low = 0; 6020 z.high = a.high; 6021 switch (STATUS(float_rounding_mode)) { 6022 case float_round_nearest_even: 6023 z.high += lastBitMask>>1; 6024 if ( ( ( z.high & roundBitsMask ) | a.low ) == 0 ) { 6025 z.high &= ~ lastBitMask; 6026 } 6027 break; 6028 case float_round_ties_away: 6029 z.high += lastBitMask>>1; 6030 break; 6031 case float_round_to_zero: 6032 break; 6033 case float_round_up: 6034 if (!extractFloat128Sign(z)) { 6035 z.high |= ( a.low != 0 ); 6036 z.high += roundBitsMask; 6037 } 6038 break; 6039 case float_round_down: 6040 if (extractFloat128Sign(z)) { 6041 z.high |= (a.low != 0); 6042 z.high += roundBitsMask; 6043 } 6044 break; 6045 default: 6046 abort(); 6047 } 6048 z.high &= ~ roundBitsMask; 6049 } 6050 if ( ( z.low != a.low ) || ( z.high != a.high ) ) { 6051 STATUS(float_exception_flags) |= float_flag_inexact; 6052 } 6053 return z; 6054 6055 } 6056 6057 /*---------------------------------------------------------------------------- 6058 | Returns the result of adding the absolute values of the quadruple-precision 6059 | floating-point values `a' and `b'. If `zSign' is 1, the sum is negated 6060 | before being returned. `zSign' is ignored if the result is a NaN. 6061 | The addition is performed according to the IEC/IEEE Standard for Binary 6062 | Floating-Point Arithmetic. 6063 *----------------------------------------------------------------------------*/ 6064 6065 static float128 addFloat128Sigs( float128 a, float128 b, flag zSign STATUS_PARAM) 6066 { 6067 int32 aExp, bExp, zExp; 6068 uint64_t aSig0, aSig1, bSig0, bSig1, zSig0, zSig1, zSig2; 6069 int32 expDiff; 6070 6071 aSig1 = extractFloat128Frac1( a ); 6072 aSig0 = extractFloat128Frac0( a ); 6073 aExp = extractFloat128Exp( a ); 6074 bSig1 = extractFloat128Frac1( b ); 6075 bSig0 = extractFloat128Frac0( b ); 6076 bExp = extractFloat128Exp( b ); 6077 expDiff = aExp - bExp; 6078 if ( 0 < expDiff ) { 6079 if ( aExp == 0x7FFF ) { 6080 if ( aSig0 | aSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6081 return a; 6082 } 6083 if ( bExp == 0 ) { 6084 --expDiff; 6085 } 6086 else { 6087 bSig0 |= LIT64( 0x0001000000000000 ); 6088 } 6089 shift128ExtraRightJamming( 6090 bSig0, bSig1, 0, expDiff, &bSig0, &bSig1, &zSig2 ); 6091 zExp = aExp; 6092 } 6093 else if ( expDiff < 0 ) { 6094 if ( bExp == 0x7FFF ) { 6095 if ( bSig0 | bSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6096 return packFloat128( zSign, 0x7FFF, 0, 0 ); 6097 } 6098 if ( aExp == 0 ) { 6099 ++expDiff; 6100 } 6101 else { 6102 aSig0 |= LIT64( 0x0001000000000000 ); 6103 } 6104 shift128ExtraRightJamming( 6105 aSig0, aSig1, 0, - expDiff, &aSig0, &aSig1, &zSig2 ); 6106 zExp = bExp; 6107 } 6108 else { 6109 if ( aExp == 0x7FFF ) { 6110 if ( aSig0 | aSig1 | bSig0 | bSig1 ) { 6111 return propagateFloat128NaN( a, b STATUS_VAR ); 6112 } 6113 return a; 6114 } 6115 add128( aSig0, aSig1, bSig0, bSig1, &zSig0, &zSig1 ); 6116 if ( aExp == 0 ) { 6117 if (STATUS(flush_to_zero)) { 6118 if (zSig0 | zSig1) { 6119 float_raise(float_flag_output_denormal STATUS_VAR); 6120 } 6121 return packFloat128(zSign, 0, 0, 0); 6122 } 6123 return packFloat128( zSign, 0, zSig0, zSig1 ); 6124 } 6125 zSig2 = 0; 6126 zSig0 |= LIT64( 0x0002000000000000 ); 6127 zExp = aExp; 6128 goto shiftRight1; 6129 } 6130 aSig0 |= LIT64( 0x0001000000000000 ); 6131 add128( aSig0, aSig1, bSig0, bSig1, &zSig0, &zSig1 ); 6132 --zExp; 6133 if ( zSig0 < LIT64( 0x0002000000000000 ) ) goto roundAndPack; 6134 ++zExp; 6135 shiftRight1: 6136 shift128ExtraRightJamming( 6137 zSig0, zSig1, zSig2, 1, &zSig0, &zSig1, &zSig2 ); 6138 roundAndPack: 6139 return roundAndPackFloat128( zSign, zExp, zSig0, zSig1, zSig2 STATUS_VAR ); 6140 6141 } 6142 6143 /*---------------------------------------------------------------------------- 6144 | Returns the result of subtracting the absolute values of the quadruple- 6145 | precision floating-point values `a' and `b'. If `zSign' is 1, the 6146 | difference is negated before being returned. `zSign' is ignored if the 6147 | result is a NaN. The subtraction is performed according to the IEC/IEEE 6148 | Standard for Binary Floating-Point Arithmetic. 6149 *----------------------------------------------------------------------------*/ 6150 6151 static float128 subFloat128Sigs( float128 a, float128 b, flag zSign STATUS_PARAM) 6152 { 6153 int32 aExp, bExp, zExp; 6154 uint64_t aSig0, aSig1, bSig0, bSig1, zSig0, zSig1; 6155 int32 expDiff; 6156 float128 z; 6157 6158 aSig1 = extractFloat128Frac1( a ); 6159 aSig0 = extractFloat128Frac0( a ); 6160 aExp = extractFloat128Exp( a ); 6161 bSig1 = extractFloat128Frac1( b ); 6162 bSig0 = extractFloat128Frac0( b ); 6163 bExp = extractFloat128Exp( b ); 6164 expDiff = aExp - bExp; 6165 shortShift128Left( aSig0, aSig1, 14, &aSig0, &aSig1 ); 6166 shortShift128Left( bSig0, bSig1, 14, &bSig0, &bSig1 ); 6167 if ( 0 < expDiff ) goto aExpBigger; 6168 if ( expDiff < 0 ) goto bExpBigger; 6169 if ( aExp == 0x7FFF ) { 6170 if ( aSig0 | aSig1 | bSig0 | bSig1 ) { 6171 return propagateFloat128NaN( a, b STATUS_VAR ); 6172 } 6173 float_raise( float_flag_invalid STATUS_VAR); 6174 z.low = float128_default_nan_low; 6175 z.high = float128_default_nan_high; 6176 return z; 6177 } 6178 if ( aExp == 0 ) { 6179 aExp = 1; 6180 bExp = 1; 6181 } 6182 if ( bSig0 < aSig0 ) goto aBigger; 6183 if ( aSig0 < bSig0 ) goto bBigger; 6184 if ( bSig1 < aSig1 ) goto aBigger; 6185 if ( aSig1 < bSig1 ) goto bBigger; 6186 return packFloat128( STATUS(float_rounding_mode) == float_round_down, 0, 0, 0 ); 6187 bExpBigger: 6188 if ( bExp == 0x7FFF ) { 6189 if ( bSig0 | bSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6190 return packFloat128( zSign ^ 1, 0x7FFF, 0, 0 ); 6191 } 6192 if ( aExp == 0 ) { 6193 ++expDiff; 6194 } 6195 else { 6196 aSig0 |= LIT64( 0x4000000000000000 ); 6197 } 6198 shift128RightJamming( aSig0, aSig1, - expDiff, &aSig0, &aSig1 ); 6199 bSig0 |= LIT64( 0x4000000000000000 ); 6200 bBigger: 6201 sub128( bSig0, bSig1, aSig0, aSig1, &zSig0, &zSig1 ); 6202 zExp = bExp; 6203 zSign ^= 1; 6204 goto normalizeRoundAndPack; 6205 aExpBigger: 6206 if ( aExp == 0x7FFF ) { 6207 if ( aSig0 | aSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6208 return a; 6209 } 6210 if ( bExp == 0 ) { 6211 --expDiff; 6212 } 6213 else { 6214 bSig0 |= LIT64( 0x4000000000000000 ); 6215 } 6216 shift128RightJamming( bSig0, bSig1, expDiff, &bSig0, &bSig1 ); 6217 aSig0 |= LIT64( 0x4000000000000000 ); 6218 aBigger: 6219 sub128( aSig0, aSig1, bSig0, bSig1, &zSig0, &zSig1 ); 6220 zExp = aExp; 6221 normalizeRoundAndPack: 6222 --zExp; 6223 return normalizeRoundAndPackFloat128( zSign, zExp - 14, zSig0, zSig1 STATUS_VAR ); 6224 6225 } 6226 6227 /*---------------------------------------------------------------------------- 6228 | Returns the result of adding the quadruple-precision floating-point values 6229 | `a' and `b'. The operation is performed according to the IEC/IEEE Standard 6230 | for Binary Floating-Point Arithmetic. 6231 *----------------------------------------------------------------------------*/ 6232 6233 float128 float128_add( float128 a, float128 b STATUS_PARAM ) 6234 { 6235 flag aSign, bSign; 6236 6237 aSign = extractFloat128Sign( a ); 6238 bSign = extractFloat128Sign( b ); 6239 if ( aSign == bSign ) { 6240 return addFloat128Sigs( a, b, aSign STATUS_VAR ); 6241 } 6242 else { 6243 return subFloat128Sigs( a, b, aSign STATUS_VAR ); 6244 } 6245 6246 } 6247 6248 /*---------------------------------------------------------------------------- 6249 | Returns the result of subtracting the quadruple-precision floating-point 6250 | values `a' and `b'. The operation is performed according to the IEC/IEEE 6251 | Standard for Binary Floating-Point Arithmetic. 6252 *----------------------------------------------------------------------------*/ 6253 6254 float128 float128_sub( float128 a, float128 b STATUS_PARAM ) 6255 { 6256 flag aSign, bSign; 6257 6258 aSign = extractFloat128Sign( a ); 6259 bSign = extractFloat128Sign( b ); 6260 if ( aSign == bSign ) { 6261 return subFloat128Sigs( a, b, aSign STATUS_VAR ); 6262 } 6263 else { 6264 return addFloat128Sigs( a, b, aSign STATUS_VAR ); 6265 } 6266 6267 } 6268 6269 /*---------------------------------------------------------------------------- 6270 | Returns the result of multiplying the quadruple-precision floating-point 6271 | values `a' and `b'. The operation is performed according to the IEC/IEEE 6272 | Standard for Binary Floating-Point Arithmetic. 6273 *----------------------------------------------------------------------------*/ 6274 6275 float128 float128_mul( float128 a, float128 b STATUS_PARAM ) 6276 { 6277 flag aSign, bSign, zSign; 6278 int32 aExp, bExp, zExp; 6279 uint64_t aSig0, aSig1, bSig0, bSig1, zSig0, zSig1, zSig2, zSig3; 6280 float128 z; 6281 6282 aSig1 = extractFloat128Frac1( a ); 6283 aSig0 = extractFloat128Frac0( a ); 6284 aExp = extractFloat128Exp( a ); 6285 aSign = extractFloat128Sign( a ); 6286 bSig1 = extractFloat128Frac1( b ); 6287 bSig0 = extractFloat128Frac0( b ); 6288 bExp = extractFloat128Exp( b ); 6289 bSign = extractFloat128Sign( b ); 6290 zSign = aSign ^ bSign; 6291 if ( aExp == 0x7FFF ) { 6292 if ( ( aSig0 | aSig1 ) 6293 || ( ( bExp == 0x7FFF ) && ( bSig0 | bSig1 ) ) ) { 6294 return propagateFloat128NaN( a, b STATUS_VAR ); 6295 } 6296 if ( ( bExp | bSig0 | bSig1 ) == 0 ) goto invalid; 6297 return packFloat128( zSign, 0x7FFF, 0, 0 ); 6298 } 6299 if ( bExp == 0x7FFF ) { 6300 if ( bSig0 | bSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6301 if ( ( aExp | aSig0 | aSig1 ) == 0 ) { 6302 invalid: 6303 float_raise( float_flag_invalid STATUS_VAR); 6304 z.low = float128_default_nan_low; 6305 z.high = float128_default_nan_high; 6306 return z; 6307 } 6308 return packFloat128( zSign, 0x7FFF, 0, 0 ); 6309 } 6310 if ( aExp == 0 ) { 6311 if ( ( aSig0 | aSig1 ) == 0 ) return packFloat128( zSign, 0, 0, 0 ); 6312 normalizeFloat128Subnormal( aSig0, aSig1, &aExp, &aSig0, &aSig1 ); 6313 } 6314 if ( bExp == 0 ) { 6315 if ( ( bSig0 | bSig1 ) == 0 ) return packFloat128( zSign, 0, 0, 0 ); 6316 normalizeFloat128Subnormal( bSig0, bSig1, &bExp, &bSig0, &bSig1 ); 6317 } 6318 zExp = aExp + bExp - 0x4000; 6319 aSig0 |= LIT64( 0x0001000000000000 ); 6320 shortShift128Left( bSig0, bSig1, 16, &bSig0, &bSig1 ); 6321 mul128To256( aSig0, aSig1, bSig0, bSig1, &zSig0, &zSig1, &zSig2, &zSig3 ); 6322 add128( zSig0, zSig1, aSig0, aSig1, &zSig0, &zSig1 ); 6323 zSig2 |= ( zSig3 != 0 ); 6324 if ( LIT64( 0x0002000000000000 ) <= zSig0 ) { 6325 shift128ExtraRightJamming( 6326 zSig0, zSig1, zSig2, 1, &zSig0, &zSig1, &zSig2 ); 6327 ++zExp; 6328 } 6329 return roundAndPackFloat128( zSign, zExp, zSig0, zSig1, zSig2 STATUS_VAR ); 6330 6331 } 6332 6333 /*---------------------------------------------------------------------------- 6334 | Returns the result of dividing the quadruple-precision floating-point value 6335 | `a' by the corresponding value `b'. The operation is performed according to 6336 | the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 6337 *----------------------------------------------------------------------------*/ 6338 6339 float128 float128_div( float128 a, float128 b STATUS_PARAM ) 6340 { 6341 flag aSign, bSign, zSign; 6342 int32 aExp, bExp, zExp; 6343 uint64_t aSig0, aSig1, bSig0, bSig1, zSig0, zSig1, zSig2; 6344 uint64_t rem0, rem1, rem2, rem3, term0, term1, term2, term3; 6345 float128 z; 6346 6347 aSig1 = extractFloat128Frac1( a ); 6348 aSig0 = extractFloat128Frac0( a ); 6349 aExp = extractFloat128Exp( a ); 6350 aSign = extractFloat128Sign( a ); 6351 bSig1 = extractFloat128Frac1( b ); 6352 bSig0 = extractFloat128Frac0( b ); 6353 bExp = extractFloat128Exp( b ); 6354 bSign = extractFloat128Sign( b ); 6355 zSign = aSign ^ bSign; 6356 if ( aExp == 0x7FFF ) { 6357 if ( aSig0 | aSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6358 if ( bExp == 0x7FFF ) { 6359 if ( bSig0 | bSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6360 goto invalid; 6361 } 6362 return packFloat128( zSign, 0x7FFF, 0, 0 ); 6363 } 6364 if ( bExp == 0x7FFF ) { 6365 if ( bSig0 | bSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6366 return packFloat128( zSign, 0, 0, 0 ); 6367 } 6368 if ( bExp == 0 ) { 6369 if ( ( bSig0 | bSig1 ) == 0 ) { 6370 if ( ( aExp | aSig0 | aSig1 ) == 0 ) { 6371 invalid: 6372 float_raise( float_flag_invalid STATUS_VAR); 6373 z.low = float128_default_nan_low; 6374 z.high = float128_default_nan_high; 6375 return z; 6376 } 6377 float_raise( float_flag_divbyzero STATUS_VAR); 6378 return packFloat128( zSign, 0x7FFF, 0, 0 ); 6379 } 6380 normalizeFloat128Subnormal( bSig0, bSig1, &bExp, &bSig0, &bSig1 ); 6381 } 6382 if ( aExp == 0 ) { 6383 if ( ( aSig0 | aSig1 ) == 0 ) return packFloat128( zSign, 0, 0, 0 ); 6384 normalizeFloat128Subnormal( aSig0, aSig1, &aExp, &aSig0, &aSig1 ); 6385 } 6386 zExp = aExp - bExp + 0x3FFD; 6387 shortShift128Left( 6388 aSig0 | LIT64( 0x0001000000000000 ), aSig1, 15, &aSig0, &aSig1 ); 6389 shortShift128Left( 6390 bSig0 | LIT64( 0x0001000000000000 ), bSig1, 15, &bSig0, &bSig1 ); 6391 if ( le128( bSig0, bSig1, aSig0, aSig1 ) ) { 6392 shift128Right( aSig0, aSig1, 1, &aSig0, &aSig1 ); 6393 ++zExp; 6394 } 6395 zSig0 = estimateDiv128To64( aSig0, aSig1, bSig0 ); 6396 mul128By64To192( bSig0, bSig1, zSig0, &term0, &term1, &term2 ); 6397 sub192( aSig0, aSig1, 0, term0, term1, term2, &rem0, &rem1, &rem2 ); 6398 while ( (int64_t) rem0 < 0 ) { 6399 --zSig0; 6400 add192( rem0, rem1, rem2, 0, bSig0, bSig1, &rem0, &rem1, &rem2 ); 6401 } 6402 zSig1 = estimateDiv128To64( rem1, rem2, bSig0 ); 6403 if ( ( zSig1 & 0x3FFF ) <= 4 ) { 6404 mul128By64To192( bSig0, bSig1, zSig1, &term1, &term2, &term3 ); 6405 sub192( rem1, rem2, 0, term1, term2, term3, &rem1, &rem2, &rem3 ); 6406 while ( (int64_t) rem1 < 0 ) { 6407 --zSig1; 6408 add192( rem1, rem2, rem3, 0, bSig0, bSig1, &rem1, &rem2, &rem3 ); 6409 } 6410 zSig1 |= ( ( rem1 | rem2 | rem3 ) != 0 ); 6411 } 6412 shift128ExtraRightJamming( zSig0, zSig1, 0, 15, &zSig0, &zSig1, &zSig2 ); 6413 return roundAndPackFloat128( zSign, zExp, zSig0, zSig1, zSig2 STATUS_VAR ); 6414 6415 } 6416 6417 /*---------------------------------------------------------------------------- 6418 | Returns the remainder of the quadruple-precision floating-point value `a' 6419 | with respect to the corresponding value `b'. The operation is performed 6420 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 6421 *----------------------------------------------------------------------------*/ 6422 6423 float128 float128_rem( float128 a, float128 b STATUS_PARAM ) 6424 { 6425 flag aSign, zSign; 6426 int32 aExp, bExp, expDiff; 6427 uint64_t aSig0, aSig1, bSig0, bSig1, q, term0, term1, term2; 6428 uint64_t allZero, alternateASig0, alternateASig1, sigMean1; 6429 int64_t sigMean0; 6430 float128 z; 6431 6432 aSig1 = extractFloat128Frac1( a ); 6433 aSig0 = extractFloat128Frac0( a ); 6434 aExp = extractFloat128Exp( a ); 6435 aSign = extractFloat128Sign( a ); 6436 bSig1 = extractFloat128Frac1( b ); 6437 bSig0 = extractFloat128Frac0( b ); 6438 bExp = extractFloat128Exp( b ); 6439 if ( aExp == 0x7FFF ) { 6440 if ( ( aSig0 | aSig1 ) 6441 || ( ( bExp == 0x7FFF ) && ( bSig0 | bSig1 ) ) ) { 6442 return propagateFloat128NaN( a, b STATUS_VAR ); 6443 } 6444 goto invalid; 6445 } 6446 if ( bExp == 0x7FFF ) { 6447 if ( bSig0 | bSig1 ) return propagateFloat128NaN( a, b STATUS_VAR ); 6448 return a; 6449 } 6450 if ( bExp == 0 ) { 6451 if ( ( bSig0 | bSig1 ) == 0 ) { 6452 invalid: 6453 float_raise( float_flag_invalid STATUS_VAR); 6454 z.low = float128_default_nan_low; 6455 z.high = float128_default_nan_high; 6456 return z; 6457 } 6458 normalizeFloat128Subnormal( bSig0, bSig1, &bExp, &bSig0, &bSig1 ); 6459 } 6460 if ( aExp == 0 ) { 6461 if ( ( aSig0 | aSig1 ) == 0 ) return a; 6462 normalizeFloat128Subnormal( aSig0, aSig1, &aExp, &aSig0, &aSig1 ); 6463 } 6464 expDiff = aExp - bExp; 6465 if ( expDiff < -1 ) return a; 6466 shortShift128Left( 6467 aSig0 | LIT64( 0x0001000000000000 ), 6468 aSig1, 6469 15 - ( expDiff < 0 ), 6470 &aSig0, 6471 &aSig1 6472 ); 6473 shortShift128Left( 6474 bSig0 | LIT64( 0x0001000000000000 ), bSig1, 15, &bSig0, &bSig1 ); 6475 q = le128( bSig0, bSig1, aSig0, aSig1 ); 6476 if ( q ) sub128( aSig0, aSig1, bSig0, bSig1, &aSig0, &aSig1 ); 6477 expDiff -= 64; 6478 while ( 0 < expDiff ) { 6479 q = estimateDiv128To64( aSig0, aSig1, bSig0 ); 6480 q = ( 4 < q ) ? q - 4 : 0; 6481 mul128By64To192( bSig0, bSig1, q, &term0, &term1, &term2 ); 6482 shortShift192Left( term0, term1, term2, 61, &term1, &term2, &allZero ); 6483 shortShift128Left( aSig0, aSig1, 61, &aSig0, &allZero ); 6484 sub128( aSig0, 0, term1, term2, &aSig0, &aSig1 ); 6485 expDiff -= 61; 6486 } 6487 if ( -64 < expDiff ) { 6488 q = estimateDiv128To64( aSig0, aSig1, bSig0 ); 6489 q = ( 4 < q ) ? q - 4 : 0; 6490 q >>= - expDiff; 6491 shift128Right( bSig0, bSig1, 12, &bSig0, &bSig1 ); 6492 expDiff += 52; 6493 if ( expDiff < 0 ) { 6494 shift128Right( aSig0, aSig1, - expDiff, &aSig0, &aSig1 ); 6495 } 6496 else { 6497 shortShift128Left( aSig0, aSig1, expDiff, &aSig0, &aSig1 ); 6498 } 6499 mul128By64To192( bSig0, bSig1, q, &term0, &term1, &term2 ); 6500 sub128( aSig0, aSig1, term1, term2, &aSig0, &aSig1 ); 6501 } 6502 else { 6503 shift128Right( aSig0, aSig1, 12, &aSig0, &aSig1 ); 6504 shift128Right( bSig0, bSig1, 12, &bSig0, &bSig1 ); 6505 } 6506 do { 6507 alternateASig0 = aSig0; 6508 alternateASig1 = aSig1; 6509 ++q; 6510 sub128( aSig0, aSig1, bSig0, bSig1, &aSig0, &aSig1 ); 6511 } while ( 0 <= (int64_t) aSig0 ); 6512 add128( 6513 aSig0, aSig1, alternateASig0, alternateASig1, (uint64_t *)&sigMean0, &sigMean1 ); 6514 if ( ( sigMean0 < 0 ) 6515 || ( ( ( sigMean0 | sigMean1 ) == 0 ) && ( q & 1 ) ) ) { 6516 aSig0 = alternateASig0; 6517 aSig1 = alternateASig1; 6518 } 6519 zSign = ( (int64_t) aSig0 < 0 ); 6520 if ( zSign ) sub128( 0, 0, aSig0, aSig1, &aSig0, &aSig1 ); 6521 return 6522 normalizeRoundAndPackFloat128( aSign ^ zSign, bExp - 4, aSig0, aSig1 STATUS_VAR ); 6523 6524 } 6525 6526 /*---------------------------------------------------------------------------- 6527 | Returns the square root of the quadruple-precision floating-point value `a'. 6528 | The operation is performed according to the IEC/IEEE Standard for Binary 6529 | Floating-Point Arithmetic. 6530 *----------------------------------------------------------------------------*/ 6531 6532 float128 float128_sqrt( float128 a STATUS_PARAM ) 6533 { 6534 flag aSign; 6535 int32 aExp, zExp; 6536 uint64_t aSig0, aSig1, zSig0, zSig1, zSig2, doubleZSig0; 6537 uint64_t rem0, rem1, rem2, rem3, term0, term1, term2, term3; 6538 float128 z; 6539 6540 aSig1 = extractFloat128Frac1( a ); 6541 aSig0 = extractFloat128Frac0( a ); 6542 aExp = extractFloat128Exp( a ); 6543 aSign = extractFloat128Sign( a ); 6544 if ( aExp == 0x7FFF ) { 6545 if ( aSig0 | aSig1 ) return propagateFloat128NaN( a, a STATUS_VAR ); 6546 if ( ! aSign ) return a; 6547 goto invalid; 6548 } 6549 if ( aSign ) { 6550 if ( ( aExp | aSig0 | aSig1 ) == 0 ) return a; 6551 invalid: 6552 float_raise( float_flag_invalid STATUS_VAR); 6553 z.low = float128_default_nan_low; 6554 z.high = float128_default_nan_high; 6555 return z; 6556 } 6557 if ( aExp == 0 ) { 6558 if ( ( aSig0 | aSig1 ) == 0 ) return packFloat128( 0, 0, 0, 0 ); 6559 normalizeFloat128Subnormal( aSig0, aSig1, &aExp, &aSig0, &aSig1 ); 6560 } 6561 zExp = ( ( aExp - 0x3FFF )>>1 ) + 0x3FFE; 6562 aSig0 |= LIT64( 0x0001000000000000 ); 6563 zSig0 = estimateSqrt32( aExp, aSig0>>17 ); 6564 shortShift128Left( aSig0, aSig1, 13 - ( aExp & 1 ), &aSig0, &aSig1 ); 6565 zSig0 = estimateDiv128To64( aSig0, aSig1, zSig0<<32 ) + ( zSig0<<30 ); 6566 doubleZSig0 = zSig0<<1; 6567 mul64To128( zSig0, zSig0, &term0, &term1 ); 6568 sub128( aSig0, aSig1, term0, term1, &rem0, &rem1 ); 6569 while ( (int64_t) rem0 < 0 ) { 6570 --zSig0; 6571 doubleZSig0 -= 2; 6572 add128( rem0, rem1, zSig0>>63, doubleZSig0 | 1, &rem0, &rem1 ); 6573 } 6574 zSig1 = estimateDiv128To64( rem1, 0, doubleZSig0 ); 6575 if ( ( zSig1 & 0x1FFF ) <= 5 ) { 6576 if ( zSig1 == 0 ) zSig1 = 1; 6577 mul64To128( doubleZSig0, zSig1, &term1, &term2 ); 6578 sub128( rem1, 0, term1, term2, &rem1, &rem2 ); 6579 mul64To128( zSig1, zSig1, &term2, &term3 ); 6580 sub192( rem1, rem2, 0, 0, term2, term3, &rem1, &rem2, &rem3 ); 6581 while ( (int64_t) rem1 < 0 ) { 6582 --zSig1; 6583 shortShift128Left( 0, zSig1, 1, &term2, &term3 ); 6584 term3 |= 1; 6585 term2 |= doubleZSig0; 6586 add192( rem1, rem2, rem3, 0, term2, term3, &rem1, &rem2, &rem3 ); 6587 } 6588 zSig1 |= ( ( rem1 | rem2 | rem3 ) != 0 ); 6589 } 6590 shift128ExtraRightJamming( zSig0, zSig1, 0, 14, &zSig0, &zSig1, &zSig2 ); 6591 return roundAndPackFloat128( 0, zExp, zSig0, zSig1, zSig2 STATUS_VAR ); 6592 6593 } 6594 6595 /*---------------------------------------------------------------------------- 6596 | Returns 1 if the quadruple-precision floating-point value `a' is equal to 6597 | the corresponding value `b', and 0 otherwise. The invalid exception is 6598 | raised if either operand is a NaN. Otherwise, the comparison is performed 6599 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 6600 *----------------------------------------------------------------------------*/ 6601 6602 int float128_eq( float128 a, float128 b STATUS_PARAM ) 6603 { 6604 6605 if ( ( ( extractFloat128Exp( a ) == 0x7FFF ) 6606 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) 6607 || ( ( extractFloat128Exp( b ) == 0x7FFF ) 6608 && ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) ) 6609 ) { 6610 float_raise( float_flag_invalid STATUS_VAR); 6611 return 0; 6612 } 6613 return 6614 ( a.low == b.low ) 6615 && ( ( a.high == b.high ) 6616 || ( ( a.low == 0 ) 6617 && ( (uint64_t) ( ( a.high | b.high )<<1 ) == 0 ) ) 6618 ); 6619 6620 } 6621 6622 /*---------------------------------------------------------------------------- 6623 | Returns 1 if the quadruple-precision floating-point value `a' is less than 6624 | or equal to the corresponding value `b', and 0 otherwise. The invalid 6625 | exception is raised if either operand is a NaN. The comparison is performed 6626 | according to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 6627 *----------------------------------------------------------------------------*/ 6628 6629 int float128_le( float128 a, float128 b STATUS_PARAM ) 6630 { 6631 flag aSign, bSign; 6632 6633 if ( ( ( extractFloat128Exp( a ) == 0x7FFF ) 6634 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) 6635 || ( ( extractFloat128Exp( b ) == 0x7FFF ) 6636 && ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) ) 6637 ) { 6638 float_raise( float_flag_invalid STATUS_VAR); 6639 return 0; 6640 } 6641 aSign = extractFloat128Sign( a ); 6642 bSign = extractFloat128Sign( b ); 6643 if ( aSign != bSign ) { 6644 return 6645 aSign 6646 || ( ( ( (uint64_t) ( ( a.high | b.high )<<1 ) ) | a.low | b.low ) 6647 == 0 ); 6648 } 6649 return 6650 aSign ? le128( b.high, b.low, a.high, a.low ) 6651 : le128( a.high, a.low, b.high, b.low ); 6652 6653 } 6654 6655 /*---------------------------------------------------------------------------- 6656 | Returns 1 if the quadruple-precision floating-point value `a' is less than 6657 | the corresponding value `b', and 0 otherwise. The invalid exception is 6658 | raised if either operand is a NaN. The comparison is performed according 6659 | to the IEC/IEEE Standard for Binary Floating-Point Arithmetic. 6660 *----------------------------------------------------------------------------*/ 6661 6662 int float128_lt( float128 a, float128 b STATUS_PARAM ) 6663 { 6664 flag aSign, bSign; 6665 6666 if ( ( ( extractFloat128Exp( a ) == 0x7FFF ) 6667 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) 6668 || ( ( extractFloat128Exp( b ) == 0x7FFF ) 6669 && ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) ) 6670 ) { 6671 float_raise( float_flag_invalid STATUS_VAR); 6672 return 0; 6673 } 6674 aSign = extractFloat128Sign( a ); 6675 bSign = extractFloat128Sign( b ); 6676 if ( aSign != bSign ) { 6677 return 6678 aSign 6679 && ( ( ( (uint64_t) ( ( a.high | b.high )<<1 ) ) | a.low | b.low ) 6680 != 0 ); 6681 } 6682 return 6683 aSign ? lt128( b.high, b.low, a.high, a.low ) 6684 : lt128( a.high, a.low, b.high, b.low ); 6685 6686 } 6687 6688 /*---------------------------------------------------------------------------- 6689 | Returns 1 if the quadruple-precision floating-point values `a' and `b' cannot 6690 | be compared, and 0 otherwise. The invalid exception is raised if either 6691 | operand is a NaN. The comparison is performed according to the IEC/IEEE 6692 | Standard for Binary Floating-Point Arithmetic. 6693 *----------------------------------------------------------------------------*/ 6694 6695 int float128_unordered( float128 a, float128 b STATUS_PARAM ) 6696 { 6697 if ( ( ( extractFloat128Exp( a ) == 0x7FFF ) 6698 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) 6699 || ( ( extractFloat128Exp( b ) == 0x7FFF ) 6700 && ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) ) 6701 ) { 6702 float_raise( float_flag_invalid STATUS_VAR); 6703 return 1; 6704 } 6705 return 0; 6706 } 6707 6708 /*---------------------------------------------------------------------------- 6709 | Returns 1 if the quadruple-precision floating-point value `a' is equal to 6710 | the corresponding value `b', and 0 otherwise. Quiet NaNs do not cause an 6711 | exception. The comparison is performed according to the IEC/IEEE Standard 6712 | for Binary Floating-Point Arithmetic. 6713 *----------------------------------------------------------------------------*/ 6714 6715 int float128_eq_quiet( float128 a, float128 b STATUS_PARAM ) 6716 { 6717 6718 if ( ( ( extractFloat128Exp( a ) == 0x7FFF ) 6719 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) 6720 || ( ( extractFloat128Exp( b ) == 0x7FFF ) 6721 && ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) ) 6722 ) { 6723 if ( float128_is_signaling_nan( a ) 6724 || float128_is_signaling_nan( b ) ) { 6725 float_raise( float_flag_invalid STATUS_VAR); 6726 } 6727 return 0; 6728 } 6729 return 6730 ( a.low == b.low ) 6731 && ( ( a.high == b.high ) 6732 || ( ( a.low == 0 ) 6733 && ( (uint64_t) ( ( a.high | b.high )<<1 ) == 0 ) ) 6734 ); 6735 6736 } 6737 6738 /*---------------------------------------------------------------------------- 6739 | Returns 1 if the quadruple-precision floating-point value `a' is less than 6740 | or equal to the corresponding value `b', and 0 otherwise. Quiet NaNs do not 6741 | cause an exception. Otherwise, the comparison is performed according to the 6742 | IEC/IEEE Standard for Binary Floating-Point Arithmetic. 6743 *----------------------------------------------------------------------------*/ 6744 6745 int float128_le_quiet( float128 a, float128 b STATUS_PARAM ) 6746 { 6747 flag aSign, bSign; 6748 6749 if ( ( ( extractFloat128Exp( a ) == 0x7FFF ) 6750 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) 6751 || ( ( extractFloat128Exp( b ) == 0x7FFF ) 6752 && ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) ) 6753 ) { 6754 if ( float128_is_signaling_nan( a ) 6755 || float128_is_signaling_nan( b ) ) { 6756 float_raise( float_flag_invalid STATUS_VAR); 6757 } 6758 return 0; 6759 } 6760 aSign = extractFloat128Sign( a ); 6761 bSign = extractFloat128Sign( b ); 6762 if ( aSign != bSign ) { 6763 return 6764 aSign 6765 || ( ( ( (uint64_t) ( ( a.high | b.high )<<1 ) ) | a.low | b.low ) 6766 == 0 ); 6767 } 6768 return 6769 aSign ? le128( b.high, b.low, a.high, a.low ) 6770 : le128( a.high, a.low, b.high, b.low ); 6771 6772 } 6773 6774 /*---------------------------------------------------------------------------- 6775 | Returns 1 if the quadruple-precision floating-point value `a' is less than 6776 | the corresponding value `b', and 0 otherwise. Quiet NaNs do not cause an 6777 | exception. Otherwise, the comparison is performed according to the IEC/IEEE 6778 | Standard for Binary Floating-Point Arithmetic. 6779 *----------------------------------------------------------------------------*/ 6780 6781 int float128_lt_quiet( float128 a, float128 b STATUS_PARAM ) 6782 { 6783 flag aSign, bSign; 6784 6785 if ( ( ( extractFloat128Exp( a ) == 0x7FFF ) 6786 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) 6787 || ( ( extractFloat128Exp( b ) == 0x7FFF ) 6788 && ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) ) 6789 ) { 6790 if ( float128_is_signaling_nan( a ) 6791 || float128_is_signaling_nan( b ) ) { 6792 float_raise( float_flag_invalid STATUS_VAR); 6793 } 6794 return 0; 6795 } 6796 aSign = extractFloat128Sign( a ); 6797 bSign = extractFloat128Sign( b ); 6798 if ( aSign != bSign ) { 6799 return 6800 aSign 6801 && ( ( ( (uint64_t) ( ( a.high | b.high )<<1 ) ) | a.low | b.low ) 6802 != 0 ); 6803 } 6804 return 6805 aSign ? lt128( b.high, b.low, a.high, a.low ) 6806 : lt128( a.high, a.low, b.high, b.low ); 6807 6808 } 6809 6810 /*---------------------------------------------------------------------------- 6811 | Returns 1 if the quadruple-precision floating-point values `a' and `b' cannot 6812 | be compared, and 0 otherwise. Quiet NaNs do not cause an exception. The 6813 | comparison is performed according to the IEC/IEEE Standard for Binary 6814 | Floating-Point Arithmetic. 6815 *----------------------------------------------------------------------------*/ 6816 6817 int float128_unordered_quiet( float128 a, float128 b STATUS_PARAM ) 6818 { 6819 if ( ( ( extractFloat128Exp( a ) == 0x7FFF ) 6820 && ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) 6821 || ( ( extractFloat128Exp( b ) == 0x7FFF ) 6822 && ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) ) 6823 ) { 6824 if ( float128_is_signaling_nan( a ) 6825 || float128_is_signaling_nan( b ) ) { 6826 float_raise( float_flag_invalid STATUS_VAR); 6827 } 6828 return 1; 6829 } 6830 return 0; 6831 } 6832 6833 /* misc functions */ 6834 float32 uint32_to_float32(uint32_t a STATUS_PARAM) 6835 { 6836 return int64_to_float32(a STATUS_VAR); 6837 } 6838 6839 float64 uint32_to_float64(uint32_t a STATUS_PARAM) 6840 { 6841 return int64_to_float64(a STATUS_VAR); 6842 } 6843 6844 uint32 float32_to_uint32( float32 a STATUS_PARAM ) 6845 { 6846 int64_t v; 6847 uint32 res; 6848 int old_exc_flags = get_float_exception_flags(status); 6849 6850 v = float32_to_int64(a STATUS_VAR); 6851 if (v < 0) { 6852 res = 0; 6853 } else if (v > 0xffffffff) { 6854 res = 0xffffffff; 6855 } else { 6856 return v; 6857 } 6858 set_float_exception_flags(old_exc_flags, status); 6859 float_raise(float_flag_invalid STATUS_VAR); 6860 return res; 6861 } 6862 6863 uint32 float32_to_uint32_round_to_zero( float32 a STATUS_PARAM ) 6864 { 6865 int64_t v; 6866 uint32 res; 6867 int old_exc_flags = get_float_exception_flags(status); 6868 6869 v = float32_to_int64_round_to_zero(a STATUS_VAR); 6870 if (v < 0) { 6871 res = 0; 6872 } else if (v > 0xffffffff) { 6873 res = 0xffffffff; 6874 } else { 6875 return v; 6876 } 6877 set_float_exception_flags(old_exc_flags, status); 6878 float_raise(float_flag_invalid STATUS_VAR); 6879 return res; 6880 } 6881 6882 int_fast16_t float32_to_int16(float32 a STATUS_PARAM) 6883 { 6884 int32_t v; 6885 int_fast16_t res; 6886 int old_exc_flags = get_float_exception_flags(status); 6887 6888 v = float32_to_int32(a STATUS_VAR); 6889 if (v < -0x8000) { 6890 res = -0x8000; 6891 } else if (v > 0x7fff) { 6892 res = 0x7fff; 6893 } else { 6894 return v; 6895 } 6896 6897 set_float_exception_flags(old_exc_flags, status); 6898 float_raise(float_flag_invalid STATUS_VAR); 6899 return res; 6900 } 6901 6902 uint_fast16_t float32_to_uint16(float32 a STATUS_PARAM) 6903 { 6904 int32_t v; 6905 uint_fast16_t res; 6906 int old_exc_flags = get_float_exception_flags(status); 6907 6908 v = float32_to_int32(a STATUS_VAR); 6909 if (v < 0) { 6910 res = 0; 6911 } else if (v > 0xffff) { 6912 res = 0xffff; 6913 } else { 6914 return v; 6915 } 6916 6917 set_float_exception_flags(old_exc_flags, status); 6918 float_raise(float_flag_invalid STATUS_VAR); 6919 return res; 6920 } 6921 6922 uint_fast16_t float32_to_uint16_round_to_zero(float32 a STATUS_PARAM) 6923 { 6924 int64_t v; 6925 uint_fast16_t res; 6926 int old_exc_flags = get_float_exception_flags(status); 6927 6928 v = float32_to_int64_round_to_zero(a STATUS_VAR); 6929 if (v < 0) { 6930 res = 0; 6931 } else if (v > 0xffff) { 6932 res = 0xffff; 6933 } else { 6934 return v; 6935 } 6936 set_float_exception_flags(old_exc_flags, status); 6937 float_raise(float_flag_invalid STATUS_VAR); 6938 return res; 6939 } 6940 6941 uint32 float64_to_uint32( float64 a STATUS_PARAM ) 6942 { 6943 uint64_t v; 6944 uint32 res; 6945 int old_exc_flags = get_float_exception_flags(status); 6946 6947 v = float64_to_uint64(a STATUS_VAR); 6948 if (v > 0xffffffff) { 6949 res = 0xffffffff; 6950 } else { 6951 return v; 6952 } 6953 set_float_exception_flags(old_exc_flags, status); 6954 float_raise(float_flag_invalid STATUS_VAR); 6955 return res; 6956 } 6957 6958 uint32 float64_to_uint32_round_to_zero( float64 a STATUS_PARAM ) 6959 { 6960 uint64_t v; 6961 uint32 res; 6962 int old_exc_flags = get_float_exception_flags(status); 6963 6964 v = float64_to_uint64_round_to_zero(a STATUS_VAR); 6965 if (v > 0xffffffff) { 6966 res = 0xffffffff; 6967 } else { 6968 return v; 6969 } 6970 set_float_exception_flags(old_exc_flags, status); 6971 float_raise(float_flag_invalid STATUS_VAR); 6972 return res; 6973 } 6974 6975 int_fast16_t float64_to_int16(float64 a STATUS_PARAM) 6976 { 6977 int64_t v; 6978 int_fast16_t res; 6979 int old_exc_flags = get_float_exception_flags(status); 6980 6981 v = float64_to_int32(a STATUS_VAR); 6982 if (v < -0x8000) { 6983 res = -0x8000; 6984 } else if (v > 0x7fff) { 6985 res = 0x7fff; 6986 } else { 6987 return v; 6988 } 6989 6990 set_float_exception_flags(old_exc_flags, status); 6991 float_raise(float_flag_invalid STATUS_VAR); 6992 return res; 6993 } 6994 6995 uint_fast16_t float64_to_uint16(float64 a STATUS_PARAM) 6996 { 6997 int64_t v; 6998 uint_fast16_t res; 6999 int old_exc_flags = get_float_exception_flags(status); 7000 7001 v = float64_to_int32(a STATUS_VAR); 7002 if (v < 0) { 7003 res = 0; 7004 } else if (v > 0xffff) { 7005 res = 0xffff; 7006 } else { 7007 return v; 7008 } 7009 7010 set_float_exception_flags(old_exc_flags, status); 7011 float_raise(float_flag_invalid STATUS_VAR); 7012 return res; 7013 } 7014 7015 uint_fast16_t float64_to_uint16_round_to_zero(float64 a STATUS_PARAM) 7016 { 7017 int64_t v; 7018 uint_fast16_t res; 7019 int old_exc_flags = get_float_exception_flags(status); 7020 7021 v = float64_to_int64_round_to_zero(a STATUS_VAR); 7022 if (v < 0) { 7023 res = 0; 7024 } else if (v > 0xffff) { 7025 res = 0xffff; 7026 } else { 7027 return v; 7028 } 7029 set_float_exception_flags(old_exc_flags, status); 7030 float_raise(float_flag_invalid STATUS_VAR); 7031 return res; 7032 } 7033 7034 /*---------------------------------------------------------------------------- 7035 | Returns the result of converting the double-precision floating-point value 7036 | `a' to the 64-bit unsigned integer format. The conversion is 7037 | performed according to the IEC/IEEE Standard for Binary Floating-Point 7038 | Arithmetic---which means in particular that the conversion is rounded 7039 | according to the current rounding mode. If `a' is a NaN, the largest 7040 | positive integer is returned. If the conversion overflows, the 7041 | largest unsigned integer is returned. If 'a' is negative, the value is 7042 | rounded and zero is returned; negative values that do not round to zero 7043 | will raise the inexact exception. 7044 *----------------------------------------------------------------------------*/ 7045 7046 uint64_t float64_to_uint64(float64 a STATUS_PARAM) 7047 { 7048 flag aSign; 7049 int_fast16_t aExp, shiftCount; 7050 uint64_t aSig, aSigExtra; 7051 a = float64_squash_input_denormal(a STATUS_VAR); 7052 7053 aSig = extractFloat64Frac(a); 7054 aExp = extractFloat64Exp(a); 7055 aSign = extractFloat64Sign(a); 7056 if (aSign && (aExp > 1022)) { 7057 float_raise(float_flag_invalid STATUS_VAR); 7058 if (float64_is_any_nan(a)) { 7059 return LIT64(0xFFFFFFFFFFFFFFFF); 7060 } else { 7061 return 0; 7062 } 7063 } 7064 if (aExp) { 7065 aSig |= LIT64(0x0010000000000000); 7066 } 7067 shiftCount = 0x433 - aExp; 7068 if (shiftCount <= 0) { 7069 if (0x43E < aExp) { 7070 float_raise(float_flag_invalid STATUS_VAR); 7071 return LIT64(0xFFFFFFFFFFFFFFFF); 7072 } 7073 aSigExtra = 0; 7074 aSig <<= -shiftCount; 7075 } else { 7076 shift64ExtraRightJamming(aSig, 0, shiftCount, &aSig, &aSigExtra); 7077 } 7078 return roundAndPackUint64(aSign, aSig, aSigExtra STATUS_VAR); 7079 } 7080 7081 uint64_t float64_to_uint64_round_to_zero (float64 a STATUS_PARAM) 7082 { 7083 signed char current_rounding_mode = STATUS(float_rounding_mode); 7084 set_float_rounding_mode(float_round_to_zero STATUS_VAR); 7085 int64_t v = float64_to_uint64(a STATUS_VAR); 7086 set_float_rounding_mode(current_rounding_mode STATUS_VAR); 7087 return v; 7088 } 7089 7090 #define COMPARE(s, nan_exp) \ 7091 static inline int float ## s ## _compare_internal( float ## s a, float ## s b, \ 7092 int is_quiet STATUS_PARAM ) \ 7093 { \ 7094 flag aSign, bSign; \ 7095 uint ## s ## _t av, bv; \ 7096 a = float ## s ## _squash_input_denormal(a STATUS_VAR); \ 7097 b = float ## s ## _squash_input_denormal(b STATUS_VAR); \ 7098 \ 7099 if (( ( extractFloat ## s ## Exp( a ) == nan_exp ) && \ 7100 extractFloat ## s ## Frac( a ) ) || \ 7101 ( ( extractFloat ## s ## Exp( b ) == nan_exp ) && \ 7102 extractFloat ## s ## Frac( b ) )) { \ 7103 if (!is_quiet || \ 7104 float ## s ## _is_signaling_nan( a ) || \ 7105 float ## s ## _is_signaling_nan( b ) ) { \ 7106 float_raise( float_flag_invalid STATUS_VAR); \ 7107 } \ 7108 return float_relation_unordered; \ 7109 } \ 7110 aSign = extractFloat ## s ## Sign( a ); \ 7111 bSign = extractFloat ## s ## Sign( b ); \ 7112 av = float ## s ## _val(a); \ 7113 bv = float ## s ## _val(b); \ 7114 if ( aSign != bSign ) { \ 7115 if ( (uint ## s ## _t) ( ( av | bv )<<1 ) == 0 ) { \ 7116 /* zero case */ \ 7117 return float_relation_equal; \ 7118 } else { \ 7119 return 1 - (2 * aSign); \ 7120 } \ 7121 } else { \ 7122 if (av == bv) { \ 7123 return float_relation_equal; \ 7124 } else { \ 7125 return 1 - 2 * (aSign ^ ( av < bv )); \ 7126 } \ 7127 } \ 7128 } \ 7129 \ 7130 int float ## s ## _compare( float ## s a, float ## s b STATUS_PARAM ) \ 7131 { \ 7132 return float ## s ## _compare_internal(a, b, 0 STATUS_VAR); \ 7133 } \ 7134 \ 7135 int float ## s ## _compare_quiet( float ## s a, float ## s b STATUS_PARAM ) \ 7136 { \ 7137 return float ## s ## _compare_internal(a, b, 1 STATUS_VAR); \ 7138 } 7139 7140 COMPARE(32, 0xff) 7141 COMPARE(64, 0x7ff) 7142 7143 static inline int floatx80_compare_internal( floatx80 a, floatx80 b, 7144 int is_quiet STATUS_PARAM ) 7145 { 7146 flag aSign, bSign; 7147 7148 if (( ( extractFloatx80Exp( a ) == 0x7fff ) && 7149 ( extractFloatx80Frac( a )<<1 ) ) || 7150 ( ( extractFloatx80Exp( b ) == 0x7fff ) && 7151 ( extractFloatx80Frac( b )<<1 ) )) { 7152 if (!is_quiet || 7153 floatx80_is_signaling_nan( a ) || 7154 floatx80_is_signaling_nan( b ) ) { 7155 float_raise( float_flag_invalid STATUS_VAR); 7156 } 7157 return float_relation_unordered; 7158 } 7159 aSign = extractFloatx80Sign( a ); 7160 bSign = extractFloatx80Sign( b ); 7161 if ( aSign != bSign ) { 7162 7163 if ( ( ( (uint16_t) ( ( a.high | b.high ) << 1 ) ) == 0) && 7164 ( ( a.low | b.low ) == 0 ) ) { 7165 /* zero case */ 7166 return float_relation_equal; 7167 } else { 7168 return 1 - (2 * aSign); 7169 } 7170 } else { 7171 if (a.low == b.low && a.high == b.high) { 7172 return float_relation_equal; 7173 } else { 7174 return 1 - 2 * (aSign ^ ( lt128( a.high, a.low, b.high, b.low ) )); 7175 } 7176 } 7177 } 7178 7179 int floatx80_compare( floatx80 a, floatx80 b STATUS_PARAM ) 7180 { 7181 return floatx80_compare_internal(a, b, 0 STATUS_VAR); 7182 } 7183 7184 int floatx80_compare_quiet( floatx80 a, floatx80 b STATUS_PARAM ) 7185 { 7186 return floatx80_compare_internal(a, b, 1 STATUS_VAR); 7187 } 7188 7189 static inline int float128_compare_internal( float128 a, float128 b, 7190 int is_quiet STATUS_PARAM ) 7191 { 7192 flag aSign, bSign; 7193 7194 if (( ( extractFloat128Exp( a ) == 0x7fff ) && 7195 ( extractFloat128Frac0( a ) | extractFloat128Frac1( a ) ) ) || 7196 ( ( extractFloat128Exp( b ) == 0x7fff ) && 7197 ( extractFloat128Frac0( b ) | extractFloat128Frac1( b ) ) )) { 7198 if (!is_quiet || 7199 float128_is_signaling_nan( a ) || 7200 float128_is_signaling_nan( b ) ) { 7201 float_raise( float_flag_invalid STATUS_VAR); 7202 } 7203 return float_relation_unordered; 7204 } 7205 aSign = extractFloat128Sign( a ); 7206 bSign = extractFloat128Sign( b ); 7207 if ( aSign != bSign ) { 7208 if ( ( ( ( a.high | b.high )<<1 ) | a.low | b.low ) == 0 ) { 7209 /* zero case */ 7210 return float_relation_equal; 7211 } else { 7212 return 1 - (2 * aSign); 7213 } 7214 } else { 7215 if (a.low == b.low && a.high == b.high) { 7216 return float_relation_equal; 7217 } else { 7218 return 1 - 2 * (aSign ^ ( lt128( a.high, a.low, b.high, b.low ) )); 7219 } 7220 } 7221 } 7222 7223 int float128_compare( float128 a, float128 b STATUS_PARAM ) 7224 { 7225 return float128_compare_internal(a, b, 0 STATUS_VAR); 7226 } 7227 7228 int float128_compare_quiet( float128 a, float128 b STATUS_PARAM ) 7229 { 7230 return float128_compare_internal(a, b, 1 STATUS_VAR); 7231 } 7232 7233 /* min() and max() functions. These can't be implemented as 7234 * 'compare and pick one input' because that would mishandle 7235 * NaNs and +0 vs -0. 7236 * 7237 * minnum() and maxnum() functions. These are similar to the min() 7238 * and max() functions but if one of the arguments is a QNaN and 7239 * the other is numerical then the numerical argument is returned. 7240 * minnum() and maxnum correspond to the IEEE 754-2008 minNum() 7241 * and maxNum() operations. min() and max() are the typical min/max 7242 * semantics provided by many CPUs which predate that specification. 7243 * 7244 * minnummag() and maxnummag() functions correspond to minNumMag() 7245 * and minNumMag() from the IEEE-754 2008. 7246 */ 7247 #define MINMAX(s) \ 7248 static inline float ## s float ## s ## _minmax(float ## s a, float ## s b, \ 7249 int ismin, int isieee, \ 7250 int ismag STATUS_PARAM) \ 7251 { \ 7252 flag aSign, bSign; \ 7253 uint ## s ## _t av, bv, aav, abv; \ 7254 a = float ## s ## _squash_input_denormal(a STATUS_VAR); \ 7255 b = float ## s ## _squash_input_denormal(b STATUS_VAR); \ 7256 if (float ## s ## _is_any_nan(a) || \ 7257 float ## s ## _is_any_nan(b)) { \ 7258 if (isieee) { \ 7259 if (float ## s ## _is_quiet_nan(a) && \ 7260 !float ## s ##_is_any_nan(b)) { \ 7261 return b; \ 7262 } else if (float ## s ## _is_quiet_nan(b) && \ 7263 !float ## s ## _is_any_nan(a)) { \ 7264 return a; \ 7265 } \ 7266 } \ 7267 return propagateFloat ## s ## NaN(a, b STATUS_VAR); \ 7268 } \ 7269 aSign = extractFloat ## s ## Sign(a); \ 7270 bSign = extractFloat ## s ## Sign(b); \ 7271 av = float ## s ## _val(a); \ 7272 bv = float ## s ## _val(b); \ 7273 if (ismag) { \ 7274 aav = float ## s ## _abs(av); \ 7275 abv = float ## s ## _abs(bv); \ 7276 if (aav != abv) { \ 7277 if (ismin) { \ 7278 return (aav < abv) ? a : b; \ 7279 } else { \ 7280 return (aav < abv) ? b : a; \ 7281 } \ 7282 } \ 7283 } \ 7284 if (aSign != bSign) { \ 7285 if (ismin) { \ 7286 return aSign ? a : b; \ 7287 } else { \ 7288 return aSign ? b : a; \ 7289 } \ 7290 } else { \ 7291 if (ismin) { \ 7292 return (aSign ^ (av < bv)) ? a : b; \ 7293 } else { \ 7294 return (aSign ^ (av < bv)) ? b : a; \ 7295 } \ 7296 } \ 7297 } \ 7298 \ 7299 float ## s float ## s ## _min(float ## s a, float ## s b STATUS_PARAM) \ 7300 { \ 7301 return float ## s ## _minmax(a, b, 1, 0, 0 STATUS_VAR); \ 7302 } \ 7303 \ 7304 float ## s float ## s ## _max(float ## s a, float ## s b STATUS_PARAM) \ 7305 { \ 7306 return float ## s ## _minmax(a, b, 0, 0, 0 STATUS_VAR); \ 7307 } \ 7308 \ 7309 float ## s float ## s ## _minnum(float ## s a, float ## s b STATUS_PARAM) \ 7310 { \ 7311 return float ## s ## _minmax(a, b, 1, 1, 0 STATUS_VAR); \ 7312 } \ 7313 \ 7314 float ## s float ## s ## _maxnum(float ## s a, float ## s b STATUS_PARAM) \ 7315 { \ 7316 return float ## s ## _minmax(a, b, 0, 1, 0 STATUS_VAR); \ 7317 } \ 7318 \ 7319 float ## s float ## s ## _minnummag(float ## s a, float ## s b STATUS_PARAM) \ 7320 { \ 7321 return float ## s ## _minmax(a, b, 1, 1, 1 STATUS_VAR); \ 7322 } \ 7323 \ 7324 float ## s float ## s ## _maxnummag(float ## s a, float ## s b STATUS_PARAM) \ 7325 { \ 7326 return float ## s ## _minmax(a, b, 0, 1, 1 STATUS_VAR); \ 7327 } 7328 7329 MINMAX(32) 7330 MINMAX(64) 7331 7332 7333 /* Multiply A by 2 raised to the power N. */ 7334 float32 float32_scalbn( float32 a, int n STATUS_PARAM ) 7335 { 7336 flag aSign; 7337 int16_t aExp; 7338 uint32_t aSig; 7339 7340 a = float32_squash_input_denormal(a STATUS_VAR); 7341 aSig = extractFloat32Frac( a ); 7342 aExp = extractFloat32Exp( a ); 7343 aSign = extractFloat32Sign( a ); 7344 7345 if ( aExp == 0xFF ) { 7346 if ( aSig ) { 7347 return propagateFloat32NaN( a, a STATUS_VAR ); 7348 } 7349 return a; 7350 } 7351 if (aExp != 0) { 7352 aSig |= 0x00800000; 7353 } else if (aSig == 0) { 7354 return a; 7355 } else { 7356 aExp++; 7357 } 7358 7359 if (n > 0x200) { 7360 n = 0x200; 7361 } else if (n < -0x200) { 7362 n = -0x200; 7363 } 7364 7365 aExp += n - 1; 7366 aSig <<= 7; 7367 return normalizeRoundAndPackFloat32( aSign, aExp, aSig STATUS_VAR ); 7368 } 7369 7370 float64 float64_scalbn( float64 a, int n STATUS_PARAM ) 7371 { 7372 flag aSign; 7373 int16_t aExp; 7374 uint64_t aSig; 7375 7376 a = float64_squash_input_denormal(a STATUS_VAR); 7377 aSig = extractFloat64Frac( a ); 7378 aExp = extractFloat64Exp( a ); 7379 aSign = extractFloat64Sign( a ); 7380 7381 if ( aExp == 0x7FF ) { 7382 if ( aSig ) { 7383 return propagateFloat64NaN( a, a STATUS_VAR ); 7384 } 7385 return a; 7386 } 7387 if (aExp != 0) { 7388 aSig |= LIT64( 0x0010000000000000 ); 7389 } else if (aSig == 0) { 7390 return a; 7391 } else { 7392 aExp++; 7393 } 7394 7395 if (n > 0x1000) { 7396 n = 0x1000; 7397 } else if (n < -0x1000) { 7398 n = -0x1000; 7399 } 7400 7401 aExp += n - 1; 7402 aSig <<= 10; 7403 return normalizeRoundAndPackFloat64( aSign, aExp, aSig STATUS_VAR ); 7404 } 7405 7406 floatx80 floatx80_scalbn( floatx80 a, int n STATUS_PARAM ) 7407 { 7408 flag aSign; 7409 int32_t aExp; 7410 uint64_t aSig; 7411 7412 aSig = extractFloatx80Frac( a ); 7413 aExp = extractFloatx80Exp( a ); 7414 aSign = extractFloatx80Sign( a ); 7415 7416 if ( aExp == 0x7FFF ) { 7417 if ( aSig<<1 ) { 7418 return propagateFloatx80NaN( a, a STATUS_VAR ); 7419 } 7420 return a; 7421 } 7422 7423 if (aExp == 0) { 7424 if (aSig == 0) { 7425 return a; 7426 } 7427 aExp++; 7428 } 7429 7430 if (n > 0x10000) { 7431 n = 0x10000; 7432 } else if (n < -0x10000) { 7433 n = -0x10000; 7434 } 7435 7436 aExp += n; 7437 return normalizeRoundAndPackFloatx80( STATUS(floatx80_rounding_precision), 7438 aSign, aExp, aSig, 0 STATUS_VAR ); 7439 } 7440 7441 float128 float128_scalbn( float128 a, int n STATUS_PARAM ) 7442 { 7443 flag aSign; 7444 int32_t aExp; 7445 uint64_t aSig0, aSig1; 7446 7447 aSig1 = extractFloat128Frac1( a ); 7448 aSig0 = extractFloat128Frac0( a ); 7449 aExp = extractFloat128Exp( a ); 7450 aSign = extractFloat128Sign( a ); 7451 if ( aExp == 0x7FFF ) { 7452 if ( aSig0 | aSig1 ) { 7453 return propagateFloat128NaN( a, a STATUS_VAR ); 7454 } 7455 return a; 7456 } 7457 if (aExp != 0) { 7458 aSig0 |= LIT64( 0x0001000000000000 ); 7459 } else if (aSig0 == 0 && aSig1 == 0) { 7460 return a; 7461 } else { 7462 aExp++; 7463 } 7464 7465 if (n > 0x10000) { 7466 n = 0x10000; 7467 } else if (n < -0x10000) { 7468 n = -0x10000; 7469 } 7470 7471 aExp += n - 1; 7472 return normalizeRoundAndPackFloat128( aSign, aExp, aSig0, aSig1 7473 STATUS_VAR ); 7474 7475 } 7476