1<html> 2<head> 3<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> 4<title>Compilers</title> 5<link rel="stylesheet" href="../math.css" type="text/css"> 6<meta name="generator" content="DocBook XSL Stylesheets V1.79.1"> 7<link rel="home" href="../index.html" title="Math Toolkit 2.12.0"> 8<link rel="up" href="../overview.html" title="Chapter 1. Overview"> 9<link rel="prev" href="error_handling.html" title="Error Handling"> 10<link rel="next" href="config_macros.html" title="Configuration Macros"> 11</head> 12<body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"> 13<table cellpadding="2" width="100%"><tr> 14<td valign="top"><img alt="Boost C++ Libraries" width="277" height="86" src="../../../../../boost.png"></td> 15<td align="center"><a href="../../../../../index.html">Home</a></td> 16<td align="center"><a href="../../../../../libs/libraries.htm">Libraries</a></td> 17<td align="center"><a href="http://www.boost.org/users/people.html">People</a></td> 18<td align="center"><a href="http://www.boost.org/users/faq.html">FAQ</a></td> 19<td align="center"><a href="../../../../../more/index.htm">More</a></td> 20</tr></table> 21<hr> 22<div class="spirit-nav"> 23<a accesskey="p" href="error_handling.html"><img src="../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../overview.html"><img src="../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../index.html"><img src="../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="config_macros.html"><img src="../../../../../doc/src/images/next.png" alt="Next"></a> 24</div> 25<div class="section"> 26<div class="titlepage"><div><div><h2 class="title" style="clear: both"> 27<a name="math_toolkit.compilers_overview"></a><a class="link" href="compilers_overview.html" title="Compilers">Compilers</a> 28</h2></div></div></div> 29<p> 30 This section contains some information about how various compilers work with 31 this library. It is not comprehensive and updated experiences are always welcome. 32 Some effort has been made to suppress unhelpful warnings but it is difficult 33 to achieve this on all systems. 34 </p> 35<div class="table"> 36<a name="math_toolkit.compilers_overview.supported_tested_compilers"></a><p class="title"><b>Table 1.9. Supported/Tested Compilers</b></p> 37<div class="table-contents"><table class="table" summary="Supported/Tested Compilers"> 38<colgroup> 39<col> 40<col> 41<col> 42<col> 43</colgroup> 44<thead><tr> 45<th> 46 <p> 47 Platform 48 </p> 49 </th> 50<th> 51 <p> 52 Compiler 53 </p> 54 </th> 55<th> 56 <p> 57 Has long double support 58 </p> 59 </th> 60<th> 61 <p> 62 Notes 63 </p> 64 </th> 65</tr></thead> 66<tbody> 67<tr> 68<td> 69 <p> 70 Windows 71 </p> 72 </td> 73<td> 74 <p> 75 MSVC 7.1 and later 76 </p> 77 </td> 78<td> 79 <p> 80 Yes 81 </p> 82 </td> 83<td> 84 <p> 85 All tests OK. 86 </p> 87 <p> 88 We aim to keep our headers warning free at level 4 with this compiler. 89 </p> 90 </td> 91</tr> 92<tr> 93<td> 94 <p> 95 Windows 96 </p> 97 </td> 98<td> 99 <p> 100 Intel 8.1 and later 101 </p> 102 </td> 103<td> 104 <p> 105 Yes 106 </p> 107 </td> 108<td> 109 <p> 110 All tests OK. 111 </p> 112 <p> 113 We aim to keep our headers warning free at level 4 with this compiler. 114 However, The tests cases tend to generate a lot of warnings relating 115 to numeric underflow of the test data: these are harmless. 116 </p> 117 </td> 118</tr> 119<tr> 120<td> 121 <p> 122 Windows 123 </p> 124 </td> 125<td> 126 <p> 127 GNU Mingw32 C++ 128 </p> 129 </td> 130<td> 131 <p> 132 Yes 133 </p> 134 </td> 135<td> 136 <p> 137 All tests OK. 138 </p> 139 <p> 140 We aim to keep our headers warning free with -Wall with this compiler. 141 </p> 142 </td> 143</tr> 144<tr> 145<td> 146 <p> 147 Windows 148 </p> 149 </td> 150<td> 151 <p> 152 GNU Cygwin C++ 153 </p> 154 </td> 155<td> 156 <p> 157 No 158 </p> 159 </td> 160<td> 161 <p> 162 All tests OK. 163 </p> 164 <p> 165 We aim to keep our headers warning free with -Wall with this compiler. 166 </p> 167 <p> 168 Long double support has been disabled because there are no native 169 long double C std library functions available. 170 </p> 171 </td> 172</tr> 173<tr> 174<td> 175 <p> 176 Windows 177 </p> 178 </td> 179<td> 180 <p> 181 Borland C++ 5.8.2 (Developer studio 2006) 182 </p> 183 </td> 184<td> 185 <p> 186 No 187 </p> 188 </td> 189<td> 190 <p> 191 We have only partial compatibility with this compiler: 192 </p> 193 <p> 194 Long double support has been disabled because the native long double 195 C standard library functions really only forward to the double versions. 196 This can result in unpredictable behaviour when using the long double 197 overloads: for example <code class="computeroutput"><span class="identifier">sqrtl</span></code> 198 applied to a finite value, can result in an infinite result. 199 </p> 200 <p> 201 Some functions still fail to compile, there are no known workarounds 202 at present. 203 </p> 204 </td> 205</tr> 206<tr> 207<td> 208 <p> 209 Windows 7/Netbeans 7.2 210 </p> 211 </td> 212<td> 213 <p> 214 Clang 3.1 215 </p> 216 </td> 217<td> 218 <p> 219 Yes 220 </p> 221 </td> 222<td> 223 <p> 224 Spot examples OK. Expect all tests to compile and run OK. 225 </p> 226 </td> 227</tr> 228<tr> 229<td> 230 <p> 231 Linux 232 </p> 233 </td> 234<td> 235 <p> 236 GNU C++ 3.4 and later 237 </p> 238 </td> 239<td> 240 <p> 241 Yes 242 </p> 243 </td> 244<td> 245 <p> 246 All tests OK. 247 </p> 248 <p> 249 We aim to keep our headers warning free with -Wall with this compiler. 250 </p> 251 </td> 252</tr> 253<tr> 254<td> 255 <p> 256 Linux 257 </p> 258 </td> 259<td> 260 <p> 261 Clang 3.2 262 </p> 263 </td> 264<td> 265 <p> 266 Yes 267 </p> 268 </td> 269<td> 270 <p> 271 All tests OK. 272 </p> 273 </td> 274</tr> 275<tr> 276<td> 277 <p> 278 Linux 279 </p> 280 </td> 281<td> 282 <p> 283 Intel C++ 10.0 and later 284 </p> 285 </td> 286<td> 287 <p> 288 Yes 289 </p> 290 </td> 291<td> 292 <p> 293 All tests OK. 294 </p> 295 <p> 296 We aim to keep our headers warning free with -Wall with this compiler. 297 However, The tests cases tend to generate a lot of warnings relating 298 to numeric underflow of the test data: these are harmless. 299 </p> 300 </td> 301</tr> 302<tr> 303<td> 304 <p> 305 Linux 306 </p> 307 </td> 308<td> 309 <p> 310 Intel C++ 8.1 and 9.1 311 </p> 312 </td> 313<td> 314 <p> 315 No 316 </p> 317 </td> 318<td> 319 <p> 320 All tests OK. 321 </p> 322 <p> 323 Long double support has been disabled with these compiler releases 324 because calling the standard library long double math functions can 325 result in a segfault. The issue is Linux distribution and glibc version 326 specific and is Intel bug report #409291. Fully up to date releases 327 of Intel 9.1 (post version l_cc_c_9.1.046) shouldn't have this problem. 328 If you need long double support with this compiler, then comment 329 out the define of BOOST_MATH_NO_LONG_DOUBLE_MATH_FUNCTIONS at line 330 55 of <a href="../../../../../boost/math/tools/config.hpp" target="_top">boost/math/tools/config.hpp</a>. 331 </p> 332 <p> 333 We aim to keep our headers warning free with -Wall with this compiler. 334 However, The tests cases tend to generate a lot of warnings relating 335 to numeric underflow of the test data: these are harmless. 336 </p> 337 </td> 338</tr> 339<tr> 340<td> 341 <p> 342 Linux 343 </p> 344 </td> 345<td> 346 <p> 347 QLogic PathScale 3.0 348 </p> 349 </td> 350<td> 351 <p> 352 Yes 353 </p> 354 </td> 355<td> 356 <p> 357 Some tests involving conceptual checks fail to build, otherwise there 358 appear to be no issues. 359 </p> 360 </td> 361</tr> 362<tr> 363<td> 364 <p> 365 Linux 366 </p> 367 </td> 368<td> 369 <p> 370 Sun Studio 12 371 </p> 372 </td> 373<td> 374 <p> 375 Yes 376 </p> 377 </td> 378<td> 379 <p> 380 Some tests involving function overload resolution fail to build, 381 these issues should be rarely encountered in practice. 382 </p> 383 </td> 384</tr> 385<tr> 386<td> 387 <p> 388 Solaris 389 </p> 390 </td> 391<td> 392 <p> 393 Sun Studio 12 394 </p> 395 </td> 396<td> 397 <p> 398 Yes 399 </p> 400 </td> 401<td> 402 <p> 403 Some tests involving function overload resolution fail to build, 404 these issues should be rarely encountered in practice. 405 </p> 406 </td> 407</tr> 408<tr> 409<td> 410 <p> 411 Solaris 412 </p> 413 </td> 414<td> 415 <p> 416 GNU C++ 4.x 417 </p> 418 </td> 419<td> 420 <p> 421 Yes 422 </p> 423 </td> 424<td> 425 <p> 426 All tests OK. 427 </p> 428 <p> 429 We aim to keep our headers warning free with -Wall with this compiler. 430 </p> 431 </td> 432</tr> 433<tr> 434<td> 435 <p> 436 HP Tru64 437 </p> 438 </td> 439<td> 440 <p> 441 Compaq C++ 7.1 442 </p> 443 </td> 444<td> 445 <p> 446 Yes 447 </p> 448 </td> 449<td> 450 <p> 451 All tests OK. 452 </p> 453 </td> 454</tr> 455<tr> 456<td> 457 <p> 458 HP-UX Itanium 459 </p> 460 </td> 461<td> 462 <p> 463 HP aCC 6.x 464 </p> 465 </td> 466<td> 467 <p> 468 Yes 469 </p> 470 </td> 471<td> 472 <p> 473 All tests OK. 474 </p> 475 <p> 476 Unfortunately this compiler emits quite a few warnings from libraries 477 upon which we depend (TR1, Array etc). 478 </p> 479 </td> 480</tr> 481<tr> 482<td> 483 <p> 484 HP-UX PA-RISC 485 </p> 486 </td> 487<td> 488 <p> 489 GNU C++ 3.4 490 </p> 491 </td> 492<td> 493 <p> 494 No 495 </p> 496 </td> 497<td> 498 <p> 499 All tests OK. 500 </p> 501 </td> 502</tr> 503<tr> 504<td> 505 <p> 506 Apple Mac OS X, Intel 507 </p> 508 </td> 509<td> 510 <p> 511 Darwin/GNU C++ 4.x 512 </p> 513 </td> 514<td> 515 <p> 516 Yes 517 </p> 518 </td> 519<td> 520 <p> 521 All tests OK. 522 </p> 523 </td> 524</tr> 525<tr> 526<td> 527 <p> 528 Apple Mac OS X, PowerPC 529 </p> 530 </td> 531<td> 532 <p> 533 Darwin/GNU C++ 4.x 534 </p> 535 </td> 536<td> 537 <p> 538 No 539 </p> 540 </td> 541<td> 542 <p> 543 All tests OK. 544 </p> 545 <p> 546 Long double support has been disabled on this platform due to the 547 rather strange nature of Darwin's 106-bit long double implementation. 548 It should be possible to make this work if someone is prepared to 549 offer assistance. 550 </p> 551 </td> 552</tr> 553<tr> 554<td> 555 <p> 556 Apple Mac OS X, 557 </p> 558 </td> 559<td> 560 <p> 561 Clang 3.2 562 </p> 563 </td> 564<td> 565 <p> 566 Yes 567 </p> 568 </td> 569<td> 570 <p> 571 All tests expected to be OK. 572 </p> 573 </td> 574</tr> 575<tr> 576<td> 577 <p> 578 IBM AIX 579 </p> 580 </td> 581<td> 582 <p> 583 IBM xlc 5.3 584 </p> 585 </td> 586<td> 587 <p> 588 Yes 589 </p> 590 </td> 591<td> 592 <p> 593 All tests pass except for our fpclassify tests which fail due to 594 a bug in <code class="computeroutput"><span class="identifier">std</span><span class="special">::</span><span class="identifier">numeric_limits</span></code>, the bug effects 595 the test code, not fpclassify itself. The IBM compiler group are 596 aware of the problem. 597 </p> 598 </td> 599</tr> 600</tbody> 601</table></div> 602</div> 603<br class="table-break"><div class="table"> 604<a name="math_toolkit.compilers_overview.unsupported_compilers"></a><p class="title"><b>Table 1.10. Unsupported Compilers</b></p> 605<div class="table-contents"><table class="table" summary="Unsupported Compilers"> 606<colgroup> 607<col> 608<col> 609</colgroup> 610<thead><tr> 611<th> 612 <p> 613 Platform 614 </p> 615 </th> 616<th> 617 <p> 618 Compiler 619 </p> 620 </th> 621</tr></thead> 622<tbody> 623<tr> 624<td> 625 <p> 626 Windows 627 </p> 628 </td> 629<td> 630 <p> 631 Borland C++ 5.9.2 (Borland Developer Studio 2007) 632 </p> 633 </td> 634</tr> 635<tr> 636<td> 637 <p> 638 Windows 639 </p> 640 </td> 641<td> 642 <p> 643 MSVC 6 and 7 644 </p> 645 </td> 646</tr> 647</tbody> 648</table></div> 649</div> 650<br class="table-break"><p> 651 If your compiler or platform is not listed above, please try running the regression 652 tests: cd into boost-root/libs/math/test and do a: 653 </p> 654<pre class="programlisting"><span class="identifier">bjam</span> <span class="identifier">mytoolset</span> 655</pre> 656<p> 657 where "mytoolset" is the name of the <a href="http://www.boost.org/doc/html/bbv2.html" target="_top">Boost.Build</a> 658 toolset used for your compiler. The chances are that <span class="bold"><strong>many 659 of the accuracy tests will fail at this stage</strong></span> - don't panic - the 660 default acceptable error tolerances are quite tight, especially for long double 661 types with an extended exponent range (these cause more extreme test cases 662 to be executed for some functions). You will need to cast an eye over the output 663 from the failing tests and make a judgement as to whether the error rates are 664 acceptable or not. 665 </p> 666</div> 667<table xmlns:rev="http://www.cs.rpi.edu/~gregod/boost/tools/doc/revision" width="100%"><tr> 668<td align="left"></td> 669<td align="right"><div class="copyright-footer">Copyright © 2006-2019 Nikhar 670 Agrawal, Anton Bikineev, Paul A. Bristow, Marco Guazzone, Christopher Kormanyos, 671 Hubert Holin, Bruno Lalande, John Maddock, Jeremy Murphy, Matthew Pulver, Johan 672 Råde, Gautam Sewani, Benjamin Sobotta, Nicholas Thompson, Thijs van den Berg, 673 Daryle Walker and Xiaogang Zhang<p> 674 Distributed under the Boost Software License, Version 1.0. (See accompanying 675 file LICENSE_1_0.txt or copy at <a href="http://www.boost.org/LICENSE_1_0.txt" target="_top">http://www.boost.org/LICENSE_1_0.txt</a>) 676 </p> 677</div></td> 678</tr></table> 679<hr> 680<div class="spirit-nav"> 681<a accesskey="p" href="error_handling.html"><img src="../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../overview.html"><img src="../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../index.html"><img src="../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="config_macros.html"><img src="../../../../../doc/src/images/next.png" alt="Next"></a> 682</div> 683</body> 684</html> 685
