1 <?xml version="1.0" encoding="utf-8"?>
2 <explicit-failures-markup xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
3 xsi:noNamespaceSchemaLocation="explicit-failures.xsd">
6 PLEASE VALIDATE THE XML BEFORE COMMITTING YOUR CHANGES!
8 Locally, the xmlint tool can be used:
10 xmllint <two-dashes>valid explicit-failures-markup.xml <two-dashes>schema explicit-failures.xsd
12 The following online services can be used to validate your changes to this
15 - http://syseng.nist.gov/b2bTestbed/projects/xmlvalidation/instance_validation.html
16 - http://xmlvalidation.com/
18 With both tools you need to provide both the explicit-failures-markup.xml
19 file as the XML document and the explicit-failures.xsd as the schema
20 document. Use the browse buttons to select them from your local hard
24 <!-- /////////////// Toolsets /////////////// -->
25 <mark-toolset name="acc" status="required"/>
26 <mark-toolset name="darwin-4.0.1" status="required"/>
27 <mark-toolset name="gcc-4.1.2_sunos_i86pc" status="required"/>
28 <mark-toolset name="gcc-4.1.3_linux" status="required"/>
29 <mark-toolset name="gcc-4.2.1" status="required"/>
30 <mark-toolset name="gcc-4.2.1_hpux_ia64" status="required"/>
31 <mark-toolset name="gcc-4.2.1_linux_x86_64" status="required"/>
32 <mark-toolset name="intel-linux-9.0" status="required"/>
33 <mark-toolset name="intel-vc8-win-10.0" status="required"/>
34 <mark-toolset name="intel-win-10.0" status="required"/>
35 <mark-toolset name="msvc-7.1" status="required"/>
36 <mark-toolset name="msvc-8.0" status="required"/>
37 <mark-toolset name="msvc-8.0_64" status="required"/>
39 <!-- /////////////// Libraries /////////////// -->
42 <library name="accumulators">
44 <toolset name="sun-5.7"/>
45 <toolset name="sun-5.8"/>
46 <toolset name="sun-5.9"/>
47 <toolset name="borland-*"/>
49 <mark-expected-failures>
50 <test name="tail_variate_means"/>
51 <test name="weighted_tail_variate_means"/>
52 <toolset name="gcc-4.2.1*"/>
53 <note author="Boris Gubenko" refid="42"/>
54 </mark-expected-failures>
55 <mark-expected-failures>
56 <test name="weighted_kurtosis"/>
58 <note author="Boris Gubenko" refid="38"/>
59 </mark-expected-failures>
60 <mark-expected-failures>
61 <test name="weighted_tail_variate_means"/>
62 <toolset name="hp_cxx-71*"/>
63 <note author="Markus Schoepflin">
64 This failure is caused by a timeout when compiling the test. It
65 passes when the timeout value is increased.
67 </mark-expected-failures>
68 <mark-expected-failures>
69 <test name="covariance"/>
70 <test name="pot_quantile"/>
71 <test name="tail_variate_means"/>
72 <test name="weighted_covariance"/>
73 <test name="weighted_pot_quantile"/>
74 <test name="weighted_tail_variate_means"/>
76 <note author="Boris Gubenko" refid="47"/>
77 </mark-expected-failures>
81 <library name="algorithm/minmax">
83 <toolset name="sunpro-5_3-sunos"/>
88 <library name="algorithm/string">
90 <toolset name="borland-5.5*"/>
91 <toolset name="msvc-6.5*"/>
92 <toolset name="msvc-7.1_stlport4"/>
93 <toolset name="iw-7_1-vc6"/>
94 <toolset name="gcc-2.95.3-linux"/>
95 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
96 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
97 <toolset name="mipspro"/>
98 <toolset name="sunpro-5_3-sunos"/>
99 <note author="P.Droba">
100 The compiler does not support features that are essential for the library.
105 <toolset name="borland-5.9*"/>
106 <toolset name="borland-5.8*"/>
107 <toolset name="borland-5.6*"/>
108 <note author="P.Droba">
109 The toolset is not supported by Boost.Regex.
117 <test name="any_to_ref_test">
119 <toolset name="msvc-6.5*"/>
120 <note author="Vladimir Prus">
121 The test fail with ICE, but the exact reason for ICE is not
122 known. A minimal example of casting any to reference type
123 seem to work. Anyone interested in using this functionality
124 with msvc is suggested to do additional testing.
132 <library name="array">
135 <toolset name="msvc-6.5"/>
136 <toolset name="msvc-6.5_stlport4"/>
137 <toolset name="msvc-7.0"/>
138 <note author="A.Meredith">
139 Compilers need to support partial template specialization
140 to work with zero length arrays.
146 <toolset name="borland-5.5*"/>
147 <toolset name="borland-5.6*"/>
148 <toolset name="msvc-6.5"/>
149 <toolset name="msvc-6.5_stlport4"/>
150 <toolset name="msvc-7.0"/>
154 <toolset name="sunpro-5_3-sunos"/>
160 <toolset name="borland-5.5*"/>
161 <toolset name="borland-5.6*"/>
162 <toolset name="msvc-6.5"/>
163 <toolset name="msvc-6.5_stlport4"/>
164 <toolset name="msvc-7.0"/>
171 <library name="asio">
173 <toolset name="borland-5.6*"/>
174 <toolset name="borland-5.8*"/>
175 <note author="Chris Kohlhoff">
176 This compiler does not support enable_if, which is needed by the
177 Boost.System library on which Boost.Asio depends.
180 <mark-expected-failures>
181 <test name="read_until"/>
182 <test name="read_until_select"/>
183 <toolset name="gcc-4.2.1_hpux_ia64"/>
184 <note author="Boris Gubenko">
185 On HP-UX 11.23 platform, these tests must be compiled with
186 _XOPEN_SOURCE_EXTENDED macro defined. It is likely related
189 </mark-expected-failures>
193 <library name="assign">
195 <toolset name="dmc-8_43-stlport-4_5_3"/>
197 <mark-expected-failures>
199 <toolset name="msvc-6.5_stlport4"/>
200 <toolset name="msvc-7.0"/>
201 <note author="Thorsten Ottosen" >
202 The test would (most likely) compile and run properly if the workaround
203 syntax .to_container( c ) was applied to all list_of() expressions.
205 </mark-expected-failures>
206 <mark-expected-failures>
207 <test name="email_example"/>
208 <toolset name="gcc-2.95.3*"/>
209 <note refid="27" author="Thorsten Ottosen"/>
210 </mark-expected-failures>
211 <mark-expected-failures>
212 <test name="list_inserter"/>
213 <toolset name="msvc-7.0"/>
214 <toolset name="sunpro-5_3-sunos"/>
215 <toolset name="hp_cxx-65*"/>
216 <note refid="6" author="Thorsten Ottosen"/>
217 </mark-expected-failures>
218 <mark-expected-failures>
219 <test name="list_inserter"/>
220 <toolset name="gcc-2.95.3*"/>
221 <note author="Thorsten Ottosen">
222 This test could probably be made to work if somebody with knowledge
223 about the compilers would submit a patch.
225 </mark-expected-failures>
226 <mark-expected-failures>
227 <test name="list_of"/>
228 <toolset name="sunpro-5_3-sunos"/>
229 <toolset name="hp_cxx-65*"/>
230 <toolset name="borland-5*"/>
231 <toolset name="msvc-6.5*"/>
232 <toolset name="msvc-7.0"/>
233 <note author="Thorsten Ottosen" >
234 The test would (most likely) compile and run properly if the workaround
235 syntax .to_container( c ) was applied to all list_of() expressions.
237 </mark-expected-failures>
239 <mark-expected-failures>
240 <test name="list_of_workaround"/>
241 <toolset name="sunpro-5_3-sunos"/>
242 <note author="Thorsten Ottosen" >
243 The test could probably be made to work if somebody submitted a patch.
245 </mark-expected-failures>
246 <mark-expected-failures>
247 <test name="multi_index_container"/>
248 <toolset name="sunpro-5_3-sunos"/>
249 <toolset name="sun-5.8"/>
250 <toolset name="hp_cxx-65*"/>
251 <toolset name="borland-5*"/>
252 <toolset name="gcc-2.95.3*"/>
253 <note refid="27" author="Thorsten Ottosen"/>
254 </mark-expected-failures>
255 <mark-expected-failures>
256 <test name="multi_index_container"/>
257 <toolset name="msvc-6.5*"/>
258 <toolset name="msvc-7.0"/>
259 <toolset name="mipspro"/>
260 <toolset name="hp_cxx-65*"/>
261 <note author="Thorsten Ottosen" >
262 The test would (most likely) compile and run properly if the workaround
263 syntax .to_container( c ) was applied to all list_of() expressions.
265 </mark-expected-failures>
266 <mark-expected-failures>
267 <test name="my_vector_example"/>
268 <toolset name="gcc-2.95.3*"/>
269 <note refid="27" author="Thorsten Ottosen"/>
270 </mark-expected-failures>
271 <mark-expected-failures>
272 <test name="ptr_list_inserter"/>
273 <toolset name="sunpro-5_3-sunos"/>
274 <toolset name="hp_cxx-65*"/>
275 <toolset name="borland-5*"/>
276 <toolset name="gcc-2.95.3*"/>
277 <toolset name="msvc-6.5*"/>
278 <toolset name="msvc-7.0"/>
279 <toolset name="mipspro"/>
280 <note author="Thorsten Ottosen" >
281 The test depends on Boost.Pointer Container which probably does not work for
284 </mark-expected-failures>
285 <mark-expected-failures>
286 <test name="ptr_list_of"/>
287 <toolset name="hp_cxx-65*"/>
288 <toolset name="borland-5*"/>
289 <toolset name="gcc-2.95.3*"/>
290 <toolset name="msvc-6.5*"/>
291 <toolset name="mipspro"/>
292 <note author="Thorsten Ottosen" >
293 The test depends on Boost.Pointer Container which probably does not work for
296 </mark-expected-failures>
297 <mark-expected-failures>
298 <test name="ptr_map_inserter"/>
299 <toolset name="hp_cxx-65*"/>
300 <toolset name="borland-5*"/>
301 <toolset name="gcc-2.95.3*"/>
302 <toolset name="msvc-6.5*"/>
303 <toolset name="mipspro"/>
304 <note author="Thorsten Ottosen" >
305 The test depends on Boost.Pointer Container which probably does not work for
308 </mark-expected-failures>
309 <mark-expected-failures>
311 <toolset name="sunpro-5_3-sunos"/>
312 <note author="Thorsten Ottosen" >
313 The test does not work for
316 </mark-expected-failures>
317 <mark-expected-failures>
318 <test name="tuple_list_of"/>
319 <toolset name="sunpro-5_3-sunos"/>
320 <toolset name="borland-5*"/>
321 <toolset name="msvc-6.5*"/>
322 <toolset name="msvc-7.0"/>
323 <note author="Thorsten Ottosen" >
324 The test depends on Boost.Tuple which probably does not work for
327 </mark-expected-failures>
331 <library name="bimap">
333 <toolset name="borland-5.6*"/>
334 <note author="J. López" date="05 Jul 2004" refid="17"/>
337 <toolset name="borland-5.8*"/>
338 <note author="Alisdair Meredith" date="26 May 2006"/>
341 <toolset name="borland-5.9*"/>
342 <note author="Alisdair Meredith" date="27 Feb 2007"/>
345 <toolset name="gcc-2.95.3-linux"/>
346 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
347 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
348 <note author="J. López" date="09 Jul 2004" refid="17"/>
351 <toolset name="*como-4_3_3-msvc"/>
352 <note author="J. López" date="30 Jul 2004">
353 The VC++ 6.0 backend runs out of internal resources while
354 trying to process the Comeau output for this library;
355 Comeau Computing has been asked about a solution.
356 On the other hand, Comeau 4.3.3 with VC++ 7.0 backend works
361 <toolset name="sunpro-5_3-sunos"/>
362 <toolset name="sunpro-5_8u1-sunos"/>
363 <note author="J. López" date="22 Apr 2005" refid="17"/>
366 <toolset name="dmc-8_43-stlport-4_5_3"/>
367 <toolset name="dmc-8_44b-stlport-4_5_3"/>
368 <toolset name="dmc-8_47-stlport-4_5_3"/>
369 <note author="J. López" date="03 Jun 2005" refid="17"/>
371 <mark-expected-failures>
372 <test name="test_bimap_assign"/>
373 <test name="test_bimap_ordered"/>
374 <test name="test_bimap_unconstrained"/>
375 <test name="test_bimap_unordered"/>
376 <toolset name="acc"/>
377 <note refid="38" author="Boris Gubenko"/>
378 </mark-expected-failures>
379 <mark-expected-failures>
380 <test name="typeof"/>
381 <toolset name="acc*"/>
382 <toolset name="intel-vc71-win*"/>
383 <toolset name="intel-vc8-win*"/>
384 <toolset name="intel-win-9.1"/>
385 <toolset name="hp_cxx*"/>
386 <note refid="39" author="Boris Gubenko"/>
387 </mark-expected-failures>
388 <mark-expected-failures>
389 <test name="test_bimap_serialization"/>
390 <toolset name="gcc-mingw-3.4.5"/>
391 <note author="Matias Capeletto">Compiler bug.</note>
392 </mark-expected-failures>
393 <mark-expected-failures>
394 <test name="test_bimap_property_map"/>
395 <toolset name="gcc-3.4.6_linux_x86_64"/>
396 <note author="Matias Capeletto">Time out.</note>
397 </mark-expected-failures>
401 <library name="bind">
402 <mark-expected-failures>
403 <test name="bind_cv_test"/>
404 <test name="bind_stateful_test"/>
405 <toolset name="intel-7.1-linux"/>
406 <toolset name="intel-7.1-stdlib-default-linux"/>
407 <note refid="2" author="Aleksey Gurtovoy"/>
408 </mark-expected-failures>
409 <mark-expected-failures>
410 <test name="bind_dm2_test"/>
411 <test name="mem_fn_dm_test"/>
412 <toolset name="msvc-6.*"/>
413 <toolset name="msvc-7.0"/>
414 <toolset name="cw-8.3"/>
415 <note refid="31" author="Peter Dimov"/>
416 </mark-expected-failures>
417 <mark-expected-failures>
418 <test name="bind_dm_test"/>
419 <toolset name="sunpro-5_3-sunos"/>
420 <note refid="31" author="Peter Dimov"/>
421 </mark-expected-failures>
422 <mark-expected-failures>
423 <test name="bind_function_test"/>
424 <toolset name="sunpro-5_8u1-sunos"/>
425 <note author="Peter Dimov">
426 This failure is caused by Boost.Function.
428 </mark-expected-failures>
429 <mark-expected-failures>
430 <test name="mem_fn_derived_test"/>
431 <toolset name="sunpro-5_3-sunos"/>
432 <note refid="31" author="Peter Dimov"/>
433 </mark-expected-failures>
434 <mark-expected-failures>
435 <test name="bind_rv_sp_test"/>
436 <toolset name="hp_cxx-65*"/>
437 <toolset name="hp_cxx-71*"/>
438 <note author="Markus Schoepflin">
439 This failure is caused by a bug in the compiler triggered by the
440 use of the debug flag '-gall'. It has been reported to the
443 </mark-expected-failures>
444 <mark-expected-failures>
445 <test name="bind_dm3_test"/>
446 <toolset name="borland-5*"/>
447 <toolset name="msvc-6.*"/>
448 <toolset name="msvc-7.0"/>
449 <note refid="31" author="Peter Dimov"/>
450 </mark-expected-failures>
451 <mark-expected-failures>
452 <test name="mem_fn_eq_test"/>
453 <toolset name="msvc-7.1"/>
454 <note author="Peter Dimov">
455 This failure is only present in release mode and is caused by /OPT:ICF.
457 </mark-expected-failures>
458 <mark-expected-failures>
459 <test name="bind_placeholder_test"/>
460 <toolset name="borland-*"/>
461 <toolset name="msvc-6.*"/>
462 <toolset name="msvc-7.0"/>
463 <note refid="31" author="Peter Dimov"/>
464 </mark-expected-failures>
468 <library name="build">
469 <test name="timedata">
471 <toolset name="msvc-7.1"/>
472 <toolset name="msvc-8.0"/>
473 <toolset name="msvc-9.0"/>
474 <note author="Vladimir Prus">
475 See https://zigzag.lvk.cs.msu.su:7813/boost.build/ticket/218
482 <library name="chrono">
484 <toolset name="borland-*"/>
485 <toolset name="pathscale*"/>
486 <note author="Vicente J. Botet Escriba">
487 The compiler does not support features that are essential for the library.
492 <!-- circular_buffer -->
493 <library name="circular_buffer">
494 <mark-expected-failures>
495 <test name="base_test"/>
496 <test name="space_optimized_test"/>
497 <toolset name="acc"/>
498 <note author="Boris Gubenko" refid="41"/>
499 </mark-expected-failures>
503 <!-- concept_check -->
504 <library name="concept_check">
505 <test name="class_concept_fail_expected">
507 <toolset name="cw-8.3*"/>
508 <note author="B. Dawes" refid="3"/>
511 <test name="class_concept_fail_expected">
513 <toolset name="borland-5*"/>
514 <toolset name="msvc-6.5*"/>
515 <toolset name="msvc-7.0"/>
516 <note author="Jeremy Siek"/>
519 <test name="stl_concept_covering">
522 <note author="Jeremy Siek" refid="1"/>
525 <test name="stl_concept_check">
527 <toolset name="hp_cxx*"/>
528 <note author="Markus Schoepflin" date="09 Dec 2007">
529 This version of the Rogue Wave library fails to provide all
530 needed addition operators for the iterator type and the
531 difference type of std::deque.
538 <library name="config">
539 <test name="config_link_test">
541 <toolset name="*como-4_3_3-vc7*"/>
542 <note author="J. Maddock" refid="3"/>
545 <test name="limits_test">
547 <toolset name="cw-8.3*"/>
548 <note author="B. Dawes" refid="3"/>
551 <test name="limits_test">
553 <toolset name="gcc-3_4_4_tru64"/>
554 <note author="John Maddock">
555 Long double NaN's are apparently handled incorrectly on this platform.
560 <test name="limits_test">
562 <toolset name="iw-7_1-vc6-stlp-4_5_3"/>
563 <note author="Aleksey Gurtovoy" refid="4"/>
566 <test name="limits_test">
568 <toolset name="borland-5.8*"/>
569 <note author="A.Meredith">
570 This failure is due to NaNs trapping.
574 <test name="limits_test">
576 <toolset name="borland-5.9*"/>
577 <note author="A.Meredith">
578 This failure is due to the compiler not recognising the long double special values for infinity and quiet NaN
583 <test name="test_thread_fail1">
585 <toolset name="sunpro-5_3-sunos"/>
586 <note author="J. Maddock" refid="3"/>
589 <test name="test_thread_fail2">
591 <toolset name="sunpro-5_3-sunos"/>
592 <note author="J. Maddock" refid="3"/>
599 <library name="conversion">
600 <test name="lexical_cast_test">
602 <toolset name="sunpro-5_3-sunos"/>
603 <note author="Douglas Gregor" refid="3"/>
606 <test name="lexical_cast_abstract_test">
608 <toolset name="borland-5.6*"/>
609 <toolset name="borland-5.8*"/>
610 <toolset name="borland-5.9*"/>
611 <note author="Alisdair Meredith">
612 This compiler does not support the is_abstract type trait
616 <test name="lexical_cast_loopback_test">
618 <toolset name="borland-5.6*"/>
619 <toolset name="borland-5.8*"/>
620 <toolset name="borland-5.9*"/>
621 <toolset name="gcc-3.4*"/>
622 <toolset name="gcc-4.1*"/>
623 <toolset name="gcc-4.2*"/>
624 <toolset name="gcc-mingw-3.4*"/>
625 <toolset name="sun-5.7*"/>
626 <toolset name="sun-5.8*"/>
627 <toolset name="sun-5.9*"/>
628 <toolset name="msvc-8.0*"/>
629 <toolset name="msvc-9.0*"/>
630 <toolset name="msvc-7.1*"/>
631 <toolset name="acc"/>
632 <note author="Alexander Nasonov">
633 Conversion double-string-double may give a different value (or even throw) on many compilers
637 <test name="lexical_cast_vc8_bug_test">
639 <toolset name="msvc-8.0*"/>
640 <note author="Alexander Nasonov">
641 I don't have an access to Windows and VC++ to
642 investigate this issue.
650 <test name="crc_test">
652 <toolset name="sunpro-5_3-sunos"/>
653 <note author="Douglas Gregor" refid="3"/>
659 <library name="date_time">
661 <toolset name="como-4_3_3-vc7_1"/>
662 <toolset name="sunpro-5_3-sunos"/>
663 <toolset name="msvc-6.5"/>
664 <toolset name="msvc-6.5_stlport5"/>
665 <toolset name="msvc-6.5_stlport4"/>
666 <toolset name="msvc-7.0"/>
667 <toolset name="msvc-7.0_stlport5"/>
668 <toolset name="iw-7_1-vc6-stlp-4_5_3"/>
669 <toolset name="iw-7_1-vc6"/>
670 <toolset name="dmc-*"/>
673 <test name="testgreg_serialize*">
675 <toolset name="gcc-2.*"/>
676 <toolset name="msvc-6.5*"/>
677 <note author="B. Garst">The serialization library does not support this compiler.
682 <test name="testgreg_serialize_xml">
684 <toolset name="msvc-7.0"/>
685 <note author="J. Garland">XML serialization is not supported on this compiler.
690 <test name="testtime_serialize*">
692 <toolset name="gcc-2.*"/>
693 <toolset name="msvc-6.5*"/>
694 <note author="B. Garst">The serialization library does not support this compiler.
699 <test name="testtime_serialize_xml*">
701 <toolset name="msvc-7.0"/>
702 <note author="J. Garland">XML serialization is not supported on this compiler.
707 <test name="testdate_iterator">
709 <toolset name="intel-7.1-stdlib-default-linux"/>
710 <toolset name="intel-7.1-linux"/>
711 <note author="J. Garland" refid="19,21"/>
714 <test name="testdate_iterator_dll">
716 <toolset name="intel-7.1-stdlib-default-linux"/>
717 <toolset name="intel-7.1-linux"/>
718 <note author="J. Garland" refid="19,21"/>
723 <test name="testgeneric_period">
725 <toolset name="intel-7.1-stdlib-default-linux"/>
726 <toolset name="intel-7.1-linux"/>
727 <note author="J. Garland">These are strange runtime failures for
728 which there is no obvious explanation. Later versions of the
729 Intel compiler (eg:8.0) seem to have resolved the issue.
734 <test name="testgreg_wstream">
736 <toolset name="msvc-6.5*"/>
737 <toolset name="msvc-7.0*"/>
738 <toolset name="cw-8.3*"/>
739 <toolset name="borland-5.6*"/>
740 <toolset name="borland-5.8*"/>
741 <toolset name="borland-5.9*"/>
742 <toolset name="mingw*"/>
743 <toolset name="*mingw*"/>
744 <toolset name="*cygwin*"/>
745 <toolset name="gcc-2.95.3-linux"/>
746 <toolset name="gcc-3.1-darwin"/>
747 <toolset name="*como-4_3_3*"/>
748 <note author="B. Garst" refid="19,21"/>
752 <test name="testdate_input_facet*">
754 <toolset name="cw-9.4"/>
755 <toolset name="cw-9.5*"/>
756 <note author="J. Garland">
757 For some reason Code Warrior has difficulty compiling some of the
758 input code. This may be related to limitations of locale handling,
759 but it's unclear at this time (2005-May-21).
765 <test name="testlocal_time_facet">
767 <toolset name="borland-5.6*"/>
768 <toolset name="borland-5.8*"/>
769 <toolset name="borland-5.9*"/>
770 <toolset name="*como-4_3_3*"/>
771 <toolset name="gcc-2.95.3-linux"/>
772 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
773 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
774 <toolset name="msvc-6.5"/>
775 <toolset name="msvc-7.0"/>
776 <note author="J. Garland">
777 Some older compilers are confused by the template code here.
778 These are new features to date-time in 1.33 and there is no
779 plan to backport to these non-compliant compilers.
784 <test name="testlocal_time">
786 <toolset name="msvc-6.5"/>
787 <toolset name="*como-4_3_3*"/>
788 <toolset name="borland-5.6*"/>
789 <toolset name="borland-5.8*"/>
790 <toolset name="borland-5.9*"/>
791 <toolset name="gcc-2.95.3-linux"/>
792 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
793 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
794 <note author="J. Garland">
795 Some older compilers are confused by the template code here.
796 These are new features to date-time in 1.33 and there is no
797 plan to backport to these non-compliant compilers.
802 <test name="testlocal_time_iterator">
804 <toolset name="msvc-6.5"/>
805 <toolset name="*como-4_3_3*"/>
806 <toolset name="borland-5.6*"/>
807 <toolset name="borland-5.8*"/>
808 <toolset name="borland-5.9*"/>
809 <toolset name="gcc-2.95.3-linux"/>
810 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
811 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
812 <note author="J. Garland">
813 Some older compilers are confused by the template code here.
814 These are new features to date-time in 1.33 and there is no
815 plan to backport to these non-compliant compilers.
820 <test name="testlocal_time_period">
822 <toolset name="msvc-6.5"/>
823 <toolset name="*como-4_3_3*"/>
824 <toolset name="borland-5.6*"/>
825 <toolset name="borland-5.8*"/>
826 <toolset name="borland-5.9*"/>
827 <toolset name="gcc-2.95.3-linux"/>
828 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
829 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
830 <note author="J. Garland">
831 Some older compilers are confused by the template code here.
832 These are new features to date-time in 1.33 and there is no
833 plan to backport to these non-compliant compilers.
838 <test name="testclocks">
840 <toolset name="*como-4_3_3*"/>
841 <toolset name="borland-5.6*"/>
842 <toolset name="borland-5.8*"/>
843 <toolset name="borland-5.9*"/>
844 <toolset name="gcc-2.95.3-linux"/>
845 <toolset name="msvc-7.0"/>
846 <toolset name="msvc-6.5"/>
847 <note author="J. Garland">
848 Some compilers are confused by the template code here.
849 These are new features to date-time in 1.33 and there is no
850 plan to backport to these non-compliant compilers.
855 <test name="testlocal_time_input_facet">
857 <toolset name="borland-5.6*"/>
858 <toolset name="borland-5.8*"/>
859 <toolset name="borland-5.9*"/>
860 <toolset name="*como-4_3_3*"/>
861 <toolset name="cw-8.3*"/>
862 <toolset name="gcc-2.95.3-linux"/>
863 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
864 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
865 <toolset name="msvc-6.5"/>
866 <toolset name="msvc-7.0"/>
867 <note author="J. Garland">
868 Some older compilers are confused by the template code here.
869 These are new features to date-time in 1.33 and there is no
870 plan to backport to these non-compliant compilers.
876 <test name="testtime_input_facet">
878 <toolset name="borland-5.6*"/>
879 <toolset name="borland-5.8*"/>
880 <toolset name="borland-5.9*"/>
881 <toolset name="*como-4_3_3*"/>
882 <toolset name="cw-8.3*"/>
883 <toolset name="gcc-2.95.3-linux"/>
884 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
885 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
886 <toolset name="msvc-6.5"/>
887 <toolset name="msvc-7.0"/>
888 <note author="J. Garland">
889 Some older compilers are confused by the template code here.
890 These are new features to date-time in 1.33 and there is no
891 plan to backport to these non-compliant compilers.
896 <test name="testcustom_time_zone">
898 <toolset name="borland-5.6*"/>
899 <toolset name="borland-5.8.1"/>
900 <toolset name="gcc-2.95.3-linux"/>
901 <toolset name="*como-4_3_3*"/>
902 <toolset name="msvc-6.5"/>
903 <note author="J. Garland">
904 Some older compilers are confused by the template code here.
905 These are new features to date-time in 1.33 and there is no
906 plan to backport to these non-compliant compilers.
911 <test name="testposix_time_zone">
913 <toolset name="borland-5.6*"/>
914 <toolset name="borland-5.8.1"/>
915 <toolset name="gcc-2.95.3-linux"/>
916 <toolset name="msvc-6.5"/>
917 <note author="J. Garland">
918 Some older compilers are confused by the template code here.
919 These are new features to date-time in 1.33 and there is no
920 plan to backport to these non-compliant compilers.
925 <test name="testtz_database">
927 <toolset name="borland-5.6*"/>
928 <toolset name="borland-5.8.1"/>
929 <toolset name="*como-4_3_3*"/>
930 <toolset name="gcc-2.95.3-linux"/>
931 <toolset name="msvc-6.5"/>
932 <note author="J. Garland">
933 Some compilers are confused by the template code here.
934 These are new features to date-time in 1.33 and there is no
935 plan to backport to these non-compliant compilers.
940 <test name="testtime_wstream">
942 <toolset name="gcc-2.95.3-linux"/>
943 <toolset name="gcc-3.1-darwin"/>
944 <toolset name="msvc-6.5"/>
945 <toolset name="msvc-7.0"/>
946 <toolset name="borland-5.6*"/>
947 <toolset name="borland-5.8*"/>
948 <toolset name="borland-5.9*"/>
949 <toolset name="mingw*"/>
950 <toolset name="*mingw*"/>
951 <toolset name="*cygwin*"/>
952 <toolset name="*como-4_3_3*"/>
953 <toolset name="hp_cxx-65*"/>
954 <note author="B. Garst" refid="19,21,22"/>
958 <test name="testtime_wstream_std_config">
960 <toolset name="gcc-2.95.3-linux"/>
961 <toolset name="gcc-3.1-darwin"/>
962 <toolset name="msvc-6.5"/>
963 <toolset name="msvc-7.0"/>
964 <toolset name="borland-5.6*"/>
965 <toolset name="borland-5.8*"/>
966 <toolset name="borland-5.9*"/>
967 <toolset name="mingw*"/>
968 <toolset name="*como-4_3_3*"/>
969 <toolset name="hp_cxx-65*"/>
970 <note author="B. Garst" refid="19,21,22"/>
973 <test name="testdate_facet_new">
975 <toolset name="gcc-2.95.3-linux"/>
976 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
977 <toolset name="borland-5.6*"/>
978 <toolset name="borland-5.8*"/>
979 <toolset name="borland-5.9*"/>
980 <toolset name="cw-8.3*"/>
981 <toolset name="msvc-6.5"/>
982 <toolset name="msvc-7.0"/>
983 <note author="J. Garland">
984 These compilers are unfortunately able to correctly compile the
985 new format-based input-output code for date time. Suitable, but
986 less flexible, alternatives are available on these compilers.
990 <test name="testdate_facet_new_dll">
992 <toolset name="gcc-2.95.3-linux"/>
993 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
994 <toolset name="borland-5.6*"/>
995 <toolset name="borland-5.8*"/>
996 <toolset name="borland-5.9*"/>
997 <toolset name="cw-8.3*"/>
998 <toolset name="msvc-6.5"/>
999 <toolset name="msvc-7.0"/>
1000 <note author="J. Garland">
1001 These compilers are unfortunately able to correctly compile the
1002 new format-based input-output code for date time. Suitable, but
1003 less flexible, alternatives are available on these compilers.
1007 <test name="testtime_facet">
1009 <toolset name="gcc-2.95.3-linux"/>
1010 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
1011 <toolset name="borland-5.6*"/>
1012 <toolset name="borland-5.8*"/>
1013 <toolset name="borland-5.9*"/>
1014 <toolset name="cw-8.3*"/>
1015 <toolset name="msvc-6.5"/>
1016 <toolset name="msvc-7.0"/>
1017 <note author="J. Garland">
1018 These compilers are unfortunately able to correctly compile the
1019 new format-based input-output code for date time. Suitable, but
1020 less flexible, alternatives are available on these compilers.
1025 <test name="testwcustom_time_zone">
1027 <toolset name="gcc-2.95.3-linux"/>
1028 <toolset name="gcc-3.4.2_mingw"/>
1029 <toolset name="gcc-3.4.5_mingw"/>
1030 <toolset name="*mingw*"/>
1031 <toolset name="*cygwin*"/>
1032 <note author="J. Garland">
1033 These compilers are unfortunately able to correctly compile the
1034 new format-based input-output code for date time. Suitable, but
1035 less flexible, alternatives are available on these compilers.
1040 <test name="testwposix_time_zone">
1042 <toolset name="gcc-2.95.3-linux"/>
1043 <toolset name="gcc-3.4.2_mingw"/>
1044 <toolset name="gcc-3.4.5_mingw"/>
1045 <toolset name="*mingw*"/>
1046 <toolset name="*cygwin*"/>
1047 <note author="J. Garland">
1048 These compilers are unfortunately able to correctly compile the
1049 new format-based input-output code for date time. Suitable, but
1050 less flexible, alternatives are available on these compilers.
1055 <test name="testfacet">
1057 <toolset name="gcc-2.95.3-linux"/>
1058 <toolset name="gcc-3.1-darwin"/>
1059 <toolset name="msvc-6.5"/>
1060 <toolset name="mingw*"/>
1061 <toolset name="*mingw*"/>
1062 <toolset name="*cygwin*"/>
1063 <toolset name="gcc-3.4.2_mingw"/>
1064 <toolset name="gcc-3.4.5_mingw"/>
1065 <toolset name="borland-5.6*"/>
1066 <toolset name="borland-5.8*"/>
1067 <toolset name="borland-5.9*"/>
1068 <note author="B. Garst" refid="18,19"/>
1071 <test name="testfacet_dll">
1073 <toolset name="gcc-2.95.3-linux"/>
1074 <toolset name="gcc-3.1-darwin"/>
1075 <toolset name="msvc-6.5"/>
1076 <toolset name="mingw*"/>
1077 <toolset name="*mingw*"/>
1078 <toolset name="*cygwin*"/>
1079 <toolset name="gcc-3.4.2_mingw"/>
1080 <toolset name="gcc-3.4.5_mingw"/>
1081 <toolset name="borland-5.6*"/>
1082 <toolset name="borland-5.8*"/>
1083 <toolset name="borland-5.9*"/>
1084 <toolset name="*como-4_3_3*"/>
1085 <note author="B. Garst" refid="18,19"/>
1088 <test name="testgreg_year_dll">
1090 <toolset name="*como-4_3_3*"/>
1093 <test name="testparse_date">
1095 <toolset name="gcc-2.95.3-linux"/>
1096 <toolset name="msvc-6.5"/>
1097 <toolset name="msvc-7.0"/>
1098 <toolset name="borland-5.6*"/>
1099 <toolset name="borland-5.8*"/>
1100 <toolset name="borland-5.9*"/>
1101 <note author="B. Garst" refid="18,20"/>
1104 <test name="testmicrosec_time_clock">
1106 <toolset name="intel-7.1-stdlib-default-linux"/>
1107 <toolset name="*como-4_3_3*"/>
1108 <toolset name="intel-7.1-linux"/>
1109 <note author="B. Garst" refid="22"/>
1112 <test name="testmicrosec_time_clock">
1114 <toolset name="borland-5.6.4"/>
1115 <toolset name="borland-5.8.2"/>
1116 <note author="J. Garland">
1117 There is apparently a bug in Borland library
1118 such that std::local_time and std::gmtime are
1119 returning a time that's 1 hour ahead GetSystemTimeAsFileTime
1120 during DST. This is a rather serious problem in that
1121 some of the date-time clock interfaces will give the wrong
1126 <test name="teststreams">
1128 <toolset name="gcc-2.95.3-linux"/>
1129 <toolset name="gcc-3.1-darwin"/>
1130 <toolset name="msvc-6.5"/>
1131 <toolset name="msvc-7.0"/>
1132 <toolset name="borland-5.6*"/>
1133 <toolset name="borland-5.8*"/>
1134 <toolset name="borland-5.9*"/>
1135 <toolset name="mingw-3*"/>
1136 <toolset name="gcc-3.4.2_mingw"/>
1137 <toolset name="gcc-3.4.5_mingw"/>
1138 <toolset name="*mingw*"/>
1139 <toolset name="*cygwin*"/>
1140 <toolset name="mingw"/>
1141 <toolset name="*como-4_3_3*"/>
1142 <note author="B. Garst" refid="18,19,20"/>
1145 <test name="testdate_dll">
1147 <toolset name="*como-4_3_3*"/>
1148 <note author="J. Garland" date="30 Jan 2004" id="24"/>
1151 <test name="testgreg_day_dll">
1153 <toolset name="*como-4_3_3*"/>
1154 <note author="J. Garland" date="30 Jan 2004" id="24"/>
1159 <toolset name="*como-4_3_3*"/>
1160 <note author="J. Garland" date="30 Jan 2004" id="24"/>
1164 <mark-expected-failures>
1165 <test name="testdate_dll"/>
1166 <test name="testdate_duration_dll"/>
1167 <test name="testdate_input_facet_dll"/>
1168 <test name="testdate_iterator_dll"/>
1169 <test name="testfacet_dll"/>
1170 <test name="testformatters_dll"/>
1171 <test name="testgenerators_dll"/>
1172 <test name="testgreg_durations_dll"/>
1173 <test name="testperiod_dll"/>
1174 <toolset name="cw-8.3*"/>
1175 <note author="R. Rivera" refid="25"/>
1176 </mark-expected-failures>
1178 <mark-expected-failures>
1179 <test name="testdate_facet_new"/>
1180 <test name="testdate_facet_new_dll"/>
1181 <test name="testdate_input_facet"/>
1182 <test name="testdate_input_facet_dll"/>
1183 <test name="testdate_facet"/>
1184 <test name="testdate_facet_dll"/>
1185 <test name="testtime_facet"/>
1186 <test name="testtime_input_facet"/>
1187 <toolset name="sun-5.8"/>
1188 <note author="J. Garland">
1189 The sun 5.8 compiler and standard library have a problem with
1190 the classic facet which causes some of the io tests for date-time
1191 to fail. Overall this should not affect most uses of the library.
1193 </mark-expected-failures>
1195 <mark-expected-failures>
1196 <test name="testdate_input_facet"/>
1197 <test name="testdate_input_facet_dll"/>
1198 <toolset name="msvc-7.1_stlport4"/>
1199 <note author="J. Garland">
1200 The STLPort standard library has issues with some custom
1201 facet settings causing an unexplained failure in these
1204 </mark-expected-failures>
1206 <mark-expected-failures>
1207 <test name="testdate_facet_new"/>
1208 <test name="testdate_facet_new_dll"/>
1209 <test name="testtime_facet"/>
1210 <toolset name="msvc-7.1_stlport4"/>
1211 <toolset name="msvc-8.0_stlport5"/>
1212 <note author="J. Garland">
1213 The STLPort standard library has issues with the handling
1214 of the classic facet which causes some fo the i/o tests
1215 for date-time to fail. Overall this should not affect
1216 most uses of the library.
1218 </mark-expected-failures>
1221 <mark-expected-failures>
1222 <test name="testgreg_wstream"/>
1223 <test name="testtime_facet"/>
1224 <test name="testtime_input_facet"/>
1225 <test name="testtime_wstream"/>
1226 <toolset name="msvc-7.1_stlport4"/>
1227 <note author="J. Garland">
1228 MSVC 7.1 with its standard library passes all date-time tests.
1229 For some reason when paired with stlport a few widestream
1230 io tests do not format output correctly. Overall this should
1231 not affect most uses of the library.
1233 </mark-expected-failures>
1235 <mark-expected-failures>
1236 <test name="testlocal_time_input_facet"/>
1237 <test name="testtime_input_facet"/>
1238 <toolset name="cw-9.4"/>
1239 <toolset name="cw-9.5*"/>
1240 <note author="J. Garland">
1241 Although these tests compile, the execution aborts for
1242 an unknown reason. Note that sometimes the excution is
1243 ok on cw-9_4. This may be fixable if someone
1244 can track down the source of the problem.
1246 </mark-expected-failures>
1248 <mark-expected-failures>
1249 <test name="testlocal_time"/>
1250 <test name="testlocal_time_input_facet"/>
1251 <test name="testtime_input_facet"/>
1252 <toolset name="msvc-8.0*"/>
1253 <note author="J. Garland">
1254 These tests are failing with the beta2 version of VC_8. At least
1255 one of them is directly a result of the new VC_8 standard library
1256 restricting the year value in a tm struct to be positive (that is
1257 greater than year 1900). This is a change from VC7_1 and Microsoft
1258 is considering removing this restriction.
1260 </mark-expected-failures>
1262 <mark-expected-failures>
1263 <test name="testtime_serialize*"/>
1264 <test name="testgreg_serialize*"/>
1265 <toolset name="vacpp"/>
1266 <note author="J. Garland">
1267 These tests are for serialization which has been marked as unusable.
1268 The issue was specifically noted on
1269 AIX version : 5.2.0.41 using IBM XL Version 8.0.0.0.
1271 </mark-expected-failures>
1273 <mark-expected-failures>
1274 <test name="testfacet"/>
1275 <test name="testfacet_dll"/>
1276 <toolset name="hp_cxx*"/>
1277 <toolset name="acc*"/>
1278 <note author="Markus Schoepflin">
1279 The failure is caused by a standard library bug. It doesn't
1280 support user defined facets which are not default
1281 constructible. This has been reported to the compiler vendor.
1283 </mark-expected-failures>
1285 <mark-expected-failures>
1286 <test name="testdate_input_facet_dll"/>
1287 <test name="testdate_input_facet"/>
1288 <test name="testtime_input_facet"/>
1289 <test name="testlocal_time_input_facet"/>
1290 <toolset name="acc*"/>
1291 <note author="Jeff Garland">
1292 These tests rely on the ability of an std::map to be
1293 instantiated on an incomplete type. The Rogue Wave
1294 version 2.2 and higher does not allow this.
1296 </mark-expected-failures>
1298 <mark-expected-failures>
1299 <test name="testtime_wstream"/>
1300 <toolset name="hp_cxx-65*"/>
1301 <note author="Jeff Garland">
1302 The failure is caused by a standard library bug. It doesn't
1303 support user defined facets which are not default
1304 constructible. This has been reported to the compiler vendor.
1306 </mark-expected-failures>
1308 <mark-expected-failures>
1309 <test name="testgreg_wstream"/>
1310 <test name="testparse_date"/>
1311 <test name="teststreams"/>
1312 <toolset name="hp_cxx*"/>
1313 <note author="Markus Schoepflin">
1314 The failure is caused by a standard library bug. The end-of-stream
1315 istream iterator can only be constructed when the istream iterator
1316 has been instantiated with char as the character type. This has
1317 been reported to the compiler vendor.
1319 </mark-expected-failures>
1321 <mark-expected-failures>
1322 <test name="testfacet"/>
1323 <test name="testfacet_dll"/>
1324 <test name="testgreg_wstream"/>
1325 <test name="teststreams"/>
1326 <test name="testtime_wstream"/>
1327 <test name="testwcustom_time_zone"/>
1328 <test name="testwposix_time_zone"/>
1329 <toolset name="qcc-3.3.5_gpp"/>
1330 <note author="Jim Douglas" date="12 Feb 06" refid="36"/>
1331 </mark-expected-failures>
1336 <!-- dynamic_bitset -->
1337 <library name="dynamic_bitset">
1338 <test name="dyn_bitset_unit_tests1">
1340 <toolset name="msvc-6.5_stlport4"/>
1341 <note author="Gennaro Prota" refid="37" />
1344 <test name="dyn_bitset_unit_tests2">
1346 <toolset name="borland-5.8*"/>
1347 <toolset name="borland-5.9*"/>
1348 <note author="Roland Schwarz">
1349 The exact reason of this (linker related) bug is unresearched. The test passes
1350 on some environments. The test was found to fail on a platform whit a german
1351 version of the compiler.
1355 <test name="dyn_bitset_unit_tests4">
1357 <toolset name="cw-9.3"/>
1358 <note author="Aleksey Gurtovoy" refid="2"/>
1361 <toolset name="cw-9.3-darwin"/>
1362 <note author="Douglas Gregor" refid="2"/>
1365 <toolset name="sunpro-5_3-sunos"/>
1366 <note author="Douglas Gregor" refid="2"/>
1373 <library name="filesystem">
1375 <toolset name="borland-5.6*"/>
1376 <toolset name="borland-5.8*"/>
1377 <note author="Beman Dawes">
1378 This compiler does not support enable_if, which is needed by the
1379 Boost.System library on which Boost.Filesystem depends.
1383 <toolset name="intel-7.1-linux"/>
1384 <toolset name="intel-7.1-stdlib-default-linux"/>
1385 <note author="Aleksey Gurtovoy">
1386 Due to standard library bugs this configuration is not supported by
1387 the most recent version of the library.
1391 <toolset name="cw-8.3"/>
1392 <note author="Beman Dawes">
1393 Due to standard library bugs, this version is not supported.
1394 More recent version of the library should work OK.
1398 <toolset name="msvc-8.0~wm5~stlport5.1"/>
1399 <note author="Beman Dawes">
1400 Due to lack of C library featues, this toolset is not supported.
1403 <mark-expected-failures>
1405 <toolset name="sun-5.7"/>
1406 <toolset name="sun-5.8"/>
1407 <note author="Beman Dawes">
1408 The library works well with versions of this compiler 5.9 and later
1410 </mark-expected-failures>
1411 <mark-expected-failures>
1412 <test name="fstream_test"/>
1413 <toolset name="msvc-6.5*"/>
1414 <note author="Beman Dawes">
1415 fstream for this compiler has serious problems and is not supported
1417 </mark-expected-failures>
1418 <mark-expected-failures>
1419 <test name="operations_test_dll"/>
1420 <test name="path_test_dll"/>
1421 <toolset name="borland-5.6*"/>
1422 <toolset name="borland-5.8*"/>
1423 <toolset name="borland-5.9*"/>
1424 <toolset name="gcc-3.4.2_mingw"/>
1425 <toolset name="gcc-3.4.2_mingw"/>
1426 <note author="Beman Dawes" refid="35"/> <!-- dll's don't work - use static -->
1427 </mark-expected-failures>
1428 <mark-expected-failures>
1429 <test name="operations_test"/>
1430 <test name="operations_test_dll"/>
1431 <toolset name="msvc-6.5*"/>
1432 <note author="Beman Dawes" refid="31"/> <!-- esoteric features don't work -->
1433 </mark-expected-failures>
1434 <mark-expected-failures>
1435 <test name="mbcopy"/>
1436 <test name="mbpath"/>
1437 <test name="wide_test"/>
1438 <toolset name="gcc-3.4.2_mingw"/>
1439 <toolset name="gcc-3.4.5_mingw"/>
1440 <toolset name="gcc-mingw-3.4.5"/>
1441 <toolset name="gcc-mingw-3.4.2"/>
1442 <toolset name="gcc-cygwin-3.4.4"/>
1443 <note author="Beman Dawes" refid="19"/> <!-- no wchar_t, wstring support -->
1444 </mark-expected-failures>
1445 <mark-expected-failures>
1446 <test name="mbcopy"/>
1447 <test name="mbpath"/>
1448 <test name="wide_test"/>
1449 <toolset name="msvc-6.5*"/>
1450 <toolset name="msvc-7.0"/>
1451 <toolset name="borland-5.6*"/>
1452 <toolset name="borland-5.8*"/>
1453 <toolset name="borland-5.9*"/>
1454 <toolset name="cw-8.3"/>
1455 <toolset name="dmc-8_4_7*"/>
1456 <note author="Beman Dawes">
1457 The library does not support wide paths on this compiler because
1458 it does not support SFINAE.
1460 </mark-expected-failures>
1461 <mark-expected-failures>
1462 <test name="mbcopy"/>
1463 <test name="mbpath"/>
1464 <test name="wide_test"/>
1465 <toolset name="qcc-3.3.5_gpp"/>
1466 <note author="Jim Douglas" refid="36"/>
1467 </mark-expected-failures>
1468 <mark-expected-failures>
1469 <test name="mbcopy"/>
1470 <test name="wide_test"/>
1471 <toolset name="sun-5.8"/>
1472 <note author="John Maddock">
1473 These failures are reported to be fixed in Sun's
1474 next compiler release.
1476 </mark-expected-failures>
1480 <library name="flyweight">
1481 <mark-expected-failures>
1482 <test name="test_intermod_holder"/>
1483 <toolset name="borland-5.*"/>
1484 <toolset name="borland-6.10.0"/>
1485 <toolset name="sun-5.*"/>
1486 <toolset name="msvc-6.5*"/>
1487 <toolset name="cw-9.*"/>
1488 <toolset name="gcc-2.95*"/>
1489 <toolset name="gcc-3.0*"/>
1490 <toolset name="gcc-3.1*"/>
1491 <toolset name="gcc-3.2*"/>
1492 <toolset name="gcc-3.3*"/>
1493 <toolset name="gcc-4.2.1_hpux_ia64"/>
1494 <toolset name="mipspro"/>
1495 <toolset name="acc*"/>
1496 <toolset name="msvc-8.0~wm5*"/>
1497 <toolset name="vacpp*"/>
1498 <toolset name="pathscale*"/>
1499 <toolset name="intel-linux-8.*"/>
1500 <toolset name="gcc-3.4.6_linux_ia64"/>
1501 <note author="J. López" date="03 Dec 2008">
1502 This compiler does not support Boost.Interprocess,
1503 on which intermodule_holder depends.
1505 </mark-expected-failures>
1509 <library name="foreach">
1511 <toolset name="dmc*"/>
1512 <toolset name="sunpro-5_3-sunos"/>
1513 <note author="Eric Niebler">
1514 This compiler does not support the Boost.Range
1515 library, on which Boost.Foreach depends.
1519 <mark-expected-failures>
1520 <test name="rvalue_const"/>
1521 <toolset name="msvc-6.5*"/>
1522 <toolset name="msvc-7.0"/>
1523 <toolset name="borland-5.6*"/>
1524 <toolset name="borland-5.8*"/>
1525 <toolset name="borland-5.9*"/>
1526 <toolset name="borland-6.0*"/>
1527 <toolset name="borland-6.1*"/>
1528 <toolset name="gcc-2*"/>
1529 <toolset name="gcc-3.2*"/>
1530 <toolset name="gcc-3_3-darwin"/>
1531 <toolset name="intel-linux"/>
1532 <toolset name="vacpp"/>
1533 <toolset name="cw-8.3"/>
1534 <toolset name="cw-9.4"/>
1535 <toolset name="cw-9.5-darwin"/>
1536 <toolset name="sunpro*"/>
1537 <toolset name="mingw"/>
1538 <toolset name="hp_cxx*"/>
1539 <toolset name="intel-win32-8_1"/>
1540 <toolset name="sun-5.7"/>
1541 <toolset name="sun-5.8"/>
1542 <toolset name="sun-5.9"/>
1543 <toolset name="pathscale*"/>
1544 <note author="Eric Niebler">
1545 This compiler does not support detection of
1548 </mark-expected-failures>
1549 <mark-expected-failures>
1550 <test name="rvalue_const_r"/>
1551 <toolset name="msvc-6.5*"/>
1552 <toolset name="msvc-7.0"/>
1553 <toolset name="borland-5.6*"/>
1554 <toolset name="borland-5.8*"/>
1555 <toolset name="borland-5.9*"/>
1556 <toolset name="borland-6.0*"/>
1557 <toolset name="borland-6.1*"/>
1558 <toolset name="gcc-2*"/>
1559 <toolset name="gcc-3.2*"/>
1560 <toolset name="gcc-3_3-darwin"/>
1561 <toolset name="intel-linux"/>
1562 <toolset name="vacpp"/>
1563 <toolset name="cw-8.3"/>
1564 <toolset name="cw-9.4"/>
1565 <toolset name="cw-9.5-darwin"/>
1566 <toolset name="sunpro*"/>
1567 <toolset name="mingw"/>
1568 <toolset name="hp_cxx*"/>
1569 <toolset name="intel-win32-8_1"/>
1570 <toolset name="sun-5.7"/>
1571 <toolset name="sun-5.8"/>
1572 <toolset name="sun-5.9"/>
1573 <toolset name="pathscale*"/>
1574 <note author="Eric Niebler">
1575 This compiler does not support detection of
1578 </mark-expected-failures>
1579 <mark-expected-failures>
1580 <test name="rvalue_nonconst"/>
1581 <toolset name="msvc-6.5*"/>
1582 <toolset name="msvc-7.0"/>
1583 <toolset name="borland-5.6*"/>
1584 <toolset name="borland-5.8*"/>
1585 <toolset name="borland-5.9*"/>
1586 <toolset name="borland-6.0*"/>
1587 <toolset name="borland-6.1*"/>
1588 <toolset name="hp_cxx*"/>
1589 <toolset name="sunpro*"/>
1590 <toolset name="sun-5.7"/>
1591 <toolset name="sun-5.8"/>
1592 <toolset name="sun-5.9"/>
1593 <note author="Eric Niebler">
1594 This compiler does not support detection of
1597 </mark-expected-failures>
1598 <mark-expected-failures>
1599 <test name="rvalue_nonconst_r"/>
1600 <toolset name="msvc-6.5*"/>
1601 <toolset name="msvc-7.0"/>
1602 <toolset name="borland-5.6*"/>
1603 <toolset name="borland-5.8*"/>
1604 <toolset name="borland-5.9*"/>
1605 <toolset name="borland-6.0*"/>
1606 <toolset name="borland-6.1*"/>
1607 <toolset name="hp_cxx*"/>
1608 <toolset name="sunpro*"/>
1609 <toolset name="sun-5.7"/>
1610 <toolset name="sun-5.8"/>
1611 <toolset name="sun-5.9"/>
1612 <note author="Eric Niebler">
1613 This compiler does not support detection of
1616 </mark-expected-failures>
1617 <mark-expected-failures>
1618 <test name="dependent_type"/>
1619 <toolset name="gcc-2.95.3*"/>
1620 <toolset name="hp_cxx-65*"/>
1621 <note author="Eric Niebler">
1622 These compilers cannot handle BOOST_FOREACH
1623 in a template, where the collection type
1624 depends on a template parameter.
1626 </mark-expected-failures>
1627 <mark-expected-failures>
1628 <test name="user_defined"/>
1629 <toolset name="msvc-6.5*"/>
1630 <toolset name="msvc-7.0*"/>
1631 <note author="Eric Niebler">
1632 This failure is because the Boost.Range extension
1633 mechanism is broken on these compilers. It requires
1634 ADL which these compilers do not support.
1636 </mark-expected-failures>
1640 <library name="format">
1642 <toolset name="iw-7_1*"/>
1643 <note author="Aleksey Gurtovoy">
1644 The failure is caused by a standard library bug: the
1645 iostream components fail to handle <code>ios::internal</code>
1650 <toolset name="sunpro-5_3-sunos"/>
1652 <mark-expected-failures>
1653 <test name="format_test2"/>
1654 <test name="format_test3"/>
1655 <toolset name="hp_cxx-65*"/>
1656 <toolset name="acc*"/>
1657 <note author="Markus Schoepflin" refid="33"/>
1658 </mark-expected-failures>
1661 <!-- function_types -->
1662 <library name="function_types">
1663 <mark-expected-failures>
1664 <test name="member_ccs"/>
1665 <test name="member_ccs_exact"/>
1667 <note author="Tobias Schwinger">
1668 Not all compilers/platforms implement nonstandard calling conventions.
1670 With GCC/x86 this failure reflects
1671 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29328 .
1673 </mark-expected-failures>
1674 <mark-expected-failures>
1675 <test name="nonmember_ccs"/>
1676 <test name="nonmember_ccs_exact"/>
1678 <note author="Tobias Schwinger">
1679 Not all compilers/platforms implement nonstandard calling conventions.
1681 </mark-expected-failures>
1682 <mark-expected-failures>
1683 <test name="interface_example"/>
1684 <toolset name="msvc-7.1*"/>
1685 <note author="Tobias Schwinger">
1686 Overload selection does not work in some assignment contexts with this compiler.
1688 </mark-expected-failures>
1691 <!-- functional/factory -->
1692 <library name="functional/factory">
1693 <mark-expected-failures>
1694 <test name="factory_with_allocator"/>
1695 <toolset name="borland-*"/>
1696 <note author="Tobias Schwinger">
1697 Probably broken const conversion with that compiler.
1699 </mark-expected-failures>
1702 <!-- functional/forward -->
1703 <library name="functional/foward">
1705 <toolset name="msvc-7.0*"/>
1706 <toolset name="msvc-7.1*"/>
1707 <toolset name="sun-5.*"/>
1708 <toolset name="vacpp*"/>
1709 <toolset name="borland-*"/>
1710 <note author="Tobias Schwinger">
1711 This compiler is currently not supported.
1716 <!-- functional/hash -->
1717 <library name="functional/hash">
1718 <mark-expected-failures>
1719 <test name="hash_value_array_test"/>
1720 <toolset name="msvc-6.5*"/>
1721 <toolset name="msvc-7.0*"/>
1722 <note author="Daniel James">
1723 hash_value is not overloaded for arrays for older versions
1724 of Visual C++. There is a work around so that
1725 boost::hash<T[N]>, boost::hash_combine and boost::hash_range
1728 </mark-expected-failures>
1730 <mark-expected-failures>
1731 <test name="hash_function_pointer_test"/>
1732 <toolset name="msvc-6.5*"/>
1733 <toolset name="msvc-7.0*"/>
1734 <note refid="2" author="Daniel James"/>
1735 </mark-expected-failures>
1737 <mark-expected-failures>
1738 <test name="hash_function_pointer_test"/>
1739 <toolset name="sun-5.7"/>
1740 <toolset name="sun-5.8"/>
1741 <toolset name="sun-5.9"/>
1742 <note author="Daniel James">
1743 On these compilers the wrong overload of hash_value is called
1744 when the argument is a hash function pointer. So calling
1745 hash_value doesn't work but boost::hash does work (and it's
1746 recommended that user never call hash_value directly so this
1747 shouldn't be a problem).
1749 </mark-expected-failures>
1751 <mark-expected-failures>
1752 <test name="hash_long_double_test"/>
1753 <toolset name="gcc-3.4.3_sunos"/>
1754 <toolset name="*pa_risc"/>
1755 <note author="Daniel James">
1756 This platform has poor support for <code>long double</code> so
1757 the hash function perform poorly for values out of the range
1758 of <code>double</code> or if they differ at a greater precision
1759 that <code>double</code> is capable of representing.
1761 </mark-expected-failures>
1763 <mark-expected-failures>
1764 <test name="point" />
1765 <test name="books" />
1766 <toolset name="msvc-6.5*"/>
1767 <toolset name="msvc-7.0*"/>
1768 <note author="Daniel James">
1769 These examples only work on compilers with support for ADL.
1770 It is possible to work around this, but I wanted to keep the
1771 example code as clean as possible.
1773 </mark-expected-failures>
1775 <mark-expected-failures>
1776 <test name="point" />
1777 <toolset name="borland-*"/>
1778 <note author="Daniel James">
1779 It appears that Borland doesn't find friend functions defined
1780 in a class by ADL. This is easily fixed but this example is
1781 meant to show the typical way of customising boost::hash, not
1784 </mark-expected-failures>
1786 <mark-expected-failures>
1787 <test name="hash_global_namespace_test" />
1788 <toolset name="borland-*"/>
1789 <note author="Daniel James">
1790 The test demonstrates a Borland bug - functions that aren't
1791 in a namespace don't appear to be found by ADL.
1793 </mark-expected-failures>
1795 <mark-expected-failures>
1796 <test name="container_fwd_gcc_debug"/>
1797 <toolset name="darwin-4.2"/>
1798 <note author="Daniel James">
1799 Debug containers aren't supported on Apple's version of gcc 4.2.
1801 </mark-expected-failures>
1807 <library name="fusion">
1809 <toolset name="gcc-2.95.3*"/>
1810 <toolset name="msvc-6.5*"/>
1811 <toolset name="msvc-7.0"/>
1812 <toolset name="borland-5*"/>
1813 <toolset name="cw-8.3"/>
1814 <toolset name="dmc*"/>
1815 <toolset name="sunpro-5_3-sunos"/>
1816 <toolset name="sun-5.7"/>
1817 <toolset name="sun-5.8"/>
1818 <toolset name="sun-5.9"/>
1819 <note author="Joel de Guzman">
1820 The compiler does not support features that are
1821 essential for the library.
1824 <mark-expected-failures>
1825 <test name="fused"/>
1826 <test name="fused_function_object"/>
1827 <test name="fused_procedure"/>
1828 <test name="make_fused"/>
1829 <test name="make_fused_function_object"/>
1830 <test name="make_fused_procedure"/>
1831 <toolset name="acc"/>
1832 <note author="Tobias Schwinger">
1833 Those failures are due to not quite const-correct overload
1834 resolution. The complaints from the test suite should rarely
1835 matter in practice - the corresponding components are basically
1836 usable. With aCC6, when compiled in strict ansi mode, the test
1839 </mark-expected-failures>
1843 <library name="gil">
1844 <mark-expected-failures>
1845 <test name="pixel"/>
1846 <toolset name="acc"/>
1847 <note author="Boris Gubenko" refid="46"/>
1848 </mark-expected-failures>
1849 <mark-expected-failures>
1850 <test name="image"/>
1851 <toolset name="acc"/>
1852 <note author="Boris Gubenko" refid="47"/>
1853 </mark-expected-failures>
1857 <library name="graph">
1859 <toolset name="borland-5.*"/>
1860 <toolset name="borland-6.*"/>
1861 <toolset name="borland-cb2009"/>
1862 <toolset name="borland-cb2010"/>
1863 <toolset name="sun-5.9"/>
1865 <mark-expected-failures>
1866 <test name="grid_graph_cc"/>
1867 <test name="grid_graph_test"/>
1868 <toolset name="msvc-7.1"/>
1869 <note author="Jeremiah Willcock" refid="2"/>
1870 </mark-expected-failures>
1874 <library name="integer">
1875 <mark-expected-failures>
1876 <test name="integer_test"/>
1877 <toolset name="acc"/>
1878 <toolset name="gcc-4.2.1_hpux_ia64"/>
1879 <note author="Boris Gubenko">
1880 When compiling with aC++, depending on system load, the compile time may exceed
1881 specified timeout value. The test passes when the timeout value is increased.
1882 When compiling with GCC, linker takes segmentation fault.
1883 In the HP bug tracking system, this issue is tracked as QuIX ID: QXCR1000836120.
1885 </mark-expected-failures>
1888 <!-- interprocess-->
1889 <library name="interprocess">
1891 <toolset name="borland-5.*"/>
1892 <toolset name="sun-5.*"/>
1893 <toolset name="msvc-6.5*"/>
1894 <toolset name="cw-9.*"/>
1895 <toolset name="gcc-2.95*"/>
1896 <toolset name="gcc-3.0*"/>
1897 <toolset name="gcc-3.1*"/>
1898 <toolset name="gcc-3.2*"/>
1899 <toolset name="gcc-3.3*"/>
1900 <toolset name="gcc-4.2.1_hpux_ia64"/>
1901 <toolset name="mipspro"/>
1902 <toolset name="acc*"/>
1903 <toolset name="msvc-8.0~wm5*"/>
1904 <toolset name="pathscale*"/>
1905 <toolset name="intel-linux-8.*"/>
1906 <toolset name="gcc-3.4.6_linux_ia64"/>
1907 <note author="Ion Gaztañaga">
1908 The compiler does not support features that are essential for the library.
1914 <library name="intrusive">
1916 <toolset name="borland-5.*"/>
1917 <toolset name="msvc-6.5*"/>
1918 <toolset name="cw-9.*"/>
1919 <toolset name="gcc-2.95*"/>
1920 <toolset name="gcc-3.0*"/>
1921 <toolset name="gcc-3.1*"/>
1922 <toolset name="gcc-3.2*"/>
1923 <toolset name="gcc-3.3*"/>
1924 <toolset name="mipspro"/>
1925 <toolset name="pathscale*"/>
1926 <toolset name="intel-linux-8.*"/>
1927 <note author="Ion Gaztañaga">
1928 The compiler does not support features that are essential for the library.
1931 <mark-expected-failures>
1932 <test name="doc_offset_ptr" />
1933 <toolset name="acc"/>
1934 <toolset name="gcc-3.4.2_hpux_pa_risc"/>
1935 <toolset name="gcc-3.4.6_linux_ia64"/>
1936 <note author="Ion Gaztañaga">
1937 The compiler is not supported by Interprocess.
1939 </mark-expected-failures>
1940 <mark-expected-failures>
1941 <test name="unordered_multiset_test"/>
1942 <test name="unordered_set_test"/>
1943 <toolset name="acc"/>
1944 <note author="Boris Gubenko" refid="47"/>
1945 </mark-expected-failures>
1950 <mark-expected-failures>
1951 <test name="ios_state_unit_test"/>
1952 <toolset name="borland-5.6*"/>
1953 <toolset name="borland-5.8*"/>
1954 <toolset name="borland-5.9*"/>
1955 <toolset name="iw-7_1-vc6*"/>
1956 <toolset name="msvc-6.5*"/>
1957 <note refid="4" author="Aleksey Gurtovoy"/>
1958 </mark-expected-failures>
1959 <mark-expected-failures>
1960 <test name="ios_state_test"/>
1961 <test name="ios_state_unit_test"/>
1962 <toolset name="hp_cxx-65*"/>
1963 <note refid="34" author="Markus Schoepflin"/>
1964 </mark-expected-failures>
1965 <mark-expected-failures>
1966 <test name="ios_state_unit_test"/>
1967 <toolset name="gcc-2.95.3-*"/>
1968 <note refid="3" author="Doug Gregor"/>
1969 </mark-expected-failures>
1970 <mark-expected-failures>
1971 <test name="ios_state_unit_test"/>
1972 <toolset name="gcc-4.1.0*"/>
1973 <note author="John Maddock">
1974 This is gcc bug 26526, and is fixed in later releases.
1976 </mark-expected-failures>
1980 <library name="iostreams">
1981 <mark-expected-failures>
1982 <test name="auto_close_test"/>
1983 <test name="component_access_test"/>
1984 <test name="compose_test"/>
1985 <test name="counter_test"/>
1986 <test name="filtering_stream_test"/>
1987 <test name="flush_test"/>
1988 <test name="line_filter_test"/>
1989 <test name="newline_test"/>
1990 <test name="pipeline_test"/>
1991 <test name="regex_filter_test"/>
1992 <test name="restrict_test"/>
1993 <test name="seekable_file_test"/>
1994 <test name="seekable_filter_test"/>
1995 <test name="sequence_test"/>
1996 <test name="slice_test"/>
1997 <test name="stdio_filter_test"/>
1998 <test name="tee_test"/>
1999 <test name="wide_stream_test"/>
2000 <toolset name="sun-5.7"/>
2001 <toolset name="sun-5.8"/>
2002 <note author="Jonathan Turkanis" date="09 Jan 2008" refid="2"/>
2003 </mark-expected-failures>
2004 <mark-expected-failures>
2005 <test name="bzip2_test"/>
2006 <toolset name="gcc-3.4.3_sunos"/>
2007 <note author="Caleb Epstein">
2008 No bzip2 support on the testing machine and no way to
2009 disable this test with BBv2 at present.
2011 </mark-expected-failures>
2012 <mark-expected-failures>
2013 <test name="bzip2_test"/>
2014 <test name="file_descriptor_test"/>
2015 <test name="mapped_file_test"/>
2016 <toolset name="*como-4_3_3*"/>
2017 <note author="Jonathan Turkanis">
2018 compiler can't compile "windows.h" in strict mode
2020 </mark-expected-failures>
2021 <mark-expected-failures>
2022 <test name="code_converter_test"/>
2023 <toolset name="pgi-7.0"/>
2024 <note author="Jonathan Turkanis">
2025 This platform lacks the placement delete operator
2026 required by the C++ standard
2028 </mark-expected-failures>
2029 <mark-expected-failures>
2030 <!-- Insufficient wide character support -->
2031 <test name="code_converter_test"/>
2032 <test name="wide_stream_test"/>
2033 <toolset name="gcc-2.95.3-linux"/>
2034 <!-- Must enumerate MinGW's since some use STLPort -->
2035 <toolset name="gcc-3.4.2_mingw"/>
2036 <toolset name="mingw-3_4_4"/>
2037 <toolset name="gcc-3.4.5_mingw"/>
2038 <toolset name="gcc-3.4.5_mingw"/>
2039 <toolset name="*cygwin*"/>
2040 <toolset name="gcc-3.3.6-osf1"/>
2041 <toolset name="gcc-3.4.2_hpux_pa_risc"/>
2042 <note author="Jonathan Turkanis" refid="19"/>
2043 </mark-expected-failures>
2044 <mark-expected-failures>
2045 <!-- Insufficient wide character support -->
2046 <test name="code_converter_test"/>
2047 <test name="wide_stream_test"/>
2048 <toolset name="qcc-3.3.5*gpp"/>
2049 <note author="Jim Douglas" date="12 Feb 06" refid="36"/>
2050 </mark-expected-failures>
2051 <mark-expected-failures>
2052 <test name="compose_test"/>
2053 <toolset name="msvc-6.5_stlport4"/>
2054 <note author="Jonathan Turkanis">
2055 These six tests pass individually but cause a compiler stack overflow
2056 when compiled as a group
2058 </mark-expected-failures>
2059 <mark-expected-failures>
2060 <test name="compose_test"/>
2061 <toolset name="gcc-3.4.6_linux_ia64"/>
2062 <note author="Boris Gubenko">
2063 On this platform, linking this test takes longer than 10 minutes
2064 which is a time limit specified for bjam. When linked manually,
2067 </mark-expected-failures>
2068 <mark-expected-failures reason="?">
2069 <test name="direct_adapter_test"/>
2070 <test name="gzip_test"/>
2071 <toolset name="gcc-2.95.3-linux"/>
2072 <note author="Jonathan Turkanis" refid="29"/>
2073 </mark-expected-failures>
2074 <mark-expected-failures>
2075 <test name="file_descriptor_test"/>
2076 <toolset name="gcc-cygwin-3.4.4"/>
2077 <note author="Vladimir Prus">
2078 The test fails at runtime for unknown reasons.
2080 </mark-expected-failures>
2081 <mark-expected-failures reason="?">
2082 <test name="file_descriptor_test"/>
2083 <toolset name="gcc-3_4_4-cygwin"/>
2084 <note author="Jonathan Turkanis" refid="29"/>
2085 </mark-expected-failures>
2086 <mark-expected-failures>
2087 <test name="finite_state_filter_test"/>
2088 <toolset name="borland-5.6*"/>
2089 <toolset name="borland-5.8*"/>
2090 <toolset name="borland-5.9*"/>
2091 <toolset name="msvc-6.5*"/>
2092 <toolset name="msvc-7.0"/>
2093 <toolset name="gcc-2.95.3*"/>
2094 <toolset name="sun-5.*"/>
2095 <toolset name="vacpp"/>
2096 <note author="Jonathan Turkanis" refid="2"/>
2097 </mark-expected-failures>
2098 <mark-expected-failures>
2099 <test name="finite_state_filter_test"/>
2100 <toolset name="cw-9.4"/>
2101 <note author="Jonathan Turkanis" date="20 Dec 06">
2102 I'm not sure whether CodeWarrior is correct to report that the member
2103 in question is inaccessible; however, when the member is made public
2104 an internal error occur that I have not been able to fix, so for
2105 now the question is moot.
2107 </mark-expected-failures>
2108 <mark-expected-failures>
2109 <test name="gzip_test"/>
2110 <test name="zlib_test"/>
2111 <toolset name="como-4_3_3-vc7_1"/>
2112 <note author="Jonathan Turkanis">
2113 The failure reflects a problem with the build system: the zlib
2114 object files are generated in the wrong directory.
2116 </mark-expected-failures>
2117 <mark-expected-failures>
2118 <test name="mapped_file_test"/>
2119 <toolset name="qcc-3.3*"/>
2120 <note author="Jim Douglas" date="19 Feb 06">
2121 Memory mapped files are not supported in QNX Neutrino version 6.3.0.
2123 </mark-expected-failures>
2124 <mark-expected-failures>
2125 <test name="restrict_test"/>
2126 <toolset name="vacpp"/>
2127 <note author="Jonathan Turkanis" date="06 Jan 2008">
2128 "restrict" is treated as a keyword on this platform (as in C99);
2129 use the alias "slice" instead, defined in
2130 "boost/iostreams/slice.hpp."
2132 </mark-expected-failures>
2133 <mark-expected-failures>
2134 <!-- STLPort bug -->
2135 <test name="seekable_file_test"/>
2136 <toolset name="borland-5.6*"/>
2137 <toolset name="iw-7_1-vc6-stlp-4_5_3"/>
2138 <toolset name="*como-4_3_3*"/>
2139 <toolset name="sun-5.*"/>
2140 <toolset name="*stlport"/>
2141 <toolset name="pgi-7.0"/>
2142 <note author="Jonathan Turkanis" refid="4"/>
2143 </mark-expected-failures>
2144 <mark-expected-failures>
2145 <test name="stdio_filter_test"/>
2146 <toolset name="*como-4_3_3*"/>
2147 <note author="Jonathan Turkanis" refid="0"/>
2148 </mark-expected-failures>
2149 <mark-expected-failures>
2150 <test name="stream_offset_64bit_test"/>
2151 <toolset name="borland-*"/>
2152 <note author="Jonathan Turkanis" date="04 Jan 2008">
2153 In the Dinkumware standard library, streampos relies on fpos_t
2154 to store stream offsets, but fpos_t is defined as a 32-bit
2155 long by the Borland runtime library. In Borland's modified
2156 version of STLPort, streampos relies on streamoff to store
2157 stream offsets, but streamoff is defined to be a 32-bit long.
2159 </mark-expected-failures>
2160 <mark-expected-failures>
2161 <test name="stream_offset_64bit_test"/>
2162 <toolset name="sun-5.*"/>
2163 <note author="Jonathan Turkanis" date="06 Jan 2008">
2164 In STLPort, streampos consists of a long together with a
2165 conversion state; on this platform, long is a 32-bit type
2167 </mark-expected-failures>
2168 <mark-expected-failures>
2169 <test name="stream_offset_64bit_test"/>
2170 <toolset name="vacpp*"/>
2171 <note author="Jonathan Turkanis" date="09 Jan 2008">
2172 On this platform, streampos is an alias for fpos, whose
2173 implementation stores stream offsets using streamsize and
2174 fpos_t; both of the latter types are 32-bit
2176 </mark-expected-failures>
2177 <mark-expected-failures>
2178 <test name="stream_offset_64bit_test"/>
2179 <toolset name="intel-win-10.0_stdcxx_421"/>
2180 <toolset name="msvc-7.1_stdcxx_421"/>
2181 <toolset name="msvc-9.0_stdcxx_421"/>
2182 <toolset name="intel-win-10.1_stdcxx_421"/>
2183 <toolset name="intel-linux-10.1_stdcxx_421"/>
2184 <toolset name="gcc-4.2.1_stdcxx_421"/>
2185 <note author="Jonathan Turkanis" date="09 Jan 2008">
2186 On this platform, streampos is an alias for ptrdiff_t, which
2187 is an alias for a 32-bit type
2189 </mark-expected-failures>
2190 <mark-expected-failures>
2191 <test name="stream_offset_64bit_test"/>
2192 <toolset name="gcc-4.2"/>
2193 <note author="Jonathan Turkanis" date="09 Jan 2008">
2194 The following applies only to gcc-4.2 using the stdcxx
2195 standard library: On this platform, streampos is an alias for
2196 ptrdiff_t, which is an alias for a 32-bit type
2198 </mark-expected-failures>
2202 <library name="lambda">
2204 <toolset name="msvc-6.5*"/>
2205 <toolset name="borland-5.5*"/>
2206 <toolset name="borland-5.6*"/>
2207 <toolset name="borland-5.8*"/>
2208 <toolset name="borland-5.9*"/>
2209 <toolset name="msvc-7.0"/>
2210 <toolset name="sunpro-5_3-sunos"/>
2214 <mark-expected-failures>
2215 <test name="bll_and_function"/>
2216 <toolset name="msvc-8.0"/>
2217 <note author="Aleksey Gurtovoy" refid="6"/>
2218 </mark-expected-failures>
2219 <mark-expected-failures>
2220 <test name="member_pointer_test"/>
2221 <toolset name="gcc-2.95.3-*"/>
2222 <note author="Doug Gregor" refid="3"/>
2223 </mark-expected-failures>
2224 <mark-expected-failures>
2225 <test name="control_structures"/>
2226 <toolset name="gcc-4.2.1*"/>
2227 <note author="Boris Gubenko" refid="42"/>
2228 </mark-expected-failures>
2232 <library name="logic">
2233 <test name="tribool_io_test">
2235 <toolset name="msvc-6.5_stlport4"/>
2236 <toolset name="gcc-2.95.3-linux"/>
2237 <toolset name="sunpro-5_3-sunos"/>
2238 <toolset name="hp_cxx-65*"/>
2239 <note author="Douglas Gregor" refid="4"/>
2246 <library name="mpl">
2249 <toolset name="sunpro-5_3-sunos"/>
2250 <note author="Aleksey Gurtovoy" date="10 Jul 2005">
2251 The compiler is not supported by the library due to an
2252 utterly broken templates support.
2256 <mark-expected-failures>
2257 <test name="as_sequence"/>
2258 <test name="is_sequence"/>
2259 <test name="has_xxx"/>
2260 <test name="no_has_xxx"/>
2261 <test name="single_view"/>
2262 <toolset name="cw-8.3*"/>
2263 <note author="Aleksey Gurtovoy" date="17 Sep 2004">
2264 This failure is caused by a deficient SFINAE implementation; the bug
2265 was fixed in the next major compiler version (CodeWarrior 9.x).
2267 </mark-expected-failures>
2269 <mark-expected-failures>
2270 <test name="is_sequence"/>
2271 <test name="as_sequence"/>
2272 <test name="has_xxx"/>
2273 <toolset name="borland-5.6*"/>
2274 <toolset name="borland-5.8*"/>
2275 <toolset name="borland-5.9*"/>
2276 <toolset name="gcc-2.95.3*"/>
2277 <note author="Aleksey Gurtovoy" date="17 Sep 2004">
2278 This failure is caused by a deficient SFINAE implementation.
2280 </mark-expected-failures>
2282 <mark-expected-failures>
2283 <test name="arithmetic"/>
2286 <test name="bitwise"/>
2287 <test name="contains"/>
2289 <test name="count"/>
2290 <test name="count_if"/>
2291 <test name="deque"/>
2292 <test name="distance"/>
2293 <test name="find_if"/>
2294 <test name="for_each"/>
2295 <test name="front"/>
2296 <test name="insert"/>
2297 <test name="insert_range"/>
2298 <test name="joint_view"/>
2299 <test name="numeric_ops"/>
2300 <test name="pair_view"/>
2301 <test name="partition"/>
2302 <test name="range_c"/>
2303 <test name="remove"/>
2304 <test name="reverse"/>
2306 <test name="stable_partition"/>
2307 <test name="transform"/>
2308 <test name="unpack_args"/>
2309 <test name="vector"/>
2310 <test name="vector_c"/>
2312 <toolset name="borland-5.8.1"/>
2314 <note author="A. Meredith" date="17 May 2006">
2315 This failure is caused by a problem with recursive templates and default template parameters, fixed in Update 2.
2317 </mark-expected-failures>
2319 <mark-expected-failures>
2320 <test name="apply"/>
2321 <test name="multiset"/>
2322 <test name="zip_view"/>
2324 <toolset name="borland-5.6*"/>
2325 <toolset name="borland-5.8*"/>
2326 <toolset name="borland-5.9*"/>
2327 <note author="Aleksey Gurtovoy" date="17 Sep 2004" refid="26"/>
2328 </mark-expected-failures>
2330 <mark-expected-failures>
2331 <test name="assert"/>
2334 <test name="front"/>
2335 <test name="has_xxx"/>
2336 <test name="multiset"/>
2337 <test name="no_has_xxx"/>
2338 <test name="zip_view"/>
2340 <toolset name="mipspro"/>
2341 <note author="Aleksey Gurtovoy" date="17 Sep 2004" refid="26"/>
2342 </mark-expected-failures>
2344 <mark-expected-failures>
2345 <test name="quote"/>
2346 <toolset name="borland-5.6*"/>
2347 <toolset name="borland-5.8*"/>
2348 <toolset name="borland-5.9*"/>
2349 <toolset name="msvc-6.5*"/>
2350 <toolset name="mipspro"/>
2351 <note author="Aleksey Gurtovoy" date="17 Sep 2004">
2352 This failure is caused by a lack of compiler support for template template
2355 </mark-expected-failures>
2357 <mark-expected-failures>
2360 <test name="set_c"/>
2361 <toolset name="borland-5.6*"/>
2362 <toolset name="borland-5.8*"/>
2363 <toolset name="borland-5.9*"/>
2364 <toolset name="gcc-2.95.3*"/>
2365 <toolset name="mipspro"/>
2366 <note author="Aleksey Gurtovoy" date="17 Sep 2004">
2367 This is an advanced functionality that hasn't been ported to the deficient
2368 compilers (yet). Patches are welcome!
2370 </mark-expected-failures>
2372 <mark-expected-failures>
2374 <toolset name="msvc-6.5*"/>
2375 <toolset name="msvc-7.0"/>
2376 <note author="Aleksey Gurtovoy" date="17 Sep 2004">
2377 This is an advanced functionality that hasn't been ported to the deficient
2378 compilers (yet). Patches are welcome!
2380 </mark-expected-failures>
2382 <mark-expected-failures>
2383 <test name="apply"/>
2384 <toolset name="gcc-4.1.*"/>
2385 <note author="Caleb Epstein">
2386 This is a regression in the gcc 4.1 series that has been
2387 fixed in gcc 4.2.0. See <a
2388 href="http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28088">bug
2389 #28088</a> for details.
2391 </mark-expected-failures>
2392 <mark-expected-failures>
2393 <test name="vector_c"/>
2394 <toolset name="sun-5.8"/>
2395 <note author="John Maddock">
2396 This is reported to be fixed in the next Sun
2399 </mark-expected-failures>
2400 <mark-expected-failures>
2402 <toolset name="acc"/>
2403 <note refid="38" author="Boris Gubenko"/>
2404 </mark-expected-failures>
2408 <!-- multi_array -->
2409 <library name="multi_array">
2411 <toolset name="borland-5.5*"/>
2412 <toolset name="borland-5.6*"/>
2413 <toolset name="borland-5.8*"/>
2414 <toolset name="borland-5.9*"/>
2415 <note author="Alisdair Meredith" date="30 Jan 2004">
2417 This library has never worked [on Borland 5.5.1 and 5.6.4], and the only tests
2418 that 'pass' are compile-fail tests failing for the wrong reasons!
2423 <toolset name="sunpro-5_3-sunos"/>
2424 <note author="Douglas Gregor" refid="3"/>
2426 <!-- RG: testing usability <mark-unusable>
2427 <toolset name="gcc-2.95.3*"/>
2428 <toolset name="gcc-2.95.3-linux"/>
2429 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
2430 <note author="Ronald Garcia" date="08 Jan 2006">
2432 These compiler/standard library combinations don't
2436 </mark-unusable> -->
2437 <test name="constructors">
2439 <toolset name="msvc-6.5"/>
2440 <note author="Ronald Garcia" date="13 Jul 2004">
2441 Known error in MSVC. see
2442 <a href="http://boost-consulting.com/boost/libs/multi_index/doc/compiler_specifics.html#msvc_60">
2443 http://boost-consulting.com/boost/libs/multi_index/doc/compiler_specifics.html#msvc_60</a>
2444 for more information.
2448 <mark-expected-failures>
2449 <test name="assign_to_array"/>
2450 <toolset name="gcc-2.95.3*"/>
2451 <note author="Aleksey Gurtovoy" date="21 Sep 2004" refid="2"/>
2452 </mark-expected-failures>
2453 <mark-expected-failures>
2454 <test name="assign"/>
2455 <test name="compare"/>
2456 <test name="concept_checks"/>
2457 <test name="constructors"/>
2458 <test name="iterators"/>
2459 <test name="resize"/>
2460 <test name="stl_interaction"/>
2461 <toolset name="gcc-2.95.3*"/>
2462 <note author="Doug Gregor" date="23 Jun 2005" refid="3"/>
2463 </mark-expected-failures>
2467 <!-- multi_index -->
2468 <library name="multi_index">
2470 <toolset name="borland-5.6*"/>
2471 <note author="J. López" date="05 Jul 2004" refid="17"/>
2474 <toolset name="borland-5.8*"/>
2475 <note author="Alisdair Meredith" date="26 May 2006"/>
2478 <toolset name="borland-5.9*"/>
2479 <note author="Alisdair Meredith" date="27 Feb 2007"/>
2482 <toolset name="gcc-2.95.3-linux"/>
2483 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
2484 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
2485 <note author="J. López" date="09 Jul 2004" refid="17"/>
2488 <toolset name="*como-4_3_3-msvc"/>
2489 <note author="J. López" date="30 Jul 2004">
2490 The VC++ 6.0 backend runs out of internal resources while
2491 trying to process the Comeau output for this library;
2492 Comeau Computing has been asked about a solution.
2493 On the other hand, Comeau 4.3.3 with VC++ 7.0 backend works
2498 <toolset name="sunpro-5_3-sunos"/>
2499 <toolset name="sunpro-5_8u1-sunos"/>
2500 <note author="J. López" date="22 Apr 2005" refid="17"/>
2503 <toolset name="dmc-8_43-stlport-4_5_3"/>
2504 <toolset name="dmc-8_44b-stlport-4_5_3"/>
2505 <toolset name="dmc-8_47-stlport-4_5_3"/>
2506 <note author="J. López" date="03 Jun 2005" refid="17"/>
2508 <mark-expected-failures>
2509 <test name="test_serialization"/>
2510 <toolset name="msvc-stlport"/>
2511 <toolset name="msvc-6.5_stlport4"/>
2512 <note author="J. López" date="10 Jan 2005">
2513 This error shows when using the dynamic version of the STLport
2514 library. The problem is reportedly fixed in STLport 5.0 (in beta
2515 stage as of this writing.)
2517 </mark-expected-failures>
2518 <mark-expected-failures>
2519 <test name="test_serialization"/>
2520 <toolset name="hp_cxx-65*"/>
2521 <toolset name="hp_cxx-71*"/>
2522 <note author="J. López" date="16 Mar 2006">
2523 This test fails due to limitations of the template
2524 instantiation model used in the testing environment
2525 (-timplicit_local) resulting in erroneous duplication of some
2526 function-static variables. The test passes with other template
2527 instantiation models.
2529 </mark-expected-failures>
2534 <library name="optional">
2535 <mark-expected-failures>
2536 <test name="optional_test_ref"/>
2537 <toolset name="msvc-6.5*"/>
2538 <toolset name="msvc-7.0"/>
2539 <note author="Aleksey Gurtovoy" refid="3"/>
2540 </mark-expected-failures>
2541 <mark-expected-failures>
2542 <test name="optional_test_ref"/>
2543 <toolset name="darwin-4.0.1"/>
2544 <toolset name="gcc-mingw-3.4.5"/>
2545 <toolset name="gcc-3.4.2_hpux_pa_risc"/>
2546 <toolset name="gcc-3.4.6_linux_ia64"/>
2547 <toolset name="gcc-4.2.*"/>
2548 <toolset name="gcc-4.1.2_sunos_i86pc"/>
2549 <note author="Fernando Cacciola" id="optional-compiler-bug">
2550 <p>This failure is caused by a compiler bug, and as far as we can
2551 tell, can't be worked around in the library, although we think
2552 the library might be made safer with respect to this bug.</p>
2554 <p>Specifics: the following simple test fails when it should succeed.</p>
2556 #include <cassert>
2563 assert(&x == &y);
2569 A a(x); // direct initialization works fine
2570 A b = x; // copy initialization causes x to be copied before it is bound
2574 The possible safety enhancement would be to cause the constructor
2575 in question to be explicit for optional<T const&>; that
2576 would prevent copy initialization.
2578 </mark-expected-failures>
2579 <mark-expected-failures>
2580 <test name="optional_test_ref_fail1"/>
2581 <toolset name="borland-5.6*"/>
2582 <toolset name="borland-5.8*"/>
2583 <toolset name="borland-5.9*"/>
2584 <note author="Fernando Cacciola" refid="optional-compiler-bug"/>
2585 </mark-expected-failures>
2586 <mark-expected-failures>
2587 <test name="optional_test_fail3a"/>
2588 <toolset name="gcc-3_3-darwin"/>
2589 <note author="Fernando Cacciola" refid="optional-compiler-bug"/>
2590 </mark-expected-failures>
2591 <mark-expected-failures>
2592 <test name="optional_test_inplace_fail2"/>
2593 <toolset name="gcc-3_3-darwin"/>
2594 <note author="Fernando Cacciola" refid="optional-compiler-bug"/>
2595 </mark-expected-failures>
2598 <library name="pool">
2600 <toolset name="gcc-2.95.3-*"/>
2601 <note author="Doug Gregor" refid="2"/>
2605 <!-- preprocessor -->
2606 <library name="preprocessor">
2607 <mark-expected-failures>
2609 <toolset name="cw-8.3"/>
2610 <note author="Paul Mensonides" refid="2"/>
2611 </mark-expected-failures>
2615 <library name="proto">
2617 <toolset name="sun-5.7"/>
2618 <toolset name="sun-5.8"/>
2619 <toolset name="sun-5.9"/>
2620 <toolset name="sun-5.10"/>
2621 <toolset name="borland-*"/>
2622 <toolset name="vacpp"/>
2627 <library name="rational">
2628 <mark-expected-failures>
2629 <test name="rational_test"/>
2630 <toolset name="sun-5.8"/>
2631 <note author="J. López" date="19 Oct 2006">
2632 The test is exposing the following known error of Sun Studio 11:
2633 overload resolution fails if
2634 a) some class has a conversion operator to a reference to
2635 a built-in type, and
2636 b) overload resolution involves a user-defined operator as well
2637 as a built-in operator, and
2638 c) the built-in operator takes the result of the conversion
2639 mentioned in a) as an operand.
2640 A fix will be reportedly included in patch no 6 of Sun Studio 11.
2642 </mark-expected-failures>
2645 <!-- serialization -->
2646 <library name="serialization">
2648 <toolset name="mipspro*" />
2649 <toolset name="dmc*" />
2650 <toolset name="sunpro*" />
2651 <note author="Robert Ramey" date="13 Jul 2007" refid="9,17,18"/>
2654 <toolset name="gcc-2.95.3-linux"/>
2655 <note author="Robert Ramey" date="12 Feb 05" refid="18,19"/>
2658 <mark-expected-failures>
2659 <test name="*_warchive"/>
2660 <test name="test_codecvt_null"/>
2661 <test name="test_utf8_codecvt"/>
2662 <toolset name="mingw*"/>
2663 <toolset name="*mingw*"/>
2664 <toolset name="*cygwin*"/>
2665 <toolset name="gcc-2.95.3-linux"/>
2666 <toolset name="*como-4_3_3*"/>
2667 <note author="Robert Ramey,Roland Schwarz" date="16 Feb 07" refid="19"/>
2668 </mark-expected-failures>
2670 <mark-expected-failures>
2671 <test name="test_void_cast*"/>
2672 <toolset name="msvc-6.5*"/>
2673 <note author="Robert Ramey" date="20 Sep 2004" refid="16,29"/>
2674 </mark-expected-failures>
2676 <mark-expected-failures>
2677 <test name="test_reset_object_address*"/>
2678 <toolset name="msvc-6.5*"/>
2679 <note author="Robert Ramey" date="12 Feb 05" refid="6,29"/>
2680 </mark-expected-failures>
2682 <mark-expected-failures>
2683 <test name="test_reset_object_address*"/>
2684 <toolset name="msvc-7.0"/>
2685 <note author="J. López" date="20 Dec 2006">
2686 This error shows when the code has become too complex for the
2687 compiler to handle. The problem has no relationship with the
2688 functionality being tested, which in fact does work for
2691 </mark-expected-failures>
2693 <mark-expected-failures>
2694 <test name="test_const"/>
2695 <toolset name="msvc-6.5*"/>
2696 <toolset name="msvc-7.0"/>
2697 <note author="Aleksey Gurtovoy" refid="29"/>
2698 </mark-expected-failures>
2700 <mark-expected-failures>
2701 <test name="test_demo_pimpl"/>
2702 <test name="test_diamond*"/>
2703 <test name="test_mult_archive_types"/>
2704 <toolset name="msvc-6.5*"/>
2705 <toolset name="msvc-7.0"/>
2706 <note author="Robert Ramey" refid="6">
2707 msvc 6 compiler failure. The facility being tested conflicts the the
2708 compiler in a fundamental way and cannnot be worked around.
2710 </mark-expected-failures>
2712 <mark-expected-failures>
2713 <test name="test_mi*"/>
2714 <toolset name="msvc-6.5*"/>
2715 <note author="Robert Ramey" refid="6">
2716 msvc 6 compiler failure. The facility being tested conflicts the the
2717 compiler in a fundamental way and cannnot be worked around.
2719 </mark-expected-failures>
2721 <mark-expected-failures>
2722 <test name="*_dll"/>
2723 <toolset name="msvc-stlport"/>
2724 <toolset name="msvc-6.5_stlport4"/>
2725 <note author="Robert Ramey">
2726 This failure appears when STLPort is built and used as a DLL with msvc 6.
2727 STLPort suggests that the next version of STLPort(5.0) will include a workaround
2730 </mark-expected-failures>
2732 <mark-expected-failures>
2734 <toolset name="gcc-2.95.3-stlport*"/>
2735 <note author="Aleksey Gurtovoy">
2736 The library is believed to work in this configuration <i>if compiled against
2737 Spirit 1.6</i>. The latter is not provided by the particular testing
2738 environment these tests have been run in.
2740 </mark-expected-failures>
2742 <mark-expected-failures>
2743 <test name="test_exported*"/>
2744 <test name="test_mi*"/>
2745 <test name="test_mult_archive_types*"/>
2746 <test name="test_no_rtti*"/>
2747 <test name="test_non_default_ctor2*"/>
2748 <test name="test_registered*"/>
2749 <test name="test_shared_ptr*"/>
2750 <test name="test_unregistered*"/>
2751 <toolset name="cw*"/>
2752 <note author="Robert Ramey" refid="29">
2753 All tests that serialize derived pointers currently fail with Metrowerks compilers.
2755 </mark-expected-failures>
2757 <mark-expected-failures>
2758 <test name="test_no_rtti_*"/>
2759 <toolset name="borland-5.6*"/>
2760 <toolset name="borland-5.8*"/>
2761 <toolset name="borland-5.9*"/>
2762 <note author="Aleksey Gurtovoy" refid="29"/>
2763 </mark-expected-failures>
2765 <mark-expected-failures>
2766 <test name="test_smart_cast"/>
2767 <toolset name="intel-7.1-linux"/>
2768 <note author="Aleksey Gurtovoy" refid="29"/>
2769 </mark-expected-failures>
2771 <mark-expected-failures>
2772 <test name="test_diamond*"/>
2773 <toolset name="cw-8*"/>
2774 <toolset name="cw-9.5-darwin"/>
2775 <note author="Rene Rivera">
2776 The CW compilers have problems with the static construction idiom used to
2777 implement the type registration in the Boost.Serialization library. In many
2778 cases CW specific work arounds are implemented in the library but this one
2779 is not immediately solvable. There is a user work around possible, please
2780 contact the library developers on the Boost list for information on the
2781 work around if needed.
2783 </mark-expected-failures>
2785 <mark-expected-failures>
2786 <test name="test_class_info_load_text*"/>
2787 <test name="test_class_info_load_xml_warchive*"/>
2788 <toolset name="cw-9.5-darwin"/>
2789 <note author="Rene Rivera" refid="29"/>
2790 </mark-expected-failures>
2792 <mark-expected-failures>
2793 <test name="test_class_info_load_text_warchive_dll"/>
2794 <toolset name="msvc-6.5"/>
2795 <note author="Doug Gregor" refid="29"/>
2796 </mark-expected-failures>
2798 <mark-expected-failures>
2799 <test name="test_variant_*"/>
2800 <toolset name="hp_cxx-65*"/>
2801 <note author="Markus Schoepflin">
2802 The variant library is not supported for this compiler version.
2803 Therefore serialization of variants doesn't work.
2805 </mark-expected-failures>
2807 <mark-expected-failures>
2808 <test name="*_warchive"/>
2809 <toolset name="qcc-3.3.5*gpp"/>
2810 <note author="Jim Douglas" date="12 Feb 06" refid="36"/>
2811 </mark-expected-failures>
2813 <mark-expected-failures>
2814 <test name="test_variant_*"/>
2815 <toolset name="borland-5.8*"/>
2816 <toolset name="borland-5.9*"/>
2817 <note author="Vladimir Prus">
2818 The compiler fails with an error supposedly related to std::fpos<>::_Stz from the
2819 <iosfwd> header. It is not known what causes the compiler to instantiate this
2820 field and what causes the instantiation to fail.
2822 </mark-expected-failures>
2828 <library name="smart_ptr">
2829 <mark-expected-failures>
2830 <test name="shared_ptr_assign_fail"/>
2831 <toolset name="gcc-2.9*"/>
2832 <toolset name="sunpro-5_3-sunos"/>
2833 <note refid="32" author="Peter Dimov"/>
2834 </mark-expected-failures>
2835 <mark-expected-failures>
2836 <test name="weak_ptr_test"/>
2837 <toolset name="hp_cxx-71_006_*"/>
2838 <note author="Markus Schoepflin" refid="3"/>
2839 </mark-expected-failures>
2840 <mark-expected-failures>
2841 <test name="auto_ptr_rv_test"/>
2842 <toolset name="gcc-2.9*"/>
2843 <toolset name="borland-5*"/>
2844 <toolset name="cw-8*"/>
2845 <note refid="31" author="Peter Dimov"/>
2846 </mark-expected-failures>
2847 <mark-expected-failures>
2848 <test name="pointer_to_other_test"/>
2849 <toolset name="msvc-6.5*"/>
2850 <toolset name="msvc-7.0"/>
2851 <note refid="31" author="Peter Dimov"/>
2852 </mark-expected-failures>
2853 <mark-expected-failures>
2854 <test name="shared_ptr_alloc2_test"/>
2855 <toolset name="msvc-6.5*"/>
2856 <note refid="31" author="Peter Dimov"/>
2857 </mark-expected-failures>
2858 <mark-expected-failures>
2859 <test name="shared_ptr_move_test"/>
2861 <note refid="40" author="Boris Gubenko"/>
2862 </mark-expected-failures>
2863 <mark-expected-failures>
2864 <test name="sp_convertible_test"/>
2865 <test name="wp_convertible_test"/>
2866 <test name="ip_convertible_test"/>
2867 <toolset name="borland-*"/>
2868 <note refid="31" author="Peter Dimov"/>
2869 </mark-expected-failures>
2873 <library name="spirit">
2875 <toolset name="msvc-6.5*"/>
2876 <toolset name="borland-5.5*"/>
2877 <toolset name="borland-5.6*"/>
2878 <toolset name="borland-5.8*"/>
2879 <toolset name="msvc-7.0"/>
2880 <toolset name="gcc-2.95.3-linux"/>
2881 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
2882 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
2883 <toolset name="sunpro-5_3-sunos"/>
2887 Historically, Spirit supported a lot of compilers, including (to some
2888 extent) poorly conforming compilers such as VC6. Spirit v1.6.x will be
2889 the last release that will support older poorly conforming compilers.
2890 Starting from Spirit v1.8.0, ill conforming compilers will not be
2891 supported. If you are still using one of these older compilers, you can
2892 still use Spirit v1.6.x.
2895 The reason why Spirit v1.6.x worked on old non-conforming compilers is
2896 that the authors laboriously took the trouble of searching for
2897 workarounds to make these compilers happy. The process takes a lot of
2898 time and energy, especially when one encounters the dreaded ICE or
2899 "Internal Compiler Error". Sometimes searching for a single workaround
2900 takes days or even weeks. Sometimes, there are no known workarounds. This
2901 stifles progress a lot. And, as the library gets more progressive and
2902 takes on more advanced C++ techniques, the difficulty is escalated to
2906 Spirit v1.6.x will still be supported. Maintenance and bug fixes will
2907 still be applied. There will still be active development for the back-
2908 porting of new features introduced in Spirit v1.8.0 (and Spirit 1.9.0)
2909 to lesser able compilers; hopefully, fueled by contributions from the
2910 community. For instance, there is already a working AST tree back-port
2911 for VC6 and VC7 by Peder Holt.
2915 <mark-expected-failures>
2916 <test name="action_tests*"/>
2917 <toolset name="iw-7_1-vc6"/>
2918 <note author="Aleksey Gurtovoy" refid="4"/>
2919 </mark-expected-failures>
2920 <mark-expected-failures>
2921 <test name="ast_calc_tests*"/>
2922 <test name="closure_tests*"/>
2923 <test name="multi_pass_compile_tests"/>
2924 <test name="repeat_ast_tests*"/>
2925 <toolset name="intel-8.0-linux"/>
2926 <toolset name="intel-8.1-linux"/>
2927 <note author="Aleksey Gurtovoy">
2928 This failure is caused by a compiler bug that manifests itself in the
2929 particular environment/hardware configuration the test has been run in.
2930 You may or may not experience this issue in your local setup.
2932 </mark-expected-failures>
2933 <mark-expected-failures>
2934 <test name="escape_char_parser_tests*"/>
2935 <toolset name="intel-7.1-linux"/>
2936 <toolset name="intel-7.1-stdlib-default-linux"/>
2937 <note author="Aleksey Gurtovoy" refid="19"/>
2938 </mark-expected-failures>
2939 <mark-expected-failures>
2940 <test name="escape_char_parser_tests*"/>
2941 <toolset name="iw-7_1-vc6*"/>
2942 <note author="Aleksey Gurtovoy" refid="28"/>
2943 </mark-expected-failures>
2944 <mark-expected-failures>
2945 <test name="chset_tests*"/>
2946 <toolset name="iw-7_1-vc6-stlp-4_5_3"/>
2947 <note author="Aleksey Gurtovoy" refid="28"/>
2948 </mark-expected-failures>
2949 <mark-expected-failures>
2950 <test name="int_numerics"/>
2951 <test name="karma_pattern*"/>
2952 <test name="karma_sequence"/>
2954 <test name="sequence"/>
2955 <toolset name="acc"/>
2956 <note author="Boris Gubenko" refid="47"/>
2957 </mark-expected-failures>
2958 <mark-expected-failures>
2959 <test name="lexertl3"/>
2960 <test name="lexertl4"/>
2961 <test name="lexertl5"/>
2962 <toolset name="gcc-3.4.6_linux_ia64"/>
2963 <note author="Boris Gubenko">
2964 With GCC 3.4.6 the test fails with ICE: internal compiler error.
2965 The footprint is similar to that in GCC Bugzilla Bug 34950
2966 except 34950 is a regression introduced in GCC 4.2.3. In any
2967 case, whatever the problem is, the GCC 4.x series does not seem
2968 to have it: the test compiles just fine with GCC 4.x compiler.
2970 </mark-expected-failures>
2974 <library name="typeof">
2976 <toolset name="gcc-2.95.*"/>
2977 <toolset name="sunpro*"/>
2978 <toolset name="borland-5.6.*"/>
2979 <note author="Arkadiy Vertleyb">
2980 This compiler is not supported.
2983 <test name="*_native" category="Native compiler support">
2985 <toolset name="acc*"/>
2986 <toolset name="intel-vc71-win*"/>
2987 <toolset name="intel-vc8-win*"/>
2988 <toolset name="como-4_3_3-vc7_1"/>
2989 <toolset name="hp_cxx*"/>
2990 <toolset name="sun-5.*"/>
2991 <toolset name="borland-5*"/>
2992 <toolset name="mipspro*"/>
2993 <note author="Arkadiy Vertleyb">
2994 Native mode is not supported for this compiler.
2998 <mark-expected-failures>
2999 <test name="*_emulation"/>
3000 <toolset name="msvc-6.5*"/>
3001 <toolset name="msvc-7.0"/>
3002 <toolset name="cw-8_*"/>
3003 <note author="Arkadiy Vertleyb">
3004 Emulation mode is not supported for this compiler.
3006 </mark-expected-failures>
3007 <mark-expected-failures>
3008 <test name="function_native"/>
3009 <test name="template_tpl_native"/>
3010 <test name="function_binding_native"/>
3011 <test name="odr_no_uns"/>
3012 <toolset name="msvc-6.5*"/>
3013 <toolset name="msvc-7.0"/>
3014 <note author="Arkadiy Vertleyb">
3015 The feature is not supported by this compiler.
3017 </mark-expected-failures>
3018 <mark-expected-failures>
3019 <test name="function_native"/>
3020 <toolset name="cw-8_*"/>
3021 <note author="Arkadiy Vertleyb">
3022 The feature is not supported by this compiler.
3024 </mark-expected-failures>
3025 <mark-expected-failures>
3026 <test name="function_binding_emulation"/>
3027 <test name="function_emulation"/>
3028 <test name="function_ptr_from_tpl_emulation"/>
3029 <test name="modifiers_emulation"/>
3030 <test name="nested_typedef_emulation"/>
3031 <toolset name="borland-5.8*"/>
3032 <note author="Peder Holt">
3033 The feature is not supported by this compiler.
3035 </mark-expected-failures>
3039 <library name="function">
3041 <toolset name="sunpro-5_3-sunos"/>
3042 <note author="Douglas Gregor" refid="3"/>
3044 <test name="allocator_test">
3046 <toolset name="msvc-6.5"/>
3047 <toolset name="msvc-7.0"/>
3048 <note author="B. Dawes" refid="5"/>
3051 <test name="contains_test">
3053 <toolset name="msvc-6.5*"/>
3054 <note refid="3" author="D. Gregor"/>
3057 <test name="function_30">
3059 <toolset name="vacpp"/>
3060 <note refid="16" author="D. Gregor"/>
3063 <test name="function_arith_cxx98">
3065 <toolset name="borland-5.6*"/>
3066 <toolset name="borland-5.8*"/>
3067 <toolset name="borland-5.9*"/>
3068 <toolset name="msvc-6.5"/>
3069 <toolset name="msvc-7.0"/>
3070 <note author="B. Dawes" refid="3"/>
3073 <test name="function_ref_cxx98">
3075 <toolset name="borland-5.6*"/>
3076 <toolset name="borland-5.8*"/>
3077 <toolset name="borland-5.9*"/>
3078 <toolset name="msvc-6.5"/>
3079 <toolset name="msvc-7.0"/>
3080 <note author="B. Dawes" refid="3"/>
3083 <test name="lambda_test">
3085 <toolset name="borland-5.6*"/>
3086 <toolset name="borland-5.8*"/>
3087 <toolset name="borland-5.9*"/>
3088 <toolset name="msvc-6.5"/>
3089 <toolset name="msvc-7.0"/>
3090 <note author="B. Dawes" refid="3"/>
3093 <toolset name="cw-8.3*"/>
3094 <note author="B. Dawes" refid="2"/>
3097 <test name="lib_function_test">
3099 <toolset name="borland-5.6*"/>
3100 <toolset name="borland-5.8*"/>
3101 <toolset name="borland-5.9*"/>
3102 <toolset name="msvc-6.5"/>
3103 <toolset name="msvc-7.0"/>
3104 <note author="B. Dawes" refid="3"/>
3107 <toolset name="cw-8.3*"/>
3108 <note author="B. Dawes" refid="2"/>
3111 <test name="mem_fun_cxx98">
3113 <toolset name="borland-5.6*"/>
3114 <toolset name="borland-5.8*"/>
3115 <toolset name="borland-5.9*"/>
3116 <toolset name="msvc-6.5"/>
3117 <toolset name="msvc-7.0"/>
3118 <note author="B. Dawes" refid="3"/>
3121 <toolset name="cw-8.3*"/>
3122 <note author="B. Dawes" refid="2"/>
3125 <test name="std_bind_cxx98">
3127 <toolset name="borland-5.6*"/>
3128 <toolset name="borland-5.8*"/>
3129 <toolset name="borland-5.9*"/>
3130 <toolset name="msvc-6.5"/>
3131 <toolset name="msvc-7.0"/>
3132 <note author="B. Dawes" refid="3"/>
3135 <test name="std_bind_portable">
3137 <toolset name="msvc-6.5"/>
3138 <note author="B. Dawes" refid="5"/>
3141 <test name="sum_avg_cxx98">
3143 <toolset name="borland-5.6*"/>
3144 <toolset name="borland-5.8*"/>
3145 <toolset name="borland-5.9*"/>
3146 <toolset name="msvc-6.5"/>
3147 <toolset name="msvc-7.0"/>
3148 <note author="B. Dawes" refid="3"/>
3155 <library name="iterator">
3156 <test name="interoperable_fail" category="Corner-case tests">
3158 <toolset name="gcc-3.3*"/>
3159 <toolset name="gcc-3.2*"/>
3160 <toolset name="gcc-2*"/>
3161 <toolset name="gcc"/>
3162 <toolset name="mingw"/>
3163 <toolset name="borland*"/>
3164 <toolset name="cw-8*"/>
3165 <toolset name="qcc-3.3*"/>
3166 <note author="D. Abrahams">
3167 This failure is caused by a compiler bug. Templated operators
3168 that combine different iterators built with iterator_facade or
3169 iterator_adaptor may be present in an overload set even when those
3170 iterators are not interoperable. The usual result is that error
3171 messages generated by illegal use of these operators will be of
3177 <test name="is_convertible_fail" category="Corner-case tests">
3179 <toolset name="gcc-2*"/>
3180 <toolset name="gcc"/>
3181 <toolset name="mingw"/>
3182 <toolset name="borland*"/>
3183 <toolset name="cw-8*"/>
3184 <toolset name="msvc-6*"/>
3185 <toolset name="msvc-7.0*"/>
3186 <note author="D. Abrahams">
3187 This failure is caused by a compiler bug.
3188 <code>is_convertible<T,U>::value</code> may be true for unrelated
3189 iterators <code>T</code> and <code>U</code>
3190 (including many of the Boost specialized adaptors) which use
3191 <code>enable_if_convertible</code> to restrict the applicability
3192 of converting constructors, even when <code>T</code> is not
3193 convertible to <code>U</code> because instantiating the
3194 conversion will cause a compilation failure.
3199 <test name="indirect_iter_member_types" category="Corner-case tests"/>
3201 <mark-expected-failures>
3202 <test name="indirect_iter_member_types"/>
3203 <test name="pointee"/>
3204 <toolset name="borland-5.6*"/>
3205 <toolset name="borland-5.8*"/>
3206 <toolset name="borland-5.9*"/>
3207 <note author="D. Abrahams">
3208 This failure is caused by a compiler bug. The
3209 compiler tends to drop const-ness and as a result
3210 some indirect_iterators will have pointer and
3211 reference members of <code>T*</code> and <code>T&</code> that should
3212 have been <code>T const*</code> and <code>T const&</code>.
3214 </mark-expected-failures>
3216 <mark-expected-failures>
3217 <test name="zip_iterator_test"/>
3218 <toolset name="borland-5.6*"/>
3219 <toolset name="borland-5.8*"/>
3220 <toolset name="borland-5.9*"/>
3221 <note author="Aleksey Gurtovoy" date="19 Sep 2004" refid="26"/>
3222 </mark-expected-failures>
3224 <mark-expected-failures>
3225 <test name="is_lvalue_iterator"/>
3226 <toolset name="acc*"/>
3227 <note author="Boris Gubenko">
3228 For some currently unknown reason, with aCC, this test can be compiled
3229 only in strict ansi mode. Since on HP-UX/aCC boost testing is done in the
3230 default compilation mode, this test fails to compile on this platform.
3232 </mark-expected-failures>
3238 <library name="math">
3240 <toolset name="gcc-2.95.3-*"/>
3241 <note author="Doug Gregor" refid="3"/>
3244 <toolset name="borland-5.9.2"/>
3245 <note author="John Maddock">
3246 Sadly Borland-5.9.2 has an even harder time compiling this
3247 library than earlier versions did. There are currently too
3248 many issues to stand a chance of porting to this compiler.
3251 <mark-expected-failures>
3252 <test name="mpfr_concept_check"/>
3253 <test name="ntl_concept_check"/>
3255 <note author="John Maddock">
3256 This test relies on external software being installed in order to pass.
3258 </mark-expected-failures>
3259 <mark-expected-failures>
3260 <test name="test_traits"/>
3261 <toolset name="gcc-3.3.6"/>
3262 <note author="John Maddock">
3263 This compiler is not sufficiently conforming to correctly handle these tests.
3265 </mark-expected-failures>
3266 <mark-expected-failures>
3267 <test name="test_tr1_long_double"/>
3268 <toolset name="darwin*"/>
3269 <toolset name="intel-linux-10.0"/>
3270 <toolset name="intel-linux-9*"/>
3271 <toolset name="intel-linux-8*"/>
3272 <note author="John Maddock">
3273 Some versions of the Darwin platform have insufficient long double support
3274 for us to be able to run this test.
3276 </mark-expected-failures>
3277 <mark-expected-failures>
3278 <test name="test_policy_2"/>
3279 <toolset name="acc"/>
3280 <toolset name="gcc-mingw-3.4.5"/>
3281 <note author="John Maddock">
3282 This test takes too long to build for this compiler and times out.
3284 </mark-expected-failures>
3285 <mark-expected-failures>
3286 <test name="test_traits"/>
3287 <toolset name="sun-5.8"/>
3288 <toolset name="sun-5.9"/>
3289 <note author="John Maddock">
3290 This is a compiler bug: it is unable to use
3291 SFINAE to detect the presence of specific
3294 </mark-expected-failures>
3295 <mark-expected-failures>
3296 <test name="std_real_concept_check"/>
3297 <test name="test_instantiate1"/>
3298 <test name="test_policy_sf"/>
3299 <toolset name="sun-5.8"/>
3300 <toolset name="sun-5.9"/>
3301 <note author="John Maddock">
3302 This is a compiler bug: it is unable to resolve the
3303 overloaded functions.
3305 </mark-expected-failures>
3306 <mark-expected-failures>
3307 <test name="test_binomial_real_concept"/>
3308 <test name="test_ibeta_inv_ab_real_concept"/>
3309 <test name="test_igamma_inva_real_concept"/>
3310 <toolset name="sun-5.9"/>
3311 <toolset name="sun-5.8"/>
3312 <note author="John Maddock">
3313 This test takes too long to execute and times out.
3315 </mark-expected-failures>
3316 <mark-expected-failures>
3317 <test name="dist_binomial_incl_test"/>
3318 <test name="dist_neg_binom_incl_test"/>
3319 <test name="dist_poisson_incl_test"/>
3320 <test name="test_binomial_double"/>
3321 <test name="test_binomial_float"/>
3322 <test name="test_binomial_long_double"/>
3323 <test name="test_binomial_real_concept"/>
3324 <test name="test_negative_binomial_double"/>
3325 <test name="test_negative_binomial_float"/>
3326 <test name="test_negative_binomial_long_double"/>
3327 <test name="test_negative_binomial_real_concept"/>
3328 <test name="test_poisson_double"/>
3329 <test name="test_poisson_float"/>
3330 <test name="test_poisson_long_double"/>
3331 <test name="test_poisson_real_concept"/>
3332 <test name="tools_roots_inc_test"/>
3333 <toolset name="sun-5.8"/>
3334 <toolset name="sun-5.7"/>
3335 <note author="John Maddock">
3336 These tests fail with an internal compiler error: there is no
3337 known workaround at present, except to use Sun-5.9 which does
3338 build this code correctly.
3340 </mark-expected-failures>
3341 <mark-expected-failures>
3342 <test name="log1p_expm1_test"/>
3343 <test name="test_bernoulli"/>
3344 <test name="test_beta_dist"/>
3345 <test name="test_binomial_float"/>
3346 <test name="test_binomial_double"/>
3347 <test name="test_binomial_coeff"/>
3348 <test name="test_carlson"/>
3349 <test name="test_cauchy"/>
3350 <test name="test_cbrt"/>
3351 <test name="test_chi_squared"/>
3352 <test name="test_classify"/>
3353 <test name="test_dist_overloads"/>
3354 <test name="test_ellint_3"/>
3355 <test name="test_exponential_dist"/>
3356 <test name="test_factorials"/>
3357 <test name="test_find_location"/>
3358 <test name="test_find_scale"/>
3359 <test name="test_fisher_f"/>
3360 <test name="test_gamma_dist"/>
3361 <test name="test_hermite"/>
3362 <test name="test_ibeta_inv_float"/>
3363 <test name="test_ibeta_inv_double"/>
3364 <test name="test_ibeta_inv_ab_float"/>
3365 <test name="test_igamma_inv_float"/>
3366 <test name="test_igamma_inv_double"/>
3367 <test name="test_igamma_inva_float"/>
3368 <test name="test_igamma_inva_double"/>
3369 <test name="test_instantiate1"/>
3370 <test name="test_instantiate1"/>
3371 <test name="test_laguerre"/>
3372 <test name="test_legendre"/>
3373 <test name="test_lognormal"/>
3374 <test name="test_negative_binomial_float"/>
3375 <test name="test_negative_binomial_double"/>
3376 <test name="test_normal"/>
3377 <test name="test_rayleigh"/>
3378 <test name="test_remez"/>
3379 <test name="test_roots"/>
3380 <test name="test_students_t"/>
3381 <test name="test_toms748_solve"/>
3382 <test name="test_triangular"/>
3383 <test name="test_uniform"/>
3384 <test name="test_policy"/>
3385 <test name="test_policy_sf"/>
3386 <test name="test_bessel_j"/>
3387 <test name="test_bessel_y"/>
3388 <test name="dist_beta_incl_test"/>
3389 <test name="dist_cauchy_incl_test"/>
3390 <test name="dist_chi_squared_incl_test"/>
3391 <test name="dist_exponential_incl_test"/>
3392 <test name="dist_fisher_f_incl_test"/>
3393 <test name="dist_gamma_incl_test"/>
3394 <test name="dist_lognormal_incl_test"/>
3395 <test name="dist_normal_incl_test"/>
3396 <test name="dist_students_t_incl_test"/>
3397 <test name="sf_beta_incl_test"/>
3398 <test name="sf_bessel_incl_test"/>
3399 <test name="sf_cbrt_incl_test"/>
3400 <test name="sf_gamma_incl_test"/>
3401 <test name="sf_legendre_incl_test"/>
3402 <test name="std_real_concept_check"/>
3403 <test name="test_traits"/>
3404 <test name="tools_remez_inc_test"/>
3405 <test name="tools_roots_inc_test"/>
3406 <test name="tools_series_inc_test"/>
3407 <test name="tools_solve_inc_test"/>
3408 <test name="tools_test_data_inc_test"/>
3409 <test name="common_factor_test"/>
3410 <test name="octonion_test"/>
3411 <test name="quaternion_test"/>
3412 <test name="complex_test"/>
3413 <toolset name="borland-5.6.*"/>
3414 <note author="John Maddock">
3415 This compiler is not sufficiently conforming to correctly handle these tests.
3417 </mark-expected-failures>
3418 <mark-expected-failures>
3419 <test name="test_bernoulli"/>
3420 <test name="test_beta_dist"/>
3421 <test name="test_binomial_float"/>
3422 <test name="test_binomial_double"/>
3423 <test name="test_binomial_coeff"/>
3424 <test name="test_cauchy"/>
3425 <test name="test_dist_overloads"/>
3426 <test name="test_ellint_3"/>
3427 <test name="test_exponential_dist"/>
3428 <test name="test_factorials"/>
3429 <test name="test_find_location"/>
3430 <test name="test_find_scale"/>
3431 <test name="test_hermite"/>
3432 <test name="test_ibeta_inv_float"/>
3433 <test name="test_ibeta_inv_double"/>
3434 <test name="test_ibeta_inv_ab_float"/>
3435 <test name="test_igamma_inva_float"/>
3436 <test name="test_igamma_inva_double"/>
3437 <test name="test_instantiate1"/>
3438 <test name="test_instantiate1"/>
3439 <test name="test_laguerre"/>
3440 <test name="test_legendre"/>
3441 <test name="test_lognormal"/>
3442 <test name="test_negative_binomial_double"/>
3443 <test name="test_normal"/>
3444 <test name="test_rayleigh"/>
3445 <test name="test_remez"/>
3446 <test name="test_roots"/>
3447 <test name="test_toms748_solve"/>
3448 <test name="test_policy"/>
3449 <test name="test_policy_sf"/>
3450 <test name="dist_cauchy_incl_test"/>
3451 <test name="dist_exponential_incl_test"/>
3452 <test name="dist_lognormal_incl_test"/>
3453 <test name="dist_normal_incl_test"/>
3454 <test name="sf_gamma_incl_test"/>
3455 <test name="sf_legendre_incl_test"/>
3456 <test name="std_real_concept_check"/>
3457 <test name="test_traits"/>
3458 <test name="tools_remez_inc_test"/>
3459 <test name="tools_roots_inc_test"/>
3460 <test name="tools_series_inc_test"/>
3461 <test name="tools_solve_inc_test"/>
3462 <test name="tools_test_data_inc_test"/>
3463 <test name="complex_test"/>
3464 <toolset name="borland-5.8.2"/>
3465 <note author="John Maddock">
3466 This compiler is not sufficiently conforming to correctly handle these tests.
3468 </mark-expected-failures>
3469 <mark-expected-failures>
3470 <test name="octonion_test"/>
3471 <test name="quaternion_test"/>
3472 <toolset name="gcc-3.4.3_sunos"/>
3473 <note author="Caleb Epstein">
3474 There appears to be a bug in gcc's <code>std::exp (long
3475 double)</code> on this platform.
3477 </mark-expected-failures>
3478 <mark-expected-failures>
3479 <test name="test_remez"/>
3480 <toolset name="hp_cxx-71_006_tru64"/>
3481 <note author="John Maddock">
3482 For some reason taking the address of std library math functions fails
3483 on this platform: this is a problem for our test code, not the library.
3485 </mark-expected-failures>
3486 <mark-expected-failures>
3487 <test name="special_functions_test"/>
3488 <test name="octonion_test"/>
3489 <test name="quaternion_test"/>
3490 <test name="quaternion_mult_incl_test"/>
3491 <toolset name="msvc-6*"/>
3492 <note author="John Maddock">
3493 This compiler is not sufficiently conforming to compile these tests.
3495 </mark-expected-failures>
3496 <mark-expected-failures>
3497 <test name="complex_test"/>
3498 <test name="log1p_expm1_test"/>
3499 <toolset name="sunpro*"/>
3500 <note author="John Maddock">
3501 std::numeric_limits<long double>::infinity() is apparently
3502 broken in this compiler: it's filed as bug 6347520 with Sun.
3504 </mark-expected-failures>
3505 <mark-expected-failures>
3506 <test name="complex_test"/>
3507 <toolset name="msvc-6*"/>
3508 <note author="John Maddock">
3509 Incomplete std::complex support make these tests pointless
3510 (the complex trig functions are absent).
3512 </mark-expected-failures>
3513 <mark-expected-failures>
3514 <test name="special_functions_test"/>
3515 <test name="octonion_test"/>
3516 <test name="quaternion_test"/>
3517 <test name="quaternion_mult_incl_test"/>
3518 <toolset name="sun-5.8"/>
3519 <note author="John Maddock">
3520 These have yet to fully investigated, but the code is known
3521 to compile with more conforming compilers, probably workarounds
3522 are possible if someone is prepared to invest the time.
3524 </mark-expected-failures>
3525 <mark-expected-failures>
3526 <test name="quaternion_test"/>
3527 <toolset name="msvc-7.1_stlport4"/>
3528 <note author="John Maddock">
3529 Appears to be a bug in STLport's complex abs function, but needs more investigation.
3531 </mark-expected-failures>
3532 <mark-expected-failures>
3533 <test name="special_functions_test"/>
3534 <toolset name="msvc-7.1_stlport4"/>
3535 <note author="John Maddock">
3536 This appears to be a problem with STLPort's abs function: the issue only effects the
3537 test code. A workaround should be possible but users should be encouraged to use
3540 </mark-expected-failures>
3541 <mark-expected-failures>
3542 <test name="quaternion_test"/>
3543 <test name="octonion_test"/>
3544 <toolset name="gcc-cygwin*"/>
3545 <note author="John Maddock">
3546 No true long double standard lib support causes these tests to fail.
3548 </mark-expected-failures>
3549 <mark-expected-failures>
3550 <test name="quaternion_test"/>
3551 <test name="complex_test"/>
3552 <test name="special_functions_test"/>
3553 <toolset name="intel-linux*"/>
3554 <note author="John Maddock">
3555 This is Intel issue 409291, it should be fixed from
3556 compiler package l_cc_c_9.1.046 onwards.
3558 </mark-expected-failures>
3559 <mark-expected-failures>
3560 <test name="complex_test"/>
3561 <toolset name="qcc-3.3.5*cpp"/>
3562 <note author="Jim Douglas" date="14 Feb 06" refid="27"/>
3563 </mark-expected-failures>
3564 <mark-expected-failures>
3565 <test name="common_factor_test"/>
3566 <toolset name="msvc-6.5_stlport*"/>
3567 <toolset name="msvc-7.1_stlport*"/>
3568 <note author="John Maddock">
3569 This failure appears to be caused by a compiler bug: please note
3570 that the issue only effects the test suite, not the library itself.
3571 A workaround is available but breaks other compilers.
3573 </mark-expected-failures>
3576 <!-- numeric/conversion -->
3577 <library name="numeric/conversion">
3578 <test name="bounds_test">
3580 <toolset name="borland-5.6*"/>
3581 <toolset name="borland-5.8*"/>
3582 <toolset name="borland-5.9*"/>
3583 <note author="Fernando Cacciola" refid="3"/>
3586 <test name="converter_test">
3588 <toolset name="gcc-3.4.5_linux_x86_64"/>
3589 <toolset name="borland-5.6*"/>
3590 <toolset name="borland-5.8*"/>
3591 <toolset name="borland-5.9*"/>
3592 <note author="Fernando Cacciola" refid="3"/>
3595 <test name="traits_test">
3597 <toolset name="borland-5.6*"/>
3598 <toolset name="borland-5.8*"/>
3599 <toolset name="borland-5.9*"/>
3600 <note author="Fernando Cacciola" refid="3"/>
3603 <test name="udt_example_0">
3605 <toolset name="msvc-6.5_stlport4"/>
3606 <toolset name="borland-5.6*"/>
3607 <toolset name="borland-5.8*"/>
3608 <toolset name="borland-5.9*"/>
3609 <toolset name="msvc-6.5*"/>
3610 <note author="Fernando Cacciola" refid="30"/>
3613 <test name="udt_support_test">
3615 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
3616 <toolset name="borland-5.6*"/>
3617 <toolset name="borland-5.8*"/>
3618 <toolset name="borland-5.9*"/>
3619 <note author="Fernando Cacciola" refid="3"/>
3624 <!-- numeric/interval -->
3625 <library name="numeric/interval">
3627 <toolset name="borland-5.6*"/>
3628 <toolset name="msvc-6.5*"/>
3629 <toolset name="msvc-7.0"/>
3631 <mark-expected-failures>
3634 <test name="overflow"/>
3635 <toolset name="hp_cxx*"/>
3636 <note author="G. Melquiond">
3637 This test ensures the inclusion property of interval
3638 arithmetic is available for built-in floating-point types
3639 <code>float</code> and <code>double</code>. If the test
3640 fails, <code>interval<float></code> and
3641 <code>interval<double></code> should not be used
3642 on this compiler/platform since there will be no
3643 numerical guarantee.
3645 </mark-expected-failures>
3646 <mark-expected-failures>
3648 <test name="integer"/>
3649 <test name="overflow"/>
3650 <toolset name="borland-5.8*"/>
3651 <toolset name="borland-5.9*"/>
3652 <note author="A.Meredith">
3653 This compiler has some problems with name looup / overload resolution.
3655 </mark-expected-failures>
3656 <mark-expected-failures>
3657 <test name="cmp_exn"/>
3658 <test name="cmp_set"/>
3659 <test name="cmp_tribool"/>
3660 <toolset name="gcc-2.95.3-linux"/>
3661 <note author="Aleksey Gurtovoy" refid="2"/>
3662 </mark-expected-failures>
3663 <mark-expected-failures>
3665 <toolset name="cw-8.3*"/>
3666 <note author="Aleksey Gurtovoy" refid="2"/>
3667 </mark-expected-failures>
3668 <mark-expected-failures>
3669 <test name="test_float"/>
3670 <toolset name="msvc-7.1_stlport4"/>
3671 <note author="Vladimir Prus">
3672 This failure is unresearched. Presumably, the problem
3673 is that the abs function is not available in the "right"
3674 namespace with this compiler/stdlib combination.
3676 </mark-expected-failures>
3681 <!-- numeric/ublas -->
3682 <library name="numeric/ublas">
3684 <toolset name="borland-5.6*"/>
3685 <toolset name="borland-5.8*"/>
3686 <toolset name="borland-5.9*"/>
3687 <toolset name="gcc-3_3-darwin"/>
3688 <note author="M.Stevens" refid="17"/>
3691 <toolset name="cw-9.4"/>
3692 <note author="M.Stevens" refid="2"/>
3695 <toolset name="sun-5.8"/>
3696 <note author="M.Stevens" refid="4"/>
3699 <toolset name="cw-8.3"/>
3700 <toolset name="msvc-6.5*"/>
3701 <toolset name="msvc-7.0"/>
3702 <toolset name="iw-7_1-vc6"/>
3703 <toolset name="gcc-2.95*"/>
3704 <note author="M.Stevens" refid="30"/>
3707 <toolset name="msvc-7.1_stlport4"/>
3708 <note author="Roland Schwarz">
3709 This old version of the stlport library causes the BOOST_NO_STDC_NAMESPACE
3710 macro to be set. But this conflicts with the requirements of the library.
3713 <mark-expected-failures>
3714 <test name="test3"/>
3715 <toolset name="qcc-3.3.5*cpp"/>
3716 <note author="Jim Douglas" date="14 Feb 06" refid="27"/>
3717 </mark-expected-failures>
3718 <mark-expected-failures>
3719 <test name="begin_end"/>
3720 <toolset name="msvc-7.1"/>
3721 <note author="Gunter Winkler" date="07 Oct 09" refid="48"/>
3722 </mark-expected-failures>
3725 <!-- program_options -->
3726 <library name="program_options">
3728 <!-- Mark unusable toolsets -->
3730 <toolset name="gcc-2.95.3-linux"/>
3732 The failure is caused by standard library deficiencies
3733 -- it lacks the basic_string class template and
3734 the <locale> header.
3739 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
3740 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
3745 <toolset name="msvc-6.5*"/>
3750 <toolset name="msvc-7.0"/>
3754 <!-- Mark expected failures -->
3756 <test name="unicode_test*">
3758 <toolset name="iw-7_1-vc6"/>
3759 <toolset name="iw-7_1-vc6-stlp-4_5_3"/>
3760 <toolset name="msvc-6.5*"/>
3761 <note>The failures are caused by problems
3762 with std::locale implementation</note>
3766 <test name="options_description_test_dll">
3768 <toolset name="msvc-6.5"/>
3769 <toolset name="iw-7_1-vc6"/>
3774 <test name="variable_map_test_dll">
3776 <toolset name="iw-7_1-vc6"/>
3783 <toolset name="cw-8.3*"/>
3790 <toolset name="*como-4_3_3*"/>
3795 <mark-expected-failures>
3796 <test name="variable_map_test"/>
3797 <test name="variable_map_test_dll"/>
3798 <toolset name="msvc-6.5*"/>
3800 The failures are caused by compiler bug: it's not possible to
3801 explicitly pass template arguments to member template function. The
3802 failure is serious and makes one of the primary interfaces
3805 </mark-expected-failures>
3807 <mark-expected-failures>
3808 <test name="cmdline_test_dll"/>
3809 <test name="options_description_test_dll"/>
3810 <test name="parsers_test_dll"/>
3811 <test name="variable_map_test_dll"/>
3812 <test name="positional_options_test_dll"/>
3813 <toolset name="mingw-3*"/>
3814 <note author="Aleksey Gurtovoy" refid="29"/>
3815 </mark-expected-failures>
3817 <mark-expected-failures>
3818 <test name="unicode_test*"/>
3819 <toolset name="mingw-3*"/>
3820 <toolset name="gcc-3.4.2_mingw"/>
3821 <toolset name="gcc-3.4.5_mingw"/>
3822 <toolset name="gcc-mingw-3.4.5"/>
3823 <toolset name="gcc-mingw-3.4.2"/>
3824 <toolset name="gcc-cygwin-3.4.4"/>
3826 </mark-expected-failures>
3828 <mark-expected-failures>
3829 <test name="unicode_test_dll"/>
3830 <toolset name="*-darwin"/>
3831 <note refid="35" author="Doug Gregor"/>
3832 </mark-expected-failures>
3834 <mark-expected-failures>
3835 <test name="unicode_test*"/>
3836 <toolset name="qcc-3.3.5*gpp"/>
3837 <note author="Jim Douglas" date="12 Feb 06" refid="36"/>
3838 </mark-expected-failures>
3843 <library name="parameter">
3844 <mark-expected-failures>
3845 <test name="duplicates"/>
3846 <toolset name="borland-5.6*"/>
3847 <toolset name="borland-5.8*"/>
3848 <toolset name="borland-5.9*"/>
3849 <note refid="32" author="David Abrahams"/>
3850 </mark-expected-failures>
3851 <mark-expected-failures>
3852 <test name="unnamed_fail"/>
3853 <toolset name="msvc-7.0"/>
3854 <toolset name="msvc-6*"/>
3855 <note refid="32" author="David Abrahams"/>
3856 </mark-expected-failures>
3857 <test name="preprocessor">
3858 <toolset name="[Ss]un-5.8"/>
3860 Several compiler bugs were worked around in order to get
3861 this test to pass, so it could be considered to be only
3862 partially working. However, the library's macro system,
3863 which is really being tested here, does work on this
3864 compiler, which is why we worked around the failures.
3866 href="http://www.boost.org/libs/parameter/test/preprocessor.cpp">test's
3867 source file</a> for details.
3871 <mark-expected-failures>
3872 <test name="maybe"/>
3873 <toolset name="borland-5.6*"/>
3874 <toolset name="borland-5.8*"/>
3875 <toolset name="borland-5.9*"/>
3876 <toolset name="msvc-6*"/>
3877 <toolset name="msvc-7.0"/>
3878 <note refid="31" author="Daniel Wallin"/>
3879 </mark-expected-failures>
3881 <mark-expected-failures>
3882 <test name="python-parameter-test"/>
3883 <toolset name="borland-5.6*"/>
3884 <toolset name="borland-5.8*"/>
3885 <toolset name="borland-5.9*"/>
3886 <toolset name="msvc-6*"/>
3887 <toolset name="msvc-7.0"/>
3888 <note refid="31" author="Daniel Wallin"/>
3889 </mark-expected-failures>
3891 <mark-expected-failures>
3892 <test name="python_test"/>
3893 <toolset name="borland-5.6*"/>
3894 <toolset name="borland-5.8*"/>
3895 <toolset name="borland-5.9*"/>
3896 <toolset name="msvc-6*"/>
3897 <toolset name="msvc-7.0"/>
3898 <note refid="31" author="Daniel Wallin"/>
3899 </mark-expected-failures>
3901 <mark-expected-failures>
3902 <test name="optional_deduced_sfinae"/>
3903 <toolset name="msvc-6*"/>
3904 <toolset name="msvc-7.0"/>
3905 <toolset name="borland-5.6*"/>
3906 <toolset name="borland-5.8*"/>
3907 <toolset name="borland-5.9*"/>
3908 <note author="Daniel Wallin">
3909 These compilers do not support SFINAE, so are expected to
3912 </mark-expected-failures>
3914 <mark-expected-failures>
3915 <test name="preprocessor_deduced"/>
3916 <toolset name="borland-5.6*"/>
3917 <toolset name="borland-5.8*"/>
3918 <toolset name="borland-5.9*"/>
3919 <note author="Daniel Wallin">
3920 Borland does not support this feature. A compatibility syntax
3921 might be developed later on.
3923 </mark-expected-failures>
3925 <mark-expected-failures>
3926 <test name="normalized_argument_types"/>
3927 <toolset name="borland-5.6*"/>
3928 <toolset name="borland-5.8*"/>
3929 <toolset name="borland-5.9*"/>
3930 <toolset name="msvc-6*"/>
3931 <toolset name="msvc-7.0"/>
3932 <note author="Daniel Wallin">
3933 This feature generally requires advanced compiler
3934 features not supported by these compilers. It might
3935 be possible to work around the issue on VC6/7, but
3936 at this time no such workaround has been done.
3938 </mark-expected-failures>
3940 <mark-expected-failures>
3941 <test name="unnamed"/>
3943 <note author="Daniel Wallin">
3944 This is old and should not be tested any more.
3946 </mark-expected-failures>
3948 <mark-expected-failures>
3949 <test name="deduced_dependent_predicate"/>
3950 <toolset name="msvc-6*"/>
3951 <note refid="31" author="Daniel Wallin"/>
3952 </mark-expected-failures>
3953 <mark-expected-failures>
3954 <test name="optional_deduced_sfinae"/>
3955 <test name="preprocessor_deduced"/>
3956 <test name="python_test"/>
3957 <toolset name="sun-5.8"/>
3958 <note author="John Maddock">
3959 These test failure are reported to be under investigation
3960 at Sun's compiler labs.
3962 </mark-expected-failures>
3964 <mark-expected-failures>
3965 <test name="result_of"/>
3966 <toolset name="msvc-6*"/>
3967 <toolset name="msvc-7.0"/>
3968 <toolset name="borland-5.6*"/>
3969 <toolset name="borland-5.8*"/>
3970 <toolset name="borland-5.9*"/>
3971 <note refid="31" author="Daniel Wallin"/>
3972 </mark-expected-failures>
3974 <mark-expected-failures>
3975 <test name="python_test"/>
3976 <toolset name="qcc-3.3.5_gpp"/>
3977 <note refid="6" author="Daniel Wallin"/>
3978 </mark-expected-failures>
3980 <mark-expected-failures>
3981 <test name="sfinae"/>
3982 <toolset name="borland-5.8*"/>
3983 <toolset name="borland-5.9*"/>
3984 <toolset name="msvc-6.5_stlport4"/>
3985 <note refid="29" author="Daniel Wallin"/>
3986 </mark-expected-failures>
3988 <mark-expected-failures>
3989 <test name="basics"/>
3990 <test name="macros"/>
3991 <test name="maybe"/>
3992 <test name="sfinae"/>
3993 <toolset name="gcc-4.2.1*"/>
3994 <note author="Boris Gubenko" refid="42"/>
3995 </mark-expected-failures>
3999 <library name="property_tree">
4001 <toolset name="borland-*"/>
4002 <note author="Sebastian Redl">Inherited from MultiIndex</note>
4005 <toolset name="sun-5.*"/>
4006 <note author="Sebastian Redl">
4007 Lots of test failures complaining about the ambiguity of a
4008 const and a non-const overload of the same function.
4012 <toolset name="vacpp"/>
4013 <note author="Sebastian Redl">
4014 This compiler seems to have very broken name lookup.
4017 <mark-expected-failures>
4018 <test name="test_property_tree"/>
4019 <test name="test_json_parser"/>
4020 <toolset name="intel-darwin-*"/>
4021 <toolset name="intel-linux-*"/>
4022 <note author="Sebastian Redl">
4023 Tend to crash the compiler (Intel 10) or simply take too long
4026 </mark-expected-failures>
4027 <mark-expected-failures>
4028 <test name="test_xml_parser_rapidxml"/>
4029 <toolset name="gcc-3.4.3"/>
4030 <note author="Sebastian Redl">
4031 This ancient GCC doesn't like local const ints as template
4032 parameters. Or something like that.
4034 </mark-expected-failures>
4037 <!-- pointer container -->
4038 <library name="ptr_container">
4040 <toolset name="gcc-2.95.3*"/>
4041 <toolset name="sunpro-5_3-sunos"/>
4042 <toolset name="hp_cxx-65*"/>
4043 <toolset name="borland-5.6*"/>
4044 <toolset name="borland-5.8*"/>
4045 <toolset name="borland-5.9*"/>
4046 <toolset name="msvc-6.5*"/>
4047 <toolset name="msvc-7.0"/>
4048 <toolset name="dmc-8_47-stlport-4_5_3"/>
4049 <toolset name="hp_cxx-65_042_tru64"/>
4051 <mark-expected-failures>
4052 <test name="ptr_list"/>
4053 <toolset name="gcc-4.0.*"/>
4054 <note author="Thorsten Ottosen">
4055 The error is due to problems in the standard library implementation.
4056 It should be fixed in newer versions of the compiler.
4058 </mark-expected-failures>
4059 <mark-expected-failures>
4060 <test name="ptr_list"/>
4061 <toolset name="gcc-4.0.0*"/>
4062 <note author="Thorsten Ottosen">
4063 The error is due to problems in the standard library implementation.
4064 It should be fixed in newer versions of the compiler.
4066 </mark-expected-failures>
4067 <mark-expected-failures>
4068 <test name="incomplete_type_test"/>
4069 <toolset name="cw-9.4"/>
4070 <note author="Thorsten Ottosen">
4071 This error seems to be a bug the compiler. Please submit a
4074 </mark-expected-failures>
4075 <mark-expected-failures>
4076 <test name="iterator_test"/>
4077 <toolset name="gcc-3.2.3*"/>
4078 <toolset name="gcc-3.3.6*"/>
4079 <toolset name="gcc"/>
4080 <toolset name="qcc-3.3.5*"/>
4081 <note author="Thorsten Ottosen">
4082 This error seems to be a bug the standard library. Please submit a
4085 </mark-expected-failures>
4086 <mark-expected-failures>
4087 <test name="no_exceptions"/>
4088 <toolset name="cw-9.4"/>
4089 <toolset name="sun-5.*"/>
4090 <note author="Thorsten Ottosen">
4091 This test fails because the test ptr_vector fails. Please see the note
4094 </mark-expected-failures>
4095 <mark-expected-failures>
4096 <test name="ptr_deque"/>
4097 <toolset name="cw-9.4"/>
4098 <toolset name="sun-5.*"/>
4099 <note author="Thorsten Ottosen">
4100 For sun the problem is that <code>insert(iterator,range)</code>
4101 is not available due to partial ordering errors (the core library remains usable).
4102 For codewarrior the problem is at least <code>std::auto_ptr</code> overloads (the core library remains usable).
4104 </mark-expected-failures>
4105 <mark-expected-failures>
4106 <test name="ptr_list"/>
4107 <toolset name="cw-9.4"/>
4108 <toolset name="sun-5.*"/>
4109 <note author="Thorsten Ottosen">
4110 For sun the problem is that <code>insert(iterator,range)</code>
4111 is not available due to partial ordering errors (the core library remains usable).
4112 For codewarrior the problem is at least <code>std::auto_ptr</code> overloads (the core library remains usable).
4114 </mark-expected-failures>
4115 <mark-expected-failures>
4116 <test name="ptr_vector"/>
4117 <toolset name="cw-9.4"/>
4118 <toolset name="sun-5.8"/>
4119 <note author="Thorsten Ottosen">
4120 For sun the problem is that <code>insert(iterator,range)</code>
4121 is not available due to partial ordering errors (the core library remains usable).
4122 For codewarrior the problem is at least <code>std::auto_ptr</code> overloads (the core library remains usable).
4124 </mark-expected-failures>
4125 <mark-expected-failures>
4126 <test name="ptr_map"/>
4127 <toolset name="hp_cxx-71_006_tru64"/>
4128 <toolset name="cw-9.4"/>
4129 <toolset name="sun-5.*"/>
4130 <note author="Thorsten Ottosen">
4131 For hp, this compiler bug is insignificant.
4132 For sun the problem is that <code>transfer(range,ptr_map)</code>
4133 is not available due to partial ordering errors (the core library remains usable).
4134 For codewarrior the problem is not known so please submit a patch.
4136 </mark-expected-failures>
4137 <mark-expected-failures>
4138 <test name="ptr_set"/>
4139 <toolset name="cw-9.4"/>
4140 <toolset name="sun-5.*"/>
4141 <note author="Thorsten Ottosen">
4142 For sun the problem is that <code>transfer(range,ptr_map)</code> and
4143 <code>insert(range)</code>code>
4144 is not available due to partial ordering errors (the core library remains usable).
4145 For codewarrior the problem is at least <code>std::auto_ptr</code> overloads (the core library remains usable)..
4147 </mark-expected-failures>
4148 <mark-expected-failures>
4149 <test name="serialization"/>
4150 <toolset name="cw*"/>
4151 <toolset name="intel-darwin-*"/>
4152 <toolset name="intel-linux-*"/>
4153 <toolset name="pathscale-3.1"/>
4154 <toolset name="sun-5.*"/>
4155 <note author="Thorsten Ottosen">
4156 For codewarrior, the cause of this problem is unknown. Please
4157 submit a patch. Other failures are due to problems with
4158 the serialization library, or to a minor problem with the use of
4161 </mark-expected-failures>
4162 <mark-expected-failures>
4163 <test name="tree_test"/>
4164 <toolset name="sun-5.8"/>
4165 <note author="Thorsten Ottosen">
4166 For sun the problem is due to Boost.Test.
4168 </mark-expected-failures>
4169 <mark-expected-failures>
4171 <toolset name="cw-9.4"/>
4172 <note author="Thorsten Ottosen">
4173 Seem like a bug in the compiler. Please submit a patch.
4175 </mark-expected-failures>
4176 <mark-expected-failures>
4177 <test name="view_example"/>
4178 <toolset name="cw-9.4"/>
4179 <note author="Thorsten Ottosen">
4180 Seem like a bug in the compiler. Please submit a patch.
4182 </mark-expected-failures>
4186 <library name="python">
4188 <toolset name="borland-5.5*"/>
4189 <toolset name="borland-5.6*"/>
4190 <toolset name="borland-5.8*"/>
4191 <toolset name="borland-5.9*"/>
4196 <toolset name="hp_cxx-65*"/>
4197 <note author="Markus Schoepflin">
4198 The library fails to compile because of an error in the C++
4199 standard library implementation on this platform. It incorrectly
4200 assumes that fpos_t is of an integral type, which is not always
4201 the case. This is fixed in a later release.
4205 <toolset name="sun-5.6*"/>
4206 <note author="David Abrahams">
4207 The old reasoning given for this markup, which applied
4208 to sun-5.8*, was as follows. However, tuple's tests
4209 seem to use the test library, which is apparently
4210 completely broken on Sun. Therefore, I've backed off
4211 the version number to sun-5.6 so I can see the actual
4212 state of the failures.
4214 <blockquote>This compiler seems to be having trouble digesting
4215 Boost.Tuple. Until it can handle Boost.Tuple there's
4216 little chance it will handle Boost.Python</blockquote>
4219 <mark-expected-failures>
4220 <test name="object"/>
4221 <toolset name="intel-10.*"/>
4222 <note author="David Abrahams">
4224 This compiler has a bug that causes silent misbehavior at runtime
4225 when each of an assignment expression follows one of the following patterns:
4226 <em>expr</em><code>.attr(</code><em>name</em><code>)</code>
4227 or <em>expr</em><code>[</code><em>item</em><code>]</code>,
4229 is-a <code>boost::python::object</code>. We've been
4230 unable to find a workaround.
4233 </mark-expected-failures>
4234 <mark-expected-failures>
4235 <test name="object"/>
4236 <toolset name="acc"/>
4237 <note author="Boris Gubenko" refid="46"/>
4238 </mark-expected-failures>
4239 <mark-expected-failures>
4241 <test name="auto_ptr"/>
4242 <test name="builtin_convertors"/>
4243 <test name="callbacks"/>
4244 <test name="crossmod_exception"/>
4245 <test name="data_members"/>
4247 <test name="exception_translator"/>
4248 <test name="extract"/>
4249 <test name="implicit"/>
4250 <test name="iterator"/>
4252 <test name="map_indexing_suite"/>
4253 <test name="object"/>
4254 <test name="opaque"/>
4255 <test name="pickle2"/>
4256 <test name="polymorphism"/>
4257 <test name="polymorphism2"/>
4258 <test name="shared_ptr"/>
4259 <test name="slice"/>
4260 <test name="test_pointer_adoption"/>
4262 <test name="vector_indexing_suite"/>
4263 <test name="virtual_functions"/>
4264 <toolset name="gcc-2.95.3-linux"/>
4265 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
4266 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
4267 <note author="D. Abrahams">
4268 The problems with GCC 2.x only occur when C++ exceptions are thrown and
4269 the framework catches them, which happens quite often in the tests.
4270 So technically GCC 2.x is usable if you're careful.
4272 </mark-expected-failures>
4273 <mark-expected-failures>
4274 <test name="pointer_vector"/>
4275 <test name="polymorphism"/>
4276 <toolset name="hp_cxx*"/>
4277 <note author="Markus Schoepflin" refid="29"/>
4278 </mark-expected-failures>
4279 <mark-expected-failures>
4280 <test name="data_members"/>
4281 <toolset name="acc"/>
4282 <note author="Boris Gubenko">
4283 This test assumes standard-compliant dependent template name lookup which
4284 is performed by aCC6 only in strict ansi mode. Since on HP-UX/aCC6 boost
4285 testing is done in the default compilation mode, this test fails to
4286 compile on this platform (in strict ansi mode, it compiles and succeeds).
4288 </mark-expected-failures>
4292 <library name="random">
4294 <toolset name="msvc"/>
4295 <toolset name="msvc-7.0"/>
4296 <note author="B. Dawes" refid="10"/>
4298 <test name="random_test">
4300 <toolset name="cw-8.3*"/>
4301 <note author="B. Dawes" refid="3"/>
4304 <toolset name="borland-5.6*"/>
4305 <toolset name="borland-5.8*"/>
4306 <toolset name="borland-5.9*"/>
4307 <note author="B. Dawes" refid="2"/>
4310 <toolset name="intel-vc71-win*"/>
4311 <toolset name="intel-vc8-win*"/>
4312 <note author="S. Slapeta" refid="1"/>
4315 <toolset name="intel-linux-9.0"/>
4316 <note author="John Maddock">
4317 Reported to Intel as issue 409291, and confirmed
4318 as a problem. Probably this relates to a specific
4319 Linux-Kernal or GLibC version.
4323 <toolset name="qcc-3.3.5*"/>
4324 <note author="Jim Douglas" date="13 Feb 06">
4325 Test fails with ranlux*_O1 RNGs when saving and recalling the state due to a bug in the
4326 double to string conversion. The problem has been reported to QNX as PR29252.
4330 <toolset name="gcc-*_tru64"/>
4331 <note author="Markus Schoepflin">
4332 This test fails because of limitations in the system assembler
4333 version used by GCC. It most probably would pass if the test
4334 were split into multiple source files.
4338 <toolset name="gcc-3.4.6_linux_ia64"/>
4339 <note author="Boris Gubenko">
4340 It looks like a compiler issue: the test fails with gcc 3.4.6
4341 and succeeds with gcc 4.2.1.
4345 <mark-expected-failures>
4346 <test name="test_ecuyer1988"/>
4347 <test name="test_hellekalek1995"/>
4348 <test name="test_rand48"/>
4349 <test name="test_minstd_rand0"/>
4350 <test name="test_minstd_rand"/>
4351 <test name="test_kreutzer1986"/>
4352 <test name="test_mt11213b"/>
4353 <test name="test_mt19937"/>
4354 <test name="test_lagged_fibonacci"/>
4355 <test name="test_lagged_fibonacci607"/>
4356 <test name="test_ranlux3"/>
4357 <test name="test_ranlux4"/>
4358 <test name="test_ranlux3_01"/>
4359 <test name="test_ranlux4_01"/>
4360 <test name="test_ranlux64_3_01"/>
4361 <test name="test_ranlux64_4_01"/>
4362 <test name="test_taus88"/>
4363 <toolset name="gcc-mingw-4.3.3"/>
4364 <note refid="5" author="Steven Watanabe"/>
4365 </mark-expected-failures>
4369 <library name="range">
4371 <toolset name="borland-*"/>
4372 <toolset name="mipspro"/>
4373 <toolset name="dmc-8_43-stlport-4_5_3"/>
4374 <toolset name="gcc-2.95.3*"/>
4375 <toolset name="sunpro-5_3-sunos"/>
4377 <mark-expected-failures>
4378 <test name="array"/>
4379 <toolset name="como-4_3_3*"/>
4380 <toolset name="sun-5.8"/>
4381 <toolset name="borland-5.6*"/>
4382 <toolset name="borland-5.8*"/>
4383 <toolset name="borland-5.9*"/>
4384 <note refid="27" author="Thorsten Ottosen"/>
4385 </mark-expected-failures>
4386 <mark-expected-failures>
4387 <test name="adl_conformance"/>
4388 <toolset name="msvc-7.1"/>
4389 <note author="Neil Groves">
4390 This test is designed to give users visibility of the ADL problems with their compilers.
4391 Lack of Argument Dependent Lookup changes how one can extend the library. The lack of
4392 ADL is worked-around internally so that most of the functionality is preserved.
4394 </mark-expected-failures>
4395 <mark-expected-failures>
4396 <test name="iterator_range"/>
4397 <toolset name="msvc-stlport"/>
4398 <toolset name="msvc-6.5_stlport4"/>
4399 <toolset name="hp_cxx-65*"/>
4400 <note author="Thorsten Ottosen">
4401 For most compilers this is due to problems
4402 with built-in arrays (notably char arrays) and operator==()
4403 and operator!=() for iterator_range. Thus, not using built-in arrays
4406 For other compilers it is simply a bug in the standard library.
4408 </mark-expected-failures>
4409 <mark-expected-failures>
4410 <test name="push_front"/>
4411 <toolset name="msvc-9.0~stlport5.2"/>
4412 <note author="Neil Groves">
4413 push_front fails the unit test in this configuration. I do not have this
4414 configuration available to determine if a work-around is possible.
4416 </mark-expected-failures>
4417 <mark-expected-failures>
4418 <test name="reversible_range"/>
4419 <toolset name="hp_cxx-65*"/>
4420 <note author="Thorsten Ottosen">
4421 This test probably fails because it uses built-in arrays. So do expect these
4422 functions to work in normal code.
4424 </mark-expected-failures>
4425 <mark-expected-failures>
4426 <test name="string"/>
4427 <toolset name="hp_cxx-65*"/>
4428 <toolset name="sun-5.8"/>
4429 <toolset name="borland-5.6*"/>
4430 <toolset name="borland-5.8*"/>
4431 <toolset name="borland-5.9*"/>
4432 <note author="Thorsten Ottosen">
4433 The string functionality is expected to work if
4434 the user employs std::string and stays away from built-in
4437 </mark-expected-failures>
4438 <mark-expected-failures>
4439 <test name="sub_range"/>
4440 <toolset name="msvc-8.0"/>
4441 <toolset name="intel-vc8-*"/>
4442 <toolset name="iw-7_1-vc6-stlp-4_5_3"/>
4443 <toolset name="msvc-6.5_stlport4"/>
4444 <toolset name="msvc-7.0"/>
4445 <toolset name="msvc-7.1_stlport4"/>
4446 <toolset name="hp_cxx-65*"/>
4447 <note refid="6" author="Thorsten Ottosen">
4448 For most compilers this is due to problems
4449 with built-in arrays (notably char arrays) and operator==()
4450 and operator!=() for iterator_range. Thus, not using built-in arrays
4453 </mark-expected-failures>
4454 <mark-expected-failures>
4455 <test name="sub_range"/>
4456 <toolset name="cw-9_5-darwin"/>
4457 <note author="Thorsten Ottosen">
4458 At the time of release I couldn't figure out why this was failing.
4459 Anyway, the failure is not very important; also, the well-definedness of
4460 "singularity" of an iterator range is likely to change.
4462 </mark-expected-failures>
4463 <mark-expected-failures>
4464 <test name="extension_mechanism"/>
4465 <toolset name="msvc-6.5*"/>
4466 <toolset name="msvc-7.0"/>
4467 <note author="Thorsten Ottosen">
4468 The test requires support for Argument Dependent Lookup (ADL)
4469 which the compiler in question does not provide.
4471 </mark-expected-failures>
4475 <library name="ratio">
4477 <toolset name="borland-*"/>
4478 <toolset name="pathscale*"/>
4479 <note author="Vicente J. Botet Escriba">
4480 The compiler does not support features that are essential for the library.
4483 <mark-expected-failures>
4484 <test name="mpl_plus_pass"/>
4485 <test name="ratio_ext_pass"/>
4486 <toolset name="intel-*-10.1"/>
4487 <toolset name="intel-*-11.0"/>
4488 <toolset name="intel-*-11.1"/>
4489 <note author="Vicente J. Botet Escriba">
4490 internal error: assertion failed: copy_template_param_expr.
4491 It seems to be already an Intel internal bug database (tracker number #82149).
4492 I have no workaround yet.
4494 </mark-expected-failures>
4498 <library name="regex">
4499 <test name="regex_token_iterator_eg_2">
4501 <toolset name="msvc-6.5"/>
4502 <note author="J. Maddock"/>
4505 <test name="posix_api_check">
4507 <toolset name="como-4_3_3-vc7*"/>
4508 <note author="J. Maddock"/>
4511 <test name="wide_posix_api_check">
4513 <toolset name="qcc-3.3.5_gpp"/>
4514 <note author="J. Maddock">
4515 No Wide character support on this platform.
4519 <test name="wide_posix_api_check_c">
4521 <toolset name="qcc-3.3.5_gpp"/>
4522 <note author="J. Maddock">
4523 No Wide character support on this platform.
4529 <toolset name="*como-4_3_3*"/>
4530 <note author="J. Maddock">
4531 This test requires features that are unsupported by Como:
4532 use and building of dll's mainly.
4536 <mark-expected-failures>
4537 <test name="static_mutex_test"/>
4538 <test name="test_grep"/>
4539 <toolset name="*como-4_3_3*"/>
4540 <note author="J. Maddock">
4541 This test requires features that are unsupported by Como:
4542 use and building of dll's mainly.
4544 </mark-expected-failures>
4545 <mark-expected-failures>
4546 <test name="regex_regress_threaded"/>
4547 <toolset name="darwin*"/>
4548 <note author="J. Maddock">
4549 This tests fails because a dependency (Boost.Test)
4550 fails to initialise correctly. The issue has been
4551 reported to the library's author.
4553 </mark-expected-failures>
4554 <mark-expected-failures>
4555 <test name="regex_regress_threaded"/>
4556 <toolset name="gcc-*_tru64"/>
4557 <note author="J. Maddock">
4558 GCC on tru64 appears not to cope with C++ exceptions
4559 thrown from within threads.
4561 </mark-expected-failures>
4562 <test name="concept_check">
4564 <toolset name="msvc-8.0"/>
4565 <toolset name="sunpro-5_3-sunos"/>
4566 <toolset name="sun-5.8"/>
4567 <toolset name="borland-5.8*"/>
4568 <toolset name="borland-5.9*"/>
4569 <toolset name="qcc-3.3.5_cpp"/>
4570 <note author="John Maddock" refid="2"/>
4573 <test name="test_grep">
4575 <toolset name="gcc-2.95.3-linux"/>
4576 <toolset name="sunpro-5_3-sunos"/>
4577 <toolset name="sun-5.8"/>
4578 <toolset name="msvc-6.5*"/>
4579 <toolset name="msvc-7.0"/>
4580 <note author="John Maddock">
4581 This test fails because a dependency (Boost.Program Options) doesn't build with this compiler.
4585 <test name="test_grep">
4587 <toolset name="borland-5.9*"/>
4588 <note author="A.Meredith">
4589 This test fails because a dependency (Boost.Program Options) which currently doesn't build with this compiler.
4593 <mark-expected-failures>
4594 <test name="regex_regress"/>
4595 <test name="regex_regress_dll"/>
4596 <toolset name="iw-7_1-vc6-stlp-4_5_3"/>
4597 <note author="John Maddock" refid="29"/>
4598 </mark-expected-failures>
4599 <mark-expected-failures>
4600 <test name="regex_regress"/>
4601 <test name="regex_regress_threaded"/>
4602 <test name="regex_regress_dll"/>
4603 <toolset name="borland*"/>
4604 <note author="John Maddock">
4605 There appears to be a linker bug that prevents these
4606 projects from building, see http://qc.borland.com/wc/qcmain.aspx?d=32020.
4608 </mark-expected-failures>
4609 <mark-expected-failures>
4610 <test name="unicode_iterator_test"/>
4611 <toolset name="borland-5.6*"/>
4612 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
4613 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
4614 <note author="John Maddock" refid="6"/>
4615 </mark-expected-failures>
4616 <mark-expected-failures>
4617 <test name="regex_regress"/>
4618 <test name="regex_regress_threaded"/>
4619 <test name="regex_regress_dll"/>
4620 <toolset name="borland*"/>
4621 <note author="John Maddock">
4622 There appears to be a linker bug that prevents these
4623 projects from building, see http://qc.borland.com/wc/qcmain.aspx?d=32020.
4625 </mark-expected-failures>
4626 <mark-expected-failures>
4627 <test name="regex_timer"/>
4628 <toolset name="msvc-6.5_stlport4"/>
4629 <note author="John Maddock">
4630 Test fails due to unresilved externals from STLport: appears to be
4631 an STLport bug. </note>
4632 </mark-expected-failures>
4633 <mark-expected-failures>
4634 <test name="regex_regress_threaded"/>
4635 <test name="static_mutex_test"/>
4636 <toolset name="msvc-6.5_stlport*"/>
4637 <toolset name="msvc-7.1_stlport*"/>
4638 <toolset name="msvc-8.0"/>
4639 <toolset name="gcc-cygwin*"/>
4640 <note author="John Maddock">
4641 These tests pass when run directly from the command line,
4642 but fail when run under the regression test script.
4643 The issue has never been fully pinned down, but appears
4644 to be related to how long the tests take to run.</note>
4645 </mark-expected-failures>
4649 <library name="scope_exit">
4650 <mark-expected-failures>
4651 <test name="emulation_tpl"/>
4652 <toolset name="intel-*-9.1"/>
4653 <toolset name="intel-*-10.0"/>
4654 <toolset name="intel-*-10.1"/>
4655 <toolset name="intel-*-11.0"/>
4656 <note author="Alexander Nasonov">
4657 The test does not compile in typeof emulation mode,
4658 most likely due to a compiler bug. Users are advised
4659 to use native typeof.
4661 </mark-expected-failures>
4662 <mark-expected-failures>
4663 <test name="native"/>
4664 <toolset name="acc"/>
4665 <note author="Alexander Nasonov" refid="39"/>
4666 </mark-expected-failures>
4667 <mark-expected-failures>
4668 <test name="native_tpl"/>
4669 <toolset name="acc"/>
4670 <note author="Alexander Nasonov" refid="39"/>
4671 </mark-expected-failures>
4672 <mark-expected-failures>
4673 <test name="native_tu_test"/>
4674 <toolset name="acc"/>
4675 <note author="Alexander Nasonov" refid="39"/>
4676 </mark-expected-failures>
4677 <mark-expected-failures>
4678 <test name="native_tu_test"/>
4679 <toolset name="msvc-7.1*"/>
4680 <toolset name="msvc-8.0*"/>
4681 <note author="Alexander Nasonov" refid="2"/>
4682 </mark-expected-failures>
4683 <mark-expected-failures>
4684 <test name="emulation_tu_test"/>
4685 <toolset name="msvc-7.1*"/>
4686 <toolset name="msvc-8.0*"/>
4687 <note author="Alexander Nasonov" refid="2"/>
4688 </mark-expected-failures>
4692 <library name="signals">
4694 <toolset name="sunpro-5_3-sunos"/>
4696 <test name="signal_test">
4698 <toolset name="cw-8.3*"/>
4699 <note author="B. Dawes" refid="2"/>
4702 <toolset name="borland-5.6*"/>
4703 <toolset name="borland-5.8*"/>
4704 <toolset name="borland-5.9*"/>
4705 <toolset name="msvc-6.5"/>
4706 <toolset name="msvc-7.0"/>
4707 <note author="B. Dawes" refid="3"/>
4713 <library name="statechart">
4715 <toolset name="borland*"/>
4716 <toolset name="sun-5*"/>
4717 <toolset name="pgi-1*"/>
4718 <note author="Andreas Huber" refid="17"/>
4720 <mark-expected-failures>
4721 <test name="DllTestNormal"/>
4722 <toolset name="*mingw*"/>
4723 <toolset name="intel-darwin-10*"/>
4724 <toolset name="intel-darwin-11*"/>
4725 <note author="Andreas Huber">
4726 A runtime failure of this test indicates that this platform
4727 <b>dynamically</b> links code in a manner such that under
4728 certain circumstances more than one instance of a
4729 header-defined static class member can exist at runtime. See
4730 <a href="http://www.boost.org/libs/statechart/doc/faq.html#Dll">FAQ</a>
4731 for more information.
4733 </mark-expected-failures>
4734 <mark-expected-failures reason="?">
4735 <test name="FifoSchedulerTest*"/>
4736 <toolset name="darwin-4.4"/>
4737 <toolset name="intel-darwin-10*"/>
4738 <note author="Andreas Huber" refid="29"/>
4739 </mark-expected-failures>
4740 <mark-expected-failures reason="?">
4741 <test name="TransitionTest*"/>
4742 <toolset name="intel-darwin-10*"/>
4743 <toolset name="intel-linux-10*"/>
4744 <toolset name="gcc-4.3.4_0x"/>
4745 <toolset name="pathscale-3.2"/>
4746 <note author="Andreas Huber" refid="29"/>
4747 </mark-expected-failures>
4750 <!-- static_assert -->
4751 <library name="static_assert">
4752 <test name="static_assert_example_2">
4754 <toolset name="sunpro-5_3-sunos"/>
4755 <note author="J. Maddock" refid="4"/>
4758 <test name="static_assert_test">
4760 <toolset name="borland-5.6*"/>
4761 <toolset name="borland-5.8*"/>
4762 <toolset name="borland-5.9*"/>
4763 <note author="A.Meredith" date="26 May 2006">
4764 This test runs without problem on Borland compilers,
4765 which means the static assertion is not being caught.
4772 <library name="system">
4774 <toolset name="borland-5.6*"/>
4775 <toolset name="borland-5.8*"/>
4776 <note author="Beman Dawes">
4777 This compiler does not support enable_if, which is required
4784 <library name="test">
4785 <mark-expected-failures>
4786 <test name="ifstream_line_iterator_test"/>
4787 <toolset name="sunpro*"/>
4788 <note author="Gennadiy Rozental" refid="2"/>
4789 </mark-expected-failures>
4791 <mark-expected-failures>
4792 <test name="custom_exception_test"/>
4793 <toolset name="msvc-6.5*"/>
4794 <note author="Gennadiy Rozental" refid="2"/>
4795 </mark-expected-failures>
4797 <mark-expected-failures>
4798 <test name="errors_handling_test"/>
4799 <toolset name="*como-4_3_3*"/>
4800 <note author="B. Dawes" refid="3"/>
4801 </mark-expected-failures>
4803 <mark-expected-failures>
4804 <test name="token_iterator_test"/>
4805 <toolset name="msvc-6.5*"/>
4806 <toolset name="iw-7_1-vc6"/>
4807 <toolset name="msvc-7.0"/>
4808 <toolset name="msvc-7.0-stlport"/>
4809 <toolset name="gcc-2.95.3-linux"/>
4810 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
4811 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
4812 <toolset name="hp_cxx-65*"/>
4813 <toolset name="sunpro*"/>
4814 <toolset name="borland*"/>
4815 <note author="Gennadiy Rozental" refid="3"/>
4816 </mark-expected-failures>
4818 <mark-expected-failures>
4819 <test name="token_iterator_test"/>
4820 <toolset name="qcc-3.3.5*gpp"/>
4821 <note author="Jim Douglas" date="14 Feb 06" refid="36"/>
4822 </mark-expected-failures>
4824 <mark-expected-failures>
4825 <test name="test_fp_comparisons"/>
4826 <toolset name="msvc-6.5*"/>
4827 <toolset name="msvc-7.0"/>
4828 <toolset name="msvc-7.0-stlport"/>
4829 <toolset name="borland-5.6*"/>
4830 <toolset name="borland-5.8*"/>
4831 <toolset name="borland-5.9*"/>
4832 <note author="Gennadiy Rozental" refid="3"/>
4833 </mark-expected-failures>
4835 <mark-expected-failures reason="?">
4836 <test name="basic_cstring_test"/>
4837 <toolset name="gcc-2.95.3-linux"/>
4839 </mark-expected-failures>
4841 <mark-expected-failures>
4842 <test name="test_tools_test"/>
4843 <toolset name="cw-9_5-darwin"/>
4844 <note refid="29" author="Doug Gregor"/>
4845 </mark-expected-failures>
4847 <mark-expected-failures>
4848 <test name="errors_handling_test"/>
4849 <toolset name="acc*"/>
4850 <toolset name="cw-9_5-darwin"/>
4851 <note refid="29" author="Doug Gregor"/>
4852 </mark-expected-failures>
4854 <mark-expected-failures>
4855 <test name="test_tools_test"/>
4856 <toolset name="cw-9.4"/>
4857 <note refid="29" author="Doug Gregor"/>
4858 </mark-expected-failures>
4860 <mark-expected-failures>
4861 <test name="prg_exec_fail2"/>
4862 <toolset name="gcc-3.4.2_hpux_pa_risc"/>
4863 <toolset name="gcc-3.4.6_linux_ia64"/>
4864 <note author="Vladimir Prus">
4865 The test verifies that Boost.Test detects division by
4866 zero. It fails on PowerPC, PA-RISC and Linux ia64. On PowerPC
4867 processors, division has an undefined result. The compiler
4868 has to emit extra code to assert that the divisor isn't zero.
4870 Compiler options -fno-trapping-math and -fnon-call-exceptions
4871 might affect this. However, in default configuration
4872 no check is done, and division by zero is not detected.
4874 </mark-expected-failures>
4876 <mark-expected-failures>
4877 <test name="prg_exec_fail3"/>
4878 <toolset name="cw-9.4"/>
4879 <toolset name="gcc-3.4.6_linux_ia64"/>
4880 <note author="Vladimir Prus">
4881 The test appears to test that failed assertion result
4882 in non-zero exit status. That seems to be not the
4883 case, for unknown reasons.
4885 </mark-expected-failures>
4887 <mark-expected-failures>
4888 <test name="sync_access_test"/>
4889 <toolset name="acc*"/>
4890 <toolset name="gcc-4.2.1_hpux_ia64"/>
4891 <note author="Boris Gubenko">
4892 On HP-UX platform, this test must be compiled/linked in multithread mode.
4893 When compiled/linked with aC++ with -mt, it succeeds. When compiled/linked
4894 with GCC with -pthread, it links cleanly but fails in run-time.
4896 </mark-expected-failures>
4902 <library name="thread">
4904 <toolset name="*como-4_3_3*"/>
4905 <note author="B. Dawes" refid="10"/>
4907 <test name="test_mutex">
4909 <toolset name="msvc-7.0"/>
4910 <note author="B. Dawes" refid="0"/>
4911 <note author="B. Dawes" refid="6"/>
4914 <mark-expected-failures>
4915 <test name="*_lib"/>
4916 <toolset name="intel-8.0-linux*"/>
4917 <note author="Aleksey Gurtovoy">
4918 This failure is caused by a conflict between the compiler
4919 and the testing environment: the tests are run on a platform with
4920 <i>too recent</i> version of glibc, which is not currently
4921 supported by the compiler vendor (Intel).
4923 If you are having the same problem and <i>really</i> want to make
4924 things work, renaming <code>strol</code> symbol in the
4925 compiler's static runtime library (<code>libcprts.a</code>) to
4926 something else is known to resolve the issue.
4928 </mark-expected-failures>
4929 <mark-expected-failures reason="?">
4930 <test name="*_lib"/>
4931 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
4932 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
4933 <note author="Aleksey Gurtovoy" refid="29"/>
4934 </mark-expected-failures>
4936 It is unclear why this has been marked as expected failure. The
4937 pthread_timedwait is giving an error code of EINVAL, which needs to
4938 be resolved, since the timed behaviour is affected by this bug.
4939 Marked as a failure again by Roland Schwarz, 2007-01-12
4940 <mark-expected-failures>
4941 <test name="test_mutex"/>
4942 <test name="test_mutex_lib"/>
4943 <toolset name="qcc-3.3*"/>
4944 <note author="Jim Douglas" date="13 Feb 06" refid="16"/>
4945 </mark-expected-failures>
4947 <mark-expected-failures>
4948 <test name="test_tss_lib"/>
4949 <toolset name="borland-*"/>
4950 <toolset name="como-win-*"/>
4951 <toolset name="msvc*wm5*"/>
4952 <toolset name="cw-9.4"/>
4953 <note author="Roland Schwarz" date="2006-12-14">
4954 When a thread ends, tss data needs to be cleaned up. This process
4955 is mostly automatic. When threads are launched by the Boost.Thread API
4956 cleanup is handled by the library implementation. For threads, launched
4957 by the native operating system API it is not possible to get this cleanup
4958 on every compiler/platform. A warning (error) will be present in this case,
4959 which cleary states this fact. It is recommended to start threads only
4960 by means of the Boost.Thread API if you need to avoid the leaks that appear
4961 on the end of the thread. If this is not possible the cleanup can be invoked
4962 from user code before the process actually ends. For library implementors
4963 this means to call these functions during library initialization and
4966 </mark-expected-failures>
4967 <mark-expected-failures>
4968 <test name="test_thread_move"/>
4969 <test name="test_thread_move_lib"/>
4970 <test name="test_move_function"/>
4971 <test name="test_move_function_lib"/>
4972 <toolset name="acc"/>
4973 <toolset name="borland-*"/>
4974 <toolset name="sun-*"/>
4975 <note author="Anthony Williams" date="2007-12-14">
4976 The Borland compiler and HP-UX aC++ compiler in default mode fail to bind rvalues to the thread move constructor,
4977 choosing instead to bind them to the private (and unimplemented) copy constructor.
4978 With aC++, the tests compile cleanly in strict ansi mode and succeed.
4980 </mark-expected-failures>
4981 <mark-expected-failures>
4982 <test name="test_thread_move_return"/>
4983 <test name="test_thread_move_return_lib"/>
4984 <test name="test_thread_return_local"/>
4985 <test name="test_thread_return_local_lib"/>
4987 <note author="Anthony Williams" date="2009-10-28">
4988 These tests will fail in most compilers that don't support rvalue references.
4990 </mark-expected-failures>
4994 <library name="tuple">
4996 <toolset name="sunpro-5_3-sunos"/>
5000 <!-- type_traits -->
5001 <library name="type_traits">
5002 <mark-expected-failures>
5003 <test name="has_operator_new_test"/>
5004 <test name="make_signed_test"/>
5005 <test name="make_unsigned_test"/>
5006 <toolset name="msvc-7.1"/>
5007 <note author="John Maddock">
5008 Apparently the compiler can't cope with these - later versions are fine though.
5009 Probably work-round-able if someone would care to look into these.
5011 </mark-expected-failures>
5012 <mark-expected-failures>
5013 <test name="function_traits_test"/>
5014 <test name="remove_bounds_test"/>
5015 <test name="remove_const_test"/>
5016 <test name="remove_cv_test"/>
5017 <test name="remove_pointer_test"/>
5018 <test name="remove_reference_test"/>
5019 <test name="remove_volatile_test"/>
5020 <test name="decay_test"/>
5021 <test name="extent_test"/>
5022 <test name="remove_extent_test"/>
5023 <test name="remove_all_extents_test"/>
5024 <test name="rank_test"/>
5025 <test name="is_unsigned_test"/>
5026 <toolset name="msvc-6.5*"/>
5027 <toolset name="msvc-7.0"/>
5028 <note author="Aleksey Gurtovoy">
5029 This failure is caused by the lack of compiler support for class template
5030 partial specialization. A limited subset of the tested functionality is
5031 available on the compiler through a user-side workaround (see
5032 <a href="http://www.boost.org/libs/type_traits/index.html#transformations">
5033 http://www.boost.org/libs/type_traits/index.html#transformations</a> for
5036 </mark-expected-failures>
5037 <mark-expected-failures>
5038 <test name="decay_test"/>
5039 <test name="extent_test"/>
5040 <test name="is_base_and_derived_test"/>
5041 <test name="is_base_of_test"/>
5042 <test name="is_convertible_test"/>
5043 <test name="rank_test"/>
5044 <test name="remove_all_extents_test"/>
5045 <test name="remove_bounds_test"/>
5046 <test name="remove_const_test"/>
5047 <test name="remove_extent_test"/>
5048 <test name="remove_pointer_test"/>
5049 <test name="remove_volatile_test"/>
5050 <test name="tricky_add_pointer_test"/>
5051 <test name="tricky_function_type_test"/>
5052 <test name="tricky_incomplete_type_test"/>
5053 <test name="make_signed_test"/>
5054 <test name="make_unsigned_test"/>
5055 <toolset name="borland-5.6*"/>
5056 <toolset name="borland-5.8*"/>
5057 <toolset name="borland-5.9*"/>
5058 <note author="John Maddock" refid="2"/>
5059 </mark-expected-failures>
5060 <mark-expected-failures>
5061 <test name="promote_basic_test"/>
5062 <test name="promote_enum_test"/>
5063 <test name="promote_mpl_test"/>
5064 <test name="tricky_partial_spec_test"/>
5065 <toolset name="borland-5.6*"/>
5066 <toolset name="borland-5.8*"/>
5067 <toolset name="borland-5.9*"/>
5068 <note author="AlisdairM" refid="2"/>
5069 </mark-expected-failures>
5070 <mark-expected-failures>
5071 <test name="promote_enum_msvc_bug_test"/>
5072 <toolset name="msvc-7.1*"/>
5073 <toolset name="msvc-8.0*"/>
5074 <toolset name="msvc-9.0*"/>
5075 <note author="Alexander Nasonov">
5076 See bug 99776 'enum UIntEnum { value = UINT_MAX } is promoted to int'
5077 http://lab.msdn.microsoft.com/ProductFeedback/viewfeedback.aspx?feedbackid=22b0a6b7-120f-4ca0-9136-fa1b25b26efe
5079 </mark-expected-failures>
5080 <test name="tricky_is_enum_test">
5082 <toolset name="borland-5.6*"/>
5083 <toolset name="borland-5.8*"/>
5084 <toolset name="borland-5.9*"/>
5085 <toolset name="msvc-6.5*"/>
5086 <toolset name="gcc-2.95.3-*"/>
5089 <test name="tricky_incomplete_type_test">
5091 <toolset name="iw-7_1*"/>
5092 <note author="John Maddock" refid="2"/>
5095 <test name="is_abstract_test">
5097 <toolset name="borland-5.6*"/>
5098 <toolset name="borland-5.8*"/>
5099 <toolset name="borland-5.9*"/>
5100 <toolset name="cw-8.3*"/>
5101 <toolset name="cw-9.3*"/>
5102 <toolset name="cw-9.4"/>
5103 <toolset name="cw-9.5"/>
5104 <toolset name="msvc-6.5*"/>
5105 <toolset name="msvc-7.0"/>
5106 <toolset name="mingw-3_3*"/>
5107 <toolset name="gcc-2*"/>
5108 <toolset name="gcc-3.2*"/>
5109 <toolset name="gcc-3.3*"/>
5110 <toolset name="qcc-3.3*"/>
5111 <toolset name="sunpro-5_3-sunos"/>
5112 <toolset name="hp_cxx-65*"/>
5113 <toolset name="darwin"/>
5114 <toolset name="mingw"/>
5115 <note author="Aleksey Gurtovoy">
5116 This functionality is available only on compilers that implement C++ Core Language
5117 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#337">Defect Report 337</a>.
5122 <mark-expected-failures>
5123 <test name="is_polymorphic_test"/>
5124 <toolset name="gcc-2.95.3-stlport-*"/>
5125 <note author="Doug Gregor" refid="3"/>
5126 </mark-expected-failures>
5128 <mark-expected-failures>
5129 <test name="decay_test"/>
5130 <test name="extent_test"/>
5131 <test name="has_nothrow_assign_test"/>
5132 <test name="has_nothrow_constr_test"/>
5133 <test name="has_nothrow_copy_test"/>
5134 <test name="has_trivial_assign_test"/>
5135 <test name="has_trivial_constr_test"/>
5136 <test name="has_trivial_copy_test"/>
5137 <test name="has_trivial_destructor_test"/>
5138 <test name="is_array_test"/>
5139 <test name="is_base_and_derived_test"/>
5140 <test name="is_base_of_test"/>
5141 <test name="is_class_test"/>
5142 <test name="is_convertible_test"/>
5143 <test name="is_object_test"/>
5144 <test name="is_pod_test"/>
5145 <test name="is_polymorphic_test"/>
5146 <test name="rank_test"/>
5147 <test name="remove_all_extents_test"/>
5148 <test name="remove_bounds_test"/>
5149 <test name="remove_extent_test"/>
5150 <toolset name="sunpro-5_3-sunos"/>
5152 <note author="John Maddock">
5153 The Type Traits library is broken when used with Sunpro-5.3 and the
5154 argument to the template is an array or function type. Most other argument types
5155 do work as expected: in other words the functionality is limited
5156 with this compiler, but not so much as to render the library unuseable.
5158 </mark-expected-failures>
5159 <mark-expected-failures>
5160 <test name="is_empty_test"/>
5161 <test name="is_function_test"/>
5162 <test name="is_member_func_test"/>
5163 <test name="is_member_obj_test"/>
5164 <test name="is_reference_test"/>
5165 <test name="tricky_function_type_test"/>
5166 <test name="tricky_incomplete_type_test"/>
5167 <test name="tricky_is_enum_test"/>
5168 <toolset name="sunpro-5_3-sunos"/>
5169 <note author="John Maddock" refid="2"/>
5170 </mark-expected-failures>
5171 <mark-expected-failures>
5172 <test name="decay_test"/>
5173 <test name="extent_test"/>
5174 <test name="is_abstract_test"/>
5175 <test name="is_empty_test"/>
5176 <test name="is_function_test"/>
5177 <test name="is_member_func_test"/>
5178 <test name="is_member_obj_test"/>
5179 <test name="is_object_test"/>
5180 <test name="is_reference_test"/>
5181 <test name="rank_test"/>
5182 <test name="tricky_function_type_test"/>
5183 <toolset name="sun-5.8"/>
5185 <note author="John Maddock">
5186 The Type Traits library is broken when used with Sunpro-5.8 and the
5187 argument to the template is a function type. Most other argument types
5188 do work as expected: in other words the functionality is limited
5189 with this compiler, but not so much as to render the library unuseable.
5191 </mark-expected-failures>
5192 <mark-expected-failures>
5193 <test name="tricky_partial_spec_test"/>
5194 <toolset name="sun-5.9"/>
5195 <note author="John Maddock">
5196 This fails with an internal compiler error,
5197 there is no workaround as yet.
5199 </mark-expected-failures>
5200 <mark-expected-failures>
5201 <test name="tricky_function_type_test"/>
5202 <test name="is_const_test"/>
5203 <test name="is_volatile_test"/>
5204 <test name="is_convertible_test"/>
5205 <toolset name="gcc-2*"/>
5206 <note author="John Maddock" refid="2"/>
5207 </mark-expected-failures>
5208 <mark-expected-failures>
5209 <test name="aligned_storage_test"/>
5210 <toolset name="cw-8.3"/>
5211 <note author="John Maddock">
5212 Older versions of MWCW incorrectly align pointers to member functions
5213 (they use 12-byte boundaries, rather than a power-of-2 boundary),
5214 leading to alignment_of / aligned_storage
5215 to fail with these types on this compiler.
5217 </mark-expected-failures>
5221 <library name="tr1">
5223 <toolset name="msvc-7.0"/>
5224 <toolset name="msvc-6*"/>
5225 <note author="John Maddock">
5226 VC6/7 has a buggy using declaration syntax which
5227 basically makes it impossible to implement the
5228 namespace forwarding that this library relies upon.
5229 See KB article 263630 here: http://support.microsoft.com/default.aspx?scid=kb;en-us;263630
5233 <toolset name="cw-*"/>
5234 <note author="John Maddock">
5235 Metrowerks Codeworrier has partial TR1 support built in
5236 which conflicts with this implementation. Porting to this
5237 compiler is almost certainly possible, but will require some
5238 work by someone who has this compiler.
5241 <mark-expected-failures>
5242 <test name="test_type_traits"/>
5243 <test name="std_test_type_traits"/>
5244 <toolset name="msvc-7.1"/>
5245 <note author="John Maddock">
5246 Later versions of MSVC are required for these tests - the issues may
5247 be work-round-able if anyone cares enough to look into them.
5249 </mark-expected-failures>
5250 <mark-expected-failures>
5251 <test name="test_mem_fn_tricky"/>
5252 <test name="test_bind_tricky"/>
5253 <test name="test_ref_wrapper_tricky"/>
5254 <test name="test_function_tricky"/>
5255 <test name="std_test_mem_fn_tricky"/>
5256 <test name="std_test_bind_tricky"/>
5257 <test name="std_test_ref_wrapper_tricky"/>
5258 <test name="std_test_function_tricky"/>
5259 <test name="std_test_reference_wrapper_tricky"/>
5260 <test name="test_reference_wrapper_tricky"/>
5261 <test name="std_test_cmath_tricky"/>
5262 <test name="test_cmath_tricky"/>
5264 <note author="John Maddock">
5265 These tests test features that are not supported in the
5266 current Boost implementations of TR1 components, they will
5267 currently fail on all compilers, unless that compiler has
5270 </mark-expected-failures>
5272 <mark-expected-failures>
5273 <test name="run_random"/>
5274 <test name="std_run_random"/>
5275 <test name="std_test_bind"/>
5276 <test name="test_bind"/>
5277 <test name="std_test_regex"/>
5278 <test name="test_regex"/>
5279 <test name="std_test_result_of"/>
5280 <test name="test_result_of"/>
5281 <test name="tr1_has_tr1_result_of_pass"/>
5282 <test name="tr1_has_trivial_constr_test"/>
5283 <test name="tr1_is_base_of_test"/>
5284 <test name="tr1_is_convertible_test"/>
5285 <test name="tr1_is_pod_test"/>
5286 <test name="tr1_is_polymorphic_test"/>
5287 <test name="tr1_tky_function_type_test"/>
5288 <toolset name="msvc-9.0"/>
5289 <note author="John Maddock">
5290 MSVC 9.0 with the optional feature pack installed includes
5291 a version of the TR1 libraries that is not as interface-conforming
5292 as the Boost version. Most of these failures are of the "annoying"
5293 rather than "unusable" kind.
5295 </mark-expected-failures>
5297 <mark-expected-failures>
5298 <test name="test_regex"/>
5299 <test name="std_test_regex"/>
5300 <test name="test_hash"/>
5301 <test name="std_test_hash"/>
5302 <toolset name="mingw*"/>
5303 <toolset name="qcc*gpp"/>
5304 <toolset name="gcc-2*"/>
5305 <note author="John Maddock">
5306 These tests fail on this platform due to a lack of
5307 wide character support.
5309 </mark-expected-failures>
5311 <mark-expected-failures>
5312 <test name="test_cmath"/>
5313 <test name="std_test_cmath"/>
5314 <toolset name="intel-linux-9.0"/>
5315 <toolset name="darwin-4.0.1"/>
5316 <note author="John Maddock">
5317 These tests fail due to a lack of adequate
5318 long double std math lib support.
5320 </mark-expected-failures>
5322 <mark-expected-failures>
5323 <test name="test_regex"/>
5324 <test name="std_test_regex"/>
5325 <toolset name="gcc-mingw*"/>
5326 <note author="John Maddock">
5327 These tests fail on this platform due to incomplete
5328 wide character support.
5330 </mark-expected-failures>
5332 <mark-expected-failures>
5333 <test name="test_hash"/>
5334 <test name="std_test_hash"/>
5335 <toolset name="gcc-cygwin*"/>
5336 <note author="John Maddock">
5337 These tests fail on this platform due to incomplete
5338 wide character support.
5340 </mark-expected-failures>
5342 <mark-expected-failures>
5343 <test name="test_array"/>
5344 <test name="std_test_array"/>
5345 <test name="test_array_tricky"/>
5346 <test name="std_test_array_tricky"/>
5347 <test name="test_complex"/>
5348 <test name="std_test_complex"/>
5349 <test name="test_function"/>
5350 <test name="std_test_function"/>
5351 <test name="test_mem_fn"/>
5352 <test name="std_test_mem_fn"/>
5353 <test name="test_random"/>
5354 <test name="std_test_random"/>
5355 <test name="test_regex"/>
5356 <test name="std_test_regex"/>
5357 <test name="test_result_of"/>
5358 <test name="std_test_result_of"/>
5359 <test name="test_shared_ptr"/>
5360 <test name="std_test_shared_ptr"/>
5361 <test name="test_tr1_include"/>
5362 <test name="std_test_tr1_include"/>
5363 <test name="test_tuple"/>
5364 <test name="std_test_tuple"/>
5365 <test name="test_tuple_tricky"/>
5366 <test name="std_test_tuple_tricky"/>
5367 <test name="test_type_traits"/>
5368 <test name="std_test_type_traits"/>
5369 <test name="run_complex_overloads"/>
5370 <test name="std_run_complex_overloads"/>
5371 <test name="run_random"/>
5372 <test name="std_run_random"/>
5373 <test name="test_tuple_tricky"/>
5374 <test name="tr1_add_const_test"/>
5375 <test name="tr1_add_cv_test"/>
5376 <test name="tr1_add_pointer_test"/>
5377 <test name="tr1_add_reference_test"/>
5378 <test name="tr1_add_volatile_test"/>
5379 <test name="tr1_aligned_storage_test"/>
5380 <test name="tr1_alignment_of_test"/>
5381 <test name="tr1_has_nothrow_assign_test"/>
5382 <test name="tr1_has_nothrow_constr_test"/>
5383 <test name="tr1_has_nothrow_copy_test"/>
5384 <test name="tr1_has_trivial_assign_test"/>
5385 <test name="tr1_has_trivial_constr_test"/>
5386 <test name="tr1_has_trivial_copy_test"/>
5387 <test name="tr1_has_trivial_destr_test"/>
5388 <test name="tr1_has_virtual_destr_test"/>
5389 <test name="tr1_is_arithmetic_test"/>
5390 <test name="tr1_is_array_test"/>
5391 <test name="tr1_is_class_test"/>
5392 <test name="tr1_is_compound_test"/>
5393 <test name="tr1_is_const_test"/>
5394 <test name="tr1_is_convertible_test"/>
5395 <test name="tr1_is_empty_test"/>
5396 <test name="tr1_is_enum_test"/>
5397 <test name="tr1_is_floating_point_test"/>
5398 <test name="tr1_is_function_test"/>
5399 <test name="tr1_is_fundamental_test"/>
5400 <test name="tr1_is_integral_test"/>
5401 <test name="tr1_is_member_func_test"/>
5402 <test name="tr1_is_member_obj_test"/>
5403 <test name="tr1_is_member_pointer_test"/>
5404 <test name="tr1_is_object_test"/>
5405 <test name="tr1_is_pod_test"/>
5406 <test name="tr1_is_pointer_test"/>
5407 <test name="tr1_is_polymorphic_test"/>
5408 <test name="tr1_is_reference_test"/>
5409 <test name="tr1_is_same_test"/>
5410 <test name="tr1_is_scalar_test"/>
5411 <test name="tr1_is_signed_test"/>
5412 <test name="tr1_is_union_test"/>
5413 <test name="tr1_is_unsigned_test"/>
5414 <test name="tr1_is_void_test"/>
5415 <test name="tr1_is_volatile_test"/>
5416 <test name="tr1_remove_const_test"/>
5417 <test name="tr1_remove_cv_test"/>
5418 <test name="tr1_remove_pointer_test"/>
5419 <test name="tr1_remove_reference_test"/>
5420 <test name="tr1_remove_volatile_test"/>
5421 <test name="tr1_tky_abstract_type_test"/>
5422 <test name="tr1_tricky_add_pointer_test"/>
5423 <test name="tr1_tky_partial_spec_test"/>
5424 <toolset name="borland-5.6*"/>
5425 <note author="John Maddock">
5426 Support for Borland C++ in the various TR1 libraries is pretty
5427 poor (due to numerous compiler bugs sadly). The TR1 concept
5428 checks are *very* strict, and are expected to fail with this
5429 compiler. In addition most of the type_traits tests fail
5430 whenever debugging support is turned on with an internal
5431 compiler error. More conservative uses are more likely to succeed
5432 with this compiler however.
5434 </mark-expected-failures>
5436 <mark-expected-failures>
5437 <test name="test_complex"/>
5438 <test name="std_test_complex"/>
5439 <test name="test_function"/>
5440 <test name="std_test_function"/>
5441 <test name="test_random"/>
5442 <test name="std_test_random"/>
5443 <test name="test_result_of"/>
5444 <test name="std_test_result_of"/>
5445 <test name="test_tuple_tricky"/>
5446 <test name="std_test_tuple_tricky"/>
5447 <test name="test_type_traits"/>
5448 <test name="std_test_type_traits"/>
5449 <test name="run_complex_overloads"/>
5450 <test name="std_run_complex_overloads"/>
5451 <test name="test_shared_ptr"/>
5452 <test name="std_test_shared_ptr"/>
5453 <test name="std_run_random"/>
5454 <test name="run_random"/>
5455 <test name="test_tuple_tricky"/>
5456 <test name="tr1_is_convertible_test"/>
5457 <test name="tr1_remove_const_test"/>
5458 <test name="tr1_remove_pointer_test"/>
5459 <test name="tr1_remove_volatile_test"/>
5460 <test name="tr1_tricky_add_pointer_test"/>
5461 <test name="tr1_tky_partial_spec_test"/>
5462 <toolset name="borland-5.8*"/>
5463 <toolset name="borland-5.9*"/>
5464 <note author="John Maddock">
5465 Support for Borland C++ in the various TR1 libraries is pretty
5466 poor (due to numerous compiler bugs sadly). The TR1 concept
5467 checks are *very* strict, and are expected to fail with this
5468 compiler. More conservative uses are more likely to succeed
5469 with this compiler however.
5471 </mark-expected-failures>
5473 <mark-expected-failures>
5474 <test name="std_test_bind"/>
5475 <test name="test_bind"/>
5476 <toolset name="gcc-4*darwin"/>
5477 <toolset name="darwin*"/>
5478 <note author="John Maddock">
5479 These tests fail on this platform due to a recuring GCC bug.
5481 </mark-expected-failures>
5483 <mark-expected-failures>
5484 <test name="test_type_traits"/>
5485 <test name="std_test_type_traits"/>
5486 <test name="tr1_is_abstract_test"/>
5487 <toolset name="gcc-3.3.*"/>
5488 <toolset name="gcc-3.2*"/>
5489 <toolset name="qcc-3.3*"/>
5490 <note author="John Maddock">
5491 These tests fail due to a known compiler bug
5492 that is fixed in more recent GNU compiler releases. Users are
5493 very unlikely to encounter this as a real problem
5496 </mark-expected-failures>
5498 <mark-expected-failures>
5499 <test name="test_regex"/>
5500 <test name="std_test_regex"/>
5501 <test name="test_complex"/>
5502 <test name="std_test_complex"/>
5503 <test name="run_complex_overloads"/>
5504 <test name="std_run_complex_overloads"/>
5505 <toolset name="gcc-2*"/>
5506 <note author="John Maddock">
5507 These tests fail due to a known compiler bug
5508 that is fixed in more recent releases. Users are
5509 very unlikely to encounter this as a real problem
5512 </mark-expected-failures>
5514 <mark-expected-failures>
5515 <test name="test_type_traits"/>
5516 <test name="std_test_type_traits"/>
5517 <test name="test_result_of"/>
5518 <test name="std_test_result_of"/>
5519 <test name="tr1_is_abstract_test"/>
5520 <test name="test_ios"/>
5521 <test name="test_istream"/>
5522 <test name="test_ostream"/>
5523 <test name="test_streambuf"/>
5524 <test name="test_limits"/>
5525 <test name="test_locale"/>
5526 <test name="test_ios_std_header"/>
5527 <test name="test_istream_std_header"/>
5528 <test name="test_limits_std_header"/>
5529 <test name="test_locale_std_header"/>
5530 <test name="test_ostream_std_header"/>
5531 <test name="test_streambuf_std_header"/>
5532 <toolset name="gcc-2*"/>
5533 <note author="John Maddock">
5534 These tests fail due to a known compiler bug
5535 that is fixed in more recent releases. This
5536 functionality may not be usable with this compiler.
5538 </mark-expected-failures>
5540 <mark-expected-failures>
5541 <test name="run_complex_overloads"/>
5542 <test name="std_run_complex_overloads"/>
5543 <test name="std_test_complex"/>
5544 <test name="test_complex"/>
5545 <toolset name="qcc-3.3.5*gpp"/>
5546 <note author="John Maddock">
5547 These tests fail due to a known stdlib bug
5548 that has been reported to the vendor.
5550 </mark-expected-failures>
5552 <mark-expected-failures>
5553 <test name="tr1_function_traits_test"/>
5554 <test name="tr1_remove_bounds_test"/>
5555 <test name="tr1_remove_const_test"/>
5556 <test name="tr1_remove_cv_test"/>
5557 <test name="tr1_remove_pointer_test"/>
5558 <test name="tr1_remove_reference_test"/>
5559 <test name="tr1_remove_volatile_test"/>
5560 <test name="tr1_decay_test"/>
5561 <test name="tr1_extent_test"/>
5562 <test name="tr1_remove_extent_test"/>
5563 <test name="tr1_remove_all_extents_test"/>
5564 <test name="tr1_rank_test"/>
5565 <test name="tr1_is_unsigned_test"/>
5566 <toolset name="msvc-6.5*"/>
5567 <toolset name="msvc-7.0"/>
5568 <note author="Aleksey Gurtovoy">
5569 This failure is caused by the lack of compiler support for class template
5570 partial specialization. A limited subset of the tested functionality is
5571 available on the compiler through a user-side workaround (see
5572 <a href="http://www.boost.org/libs/type_traits/index.html#transformations">
5573 http://www.boost.org/libs/type_traits/index.html#transformations</a> for
5576 </mark-expected-failures>
5577 <mark-expected-failures>
5578 <test name="tr1_tky_incomplete_type_test"/>
5579 <test name="tr1_tky_incomp_type_test"/>
5580 <test name="tr1_decay_test"/>
5581 <test name="tr1_extent_test"/>
5582 <test name="tr1_is_base_of_test"/>
5583 <test name="tr1_rank_test"/>
5584 <test name="tr1_remove_all_extents_test"/>
5585 <test name="tr1_remove_extent_test"/>
5586 <test name="tr1_tky_function_type_test"/>
5587 <toolset name="borland-5.6*"/>
5588 <toolset name="borland-5.8*"/>
5589 <toolset name="borland-5.9*"/>
5590 <note author="John Maddock" refid="2"/>
5591 </mark-expected-failures>
5592 <test name="tr1_tricky_is_enum_test">
5594 <toolset name="borland-5.6*"/>
5595 <toolset name="borland-5.8*"/>
5596 <toolset name="borland-5.9*"/>
5597 <toolset name="msvc-6.5*"/>
5598 <toolset name="gcc-2.95.3-*"/>
5601 <test name="tr1_tricky_incomplete_type_test">
5603 <toolset name="iw-7_1*"/>
5604 <note author="John Maddock" refid="2"/>
5607 <test name="tr1_tricky_incomp_type_test">
5609 <toolset name="iw-7_1*"/>
5610 <note author="John Maddock" refid="2"/>
5613 <test name="tr1_tky_incomp_type_test">
5615 <toolset name="iw-7_1*"/>
5616 <note author="John Maddock" refid="2"/>
5619 <test name="tr1_is_abstract_test">
5621 <toolset name="borland-5.6*"/>
5622 <toolset name="borland-5.8*"/>
5623 <toolset name="borland-5.9*"/>
5624 <toolset name="cw-8.3*"/>
5625 <toolset name="cw-9.3*"/>
5626 <toolset name="cw-9.4*"/>
5627 <toolset name="cw-9.5*"/>
5628 <toolset name="msvc-6.5*"/>
5629 <toolset name="msvc-7.0"/>
5630 <toolset name="mingw-3_3*"/>
5631 <toolset name="gcc-2*"/>
5632 <toolset name="gcc-3.2*"/>
5633 <toolset name="gcc-3.3*"/>
5634 <toolset name="gcc-3_3*"/>
5635 <toolset name="qcc-3_3*"/>
5636 <toolset name="sunpro-5_3-sunos"/>
5637 <toolset name="hp_cxx-65*"/>
5638 <toolset name="darwin"/>
5639 <toolset name="mingw"/>
5640 <note author="Aleksey Gurtovoy">
5641 This functionality is available only on compilers that implement C++ Core Language
5642 <a href="http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#337">Defect Report 337</a>.
5647 <mark-expected-failures>
5648 <test name="tr1_is_polymorphic_test"/>
5649 <toolset name="gcc-2.95.3-stlport-*"/>
5650 <note author="Doug Gregor" refid="3"/>
5651 </mark-expected-failures>
5653 <mark-expected-failures>
5654 <test name="tr1_decay_test"/>
5655 <test name="tr1_extent_test"/>
5656 <test name="tr1_has_nothrow_assign_test"/>
5657 <test name="tr1_has_nothrow_constr_test"/>
5658 <test name="tr1_has_nothrow_copy_test"/>
5659 <test name="tr1_has_trivial_assign_test"/>
5660 <test name="tr1_has_trivial_constr_test"/>
5661 <test name="tr1_has_trivial_copy_test"/>
5662 <test name="tr1_has_trivial_destr_test"/>
5663 <test name="tr1_is_array_test"/>
5664 <test name="tr1_is_base_and_derived_test"/>
5665 <test name="tr1_is_base_of_test"/>
5666 <test name="tr1_is_class_test"/>
5667 <test name="tr1_is_convertible_test"/>
5668 <test name="tr1_is_object_test"/>
5669 <test name="tr1_is_pod_test"/>
5670 <test name="tr1_is_polymorphic_test"/>
5671 <test name="tr1_rank_test"/>
5672 <test name="tr1_remove_all_extents_test"/>
5673 <test name="tr1_remove_bounds_test"/>
5674 <test name="tr1_remove_extent_test"/>
5675 <toolset name="sunpro-5_3-sunos"/>
5677 <note author="John Maddock">
5678 The Type Traits library is broken when used with Sunpro-5.3 and the
5679 argument to the template is an array or function type. Most other argument types
5680 do work as expected: in other words the functionality is limited
5681 with this compiler, but not so much as to render the library unuseable.
5683 </mark-expected-failures>
5684 <mark-expected-failures>
5685 <test name="tr1_decay_test"/>
5686 <test name="tr1_extent_test"/>
5687 <test name="tr1_is_abstract_test"/>
5688 <test name="tr1_is_empty_test"/>
5689 <test name="tr1_is_function_test"/>
5690 <test name="tr1_is_member_func_test"/>
5691 <test name="tr1_is_member_obj_test"/>
5692 <test name="tr1_is_object_test"/>
5693 <test name="tr1_is_reference_test"/>
5694 <test name="tr1_rank_test"/>
5695 <test name="tr1_tricky_function_type_test"/>
5696 <test name="tr1_tky_function_type_test"/>
5697 <test name="test_type_traits"/>
5698 <test name="std_test_type_traits"/>
5699 <toolset name="sun-5.8"/>
5701 <note author="John Maddock">
5702 The Type Traits library is broken when used with Sunpro-5.8 and the
5703 argument to the template is a function type. Most other argument types
5704 do work as expected: in other words the functionality is limited
5705 with this compiler, but not so much as to render the library unuseable.
5707 </mark-expected-failures>
5708 <mark-expected-failures>
5709 <test name="test_random"/>
5710 <test name="std_test_random"/>
5711 <toolset name="sun-5.8"/>
5712 <toolset name="sun-5.9"/>
5714 <note author="John Maddock">
5715 These failures appear to represent a genuine issue with the
5716 Boost.Random library that has yet to be addressed.
5718 </mark-expected-failures>
5719 <mark-expected-failures>
5720 <test name="test_tuple_tricky"/>
5721 <test name="std_test_tuple_tricky"/>
5722 <toolset name="sun-5.8"/>
5724 <note author="John Maddock">
5725 These fail with an internal compiler error: there's no
5728 </mark-expected-failures>
5729 <mark-expected-failures>
5730 <test name="tr1_tky_partial_spec_test"/>
5731 <toolset name="sun-5.9"/>
5732 <note author="John Maddock">
5733 This fails with an internal compiler error: there's no
5736 </mark-expected-failures>
5737 <mark-expected-failures>
5738 <test name="test_boost"/>
5739 <test name="test_hash"/>
5740 <test name="test_random"/>
5741 <test name="test_regex"/>
5742 <toolset name="msvc-7.1_stlport4"/>
5744 <note author="John Maddock">
5745 These failures are completely spurious: they're caused by the tests
5746 being run with bjam -j2 and the post-processing not coping with the
5747 resulting output. These failures should clear if these tests
5748 are re-run at some point in the future.
5750 </mark-expected-failures>
5751 <mark-expected-failures>
5752 <test name="tr1_is_empty_test"/>
5753 <test name="tr1_is_function_test"/>
5754 <test name="tr1_is_member_func_test"/>
5755 <test name="tr1_is_member_obj_test"/>
5756 <test name="tr1_is_reference_test"/>
5757 <test name="tr1_tricky_function_type_test"/>
5758 <test name="tr1_tricky_incomplete_type_test"/>
5759 <test name="tr1_tricky_incomp_type_test"/>
5760 <test name="tr1_tricky_is_enum_test"/>
5761 <toolset name="sunpro-5_3-sunos"/>
5762 <note author="John Maddock" refid="2"/>
5763 </mark-expected-failures>
5764 <mark-expected-failures>
5765 <test name="tr1_tricky_function_type_test"/>
5766 <test name="tr1_is_const_test"/>
5767 <test name="tr1_is_volatile_test"/>
5768 <test name="tr1_is_convertible_test"/>
5769 <toolset name="gcc-2*"/>
5770 <note author="John Maddock" refid="2"/>
5771 </mark-expected-failures>
5772 <mark-expected-failures>
5773 <test name="test_array"/>
5774 <test name="std_test_array"/>
5775 <test name="test_array_tricky"/>
5776 <test name="std_test_array_tricky"/>
5777 <test name="test_bind"/>
5778 <test name="std_test_bind"/>
5779 <test name="test_complex"/>
5780 <test name="std_test_complex"/>
5781 <test name="test_function"/>
5782 <test name="std_test_function"/>
5783 <test name="test_random"/>
5784 <test name="std_test_random"/>
5785 <test name="test_reference_wrapper"/>
5786 <test name="std_test_reference_wrapper"/>
5787 <test name="test_regex"/>
5788 <test name="std_test_regex"/>
5789 <test name="test_result_of"/>
5790 <test name="std_test_result_of"/>
5791 <test name="test_shared_ptr"/>
5792 <test name="std_test_shared_ptr"/>
5793 <test name="test_tuple"/>
5794 <test name="std_test_tuple"/>
5795 <toolset name="vc-7"/>
5796 <note author="John Maddock">
5797 This library is almost unusable with VC7 due to name lookup issues.
5799 </mark-expected-failures>
5800 <mark-expected-failures>
5801 <test name="tr1_aligned_storage_test"/>
5802 <toolset name="cw-8.3"/>
5803 <note author="John Maddock">
5804 Older versions of MWCW incorrectly align pointers to member functions
5805 (they use 12-byte boundaries, rather than a power-of-2 boundary),
5806 leading to alignment_of / aligned_storage
5807 to fail with these types on this compiler.
5809 </mark-expected-failures>
5813 <library name="units">
5814 <mark-expected-failures>
5815 <test name="fail_base_dimension"/>
5816 <toolset name="vacpp"/>
5817 <note author="Steven Watanabe" refid="16"/>
5818 </mark-expected-failures>
5822 <library name="unordered">
5823 <mark-expected-failures>
5824 <test name="unnecessary_copy_tests"/>
5825 <toolset name="borland-*"/>
5826 <toolset name="sun-*"/>
5827 <note author="Daniel James">
5828 This tests whether inserting elements creates as few copies as I think
5829 is possible. If this fails it just means that the container might be
5830 a little inefficient.
5832 </mark-expected-failures>
5834 <mark-expected-failures>
5835 <test name="compile_map"/>
5836 <test name="compile_set"/>
5837 <toolset name="gcc-open64"/>
5838 <toolset name="pathscale-*"/>
5839 <note author="Daniel James">
5840 Concept checks don't seem to work on pathscale.
5842 </mark-expected-failures>
5845 <!-- utility/enable_if -->
5846 <library name="utility/enable_if">
5848 <toolset name="borland-5.6*"/>
5849 <toolset name="borland-5.8*"/>
5850 <toolset name="cw-8.3*"/>
5851 <toolset name="msvc-6.5*"/>
5852 <toolset name="msvc-7.0"/>
5853 <toolset name="gcc-2.95.3-*"/>
5857 <mark-expected-failures>
5858 <test name="no_disambiguation"/>
5859 <toolset name="gcc-3.2.*"/>
5861 </mark-expected-failures>
5863 <mark-expected-failures>
5864 <test name="partial_specializations"/>
5865 <toolset name="borland-5.9*"/>
5866 <note author="Alisdair Meredith" refid="29"/>
5867 </mark-expected-failures>
5870 <!-- utility/swap -->
5871 <library name="utility/swap">
5872 <mark-expected-failures>
5873 <test name="array_of_array_of_class"/>
5874 <test name="array_of_class"/>
5875 <test name="specialized_in_std"/>
5876 <toolset name="borland-6.10.0"/>
5877 <note refid="3" author="Niels Dekker" date="2008-11-11">
5878 The definition of a custom template specialization of std::swap
5879 appears to trigger an internal compiler error ("Fatal F1004") on
5880 CodeGear 6.10.0 (formerly named Borland), as I reported,
5881 with help from Nicola Musatti and David Dean.
5882 Related Boost mailing list discussion:
5883 http://lists.boost.org/Archives/boost/2008/11/144465.php
5884 CodeGear bug reports on this issue:
5885 http://qc.codegear.com/wc/qcmain.aspx?d=68959
5886 http://qc.codegear.com/wc/qcmain.aspx?d=69196
5888 </mark-expected-failures>
5889 <mark-expected-failures>
5890 <test name="array_of_array_of_class"/>
5891 <test name="array_of_array_of_int"/>
5892 <toolset name="borland-5.9.3"/>
5893 <note refid="3" author="Niels Dekker" date="2008-12-09">
5894 Borland 5.9.3 has an error (E2285) when trying to pass a
5895 multi-dimensional array by reference to a function template.
5896 A bug report by Christopher Yeleighton appears related:
5897 "The compiler obligatorily converts member arrays to pointers"
5898 http://qc.codegear.com/wc/qcmain.aspx?d=10267
5900 </mark-expected-failures>
5904 <library name="utility">
5905 <test name="addressof_test">
5907 <toolset name="sunpro-5_3-sunos"/>
5908 <note author="D. Gregor" refid="3"/>
5911 <test name="fun_out_iter_example">
5913 <toolset name="como-win32"/>
5914 <note author="B. Dawes" refid="4"/>
5917 <test name="numeric_traits_test">
5919 <toolset name="borland-5.6*"/>
5920 <toolset name="borland-5.8*"/>
5921 <toolset name="borland-5.9*"/>
5922 <note author="A.Meredith">
5923 Compiler has a problem with BOOST_STATIC_CONSTANT in nested templates
5924 inside class template specializations.
5928 <test name="result_of_test">
5930 <toolset name="borland-5*"/>
5931 <toolset name="cw-8.3*"/>
5932 <toolset name="msvc-6.5*"/>
5933 <toolset name="msvc-7.0"/>
5934 <toolset name="gcc-2.95.3*"/>
5935 <toolset name="sunpro-5_3-sunos"/>
5936 <note refid="3" author="D. Gregor"/>
5939 <test name="value_init_test">
5941 <toolset name="intel-9.0*"/>
5942 <toolset name="intel-linux-9.0*"/>
5943 <note author="Niels Dekker" date="2009-07-01">
5944 When I made the conversion from value_initialized<T> to T&
5945 const-correct, this specific compiler version gave compile errors.
5946 See also: Ticket #2548 - "Let's fix the logical constness of value_initialized!"
5950 <mark-expected-failures>
5951 <test name="value_init_test"/>
5952 <toolset name="msvc-6.5*"/>
5953 <toolset name="msvc-7.0"/>
5954 <note author="Aleksey Gurtovoy">
5955 This failure is caused by a compiler bug (default-constructed scalar
5956 types are not zero-initialized) that has been fixed in the latest
5957 versions of the compiler (VC 7.1 and greater).
5959 </mark-expected-failures>
5960 <mark-expected-failures>
5961 <test name="operators_test"/>
5962 <toolset name="gcc-3.4.5_linux_x86_64"/>
5963 <note author="Vladimir Prus">
5964 The test takes more that 30 minutes to compile and the
5965 compilation is automatically killed. It is likely caused
5966 by the compiler bug, but it unknown how much this
5967 bug affects regular use of the operators library. Is it
5968 also unknown if the test can be refactored so that
5969 not to trigger this bug.
5971 </mark-expected-failures>
5975 <library name="uuid">
5976 <mark-expected-failures>
5977 <test name="test_serialization"/>
5978 <toolset name="cuda-2.2"/>
5979 <toolset name="borland-cb2009"/>
5980 <toolset name="borland-cb2010"/>
5981 <toolset name="borland-5.9.3"/>
5982 <toolset name="borland-6.1.3"/>
5983 <toolset name="borland-6.2.1"/>
5984 <note author="Andy Tompkins">
5985 The test relies on Boost.Serialization which is not
5986 supported on this toolset.
5988 </mark-expected-failures>
5989 <mark-expected-failures>
5990 <test name="test_random_generator"/>
5991 <test name="test_tagging"/>
5992 <test name="test_uuid_class"/>
5993 <toolset name="borland-cb2009"/>
5994 <toolset name="borland-cb2010"/>
5995 <toolset name="borland-5.9.3"/>
5996 <toolset name="borland-6.1.3"/>
5997 <toolset name="borland-6.2.1"/>
5998 <note author="Andy Tompkins">
5999 The test relies on Boost.Random which is not supported
6002 </mark-expected-failures>
6003 <mark-expected-failures>
6004 <test name="test_uuid"/>
6005 <toolset name="borland-cb2009"/>
6006 <toolset name="borland-cb2010"/>
6007 <toolset name="borland-5.9.3"/>
6008 <toolset name="borland-6.1.3"/>
6009 <toolset name="borland-6.2.1"/>
6010 <note author="Andy Tompkins" refid="28"/>
6011 </mark-expected-failures>
6012 <mark-expected-failures>
6013 <test name="compile_random_generator"/>
6014 <test name="compile_uuid_generators"/>
6015 <test name="test_include1"/>
6016 <toolset name="borland-6.2.1"/>
6017 <note author="Andy Tompkins">
6018 The test relies on Boost.Iterator (iterator_facade)
6019 which is not supported on this toolset.
6021 </mark-expected-failures>
6025 <library name="variant">
6027 <toolset name="mipspro"/>
6028 <toolset name="sunpro-5_3-sunos"/>
6029 <toolset name="hp_cxx-65*"/>
6032 <test name="recursive_variant_test">
6034 <toolset name="como-win32"/>
6035 <toolset name="msvc-6.5*"/>
6036 <toolset name="msvc-7.0"/>
6040 <mark-expected-failures>
6041 <test name="recursive_variant_test"/>
6042 <test name="variant_test1"/>
6043 <test name="variant_test5"/>
6044 <test name="variant_visit_test"/>
6045 <toolset name="borland-5.6*"/>
6046 <toolset name="borland-5.8*"/>
6047 <toolset name="borland-5.9*"/>
6048 <note author="Aleksey Gurtovoy" refid="3"/>
6049 </mark-expected-failures>
6050 <test name="variant_reference_test">
6052 <toolset name="cw-8.3*"/>
6053 <toolset name="msvc-6.5*"/>
6054 <toolset name="msvc-7.0"/>
6058 <toolset name="iw-7_1*"/>
6059 <toolset name="intel-7.1*"/>
6066 <library name="wave">
6068 <toolset name="msvc-6.5*"/>
6069 <toolset name="sunpro-5_3-sunos"/>
6070 <toolset name="borland-5.5*"/>
6071 <toolset name="borland-5.6*"/>
6072 <toolset name="borland-5.8*"/>
6073 <toolset name="gcc-2.95.3-linux"/>
6074 <toolset name="gcc-2.95.3-stlport-4.5.3-linux"/>
6075 <toolset name="gcc-2.95.3-stlport-4.6.2-linux"/>
6076 <toolset name="hp_cxx-65*"/>
6080 <toolset name="msvc-7.0"/>
6082 This toolset isn't supported because of the used Spirit V1.8.x, which in turn is
6083 not usable with this toolset.
6088 <toolset name="borland-5.9*"/>
6089 <note author="Alisdair Meredith">
6090 This toolset isn't supported because of the used multi_index library, which in turn is
6091 not usable with this toolset.
6095 <mark-expected-failures>
6096 <test name="testwave"/>
6097 <!-- toolset name="cw-9_5-darwin"/ -->
6098 <toolset name="cw-8*"/>
6099 <note author="Rene Rivera" refid="29"/>
6100 </mark-expected-failures>
6102 <mark-expected-failures>
6103 <test name="testwave"/>
6104 <toolset name="gcc-3.2.3-linux"/>
6105 <toolset name="gcc-3.2.3_linux"/>
6106 <toolset name="gcc-3.3.6-linux"/>
6107 <toolset name="gcc-3.3.6"/>
6108 <note author="Hartmut Kaiser" refid="29"/>
6109 </mark-expected-failures>
6111 <mark-expected-failures>
6112 <test name="testwave"/>
6113 <!-- <toolset name="qcc-3.3.5_gpp"/> -->
6114 <toolset name="qcc-3.3.5*gpp"/>
6115 <note author="Hartmut Kaiser" refid="29"/>
6116 </mark-expected-failures>
6118 <mark-expected-failures>
6119 <test name="testwave_dll"/>
6120 <toolset name="mingw-3*"/>
6121 <note author="Hartmut Kaiser" refid="29"/>
6122 </mark-expected-failures>
6124 <mark-expected-failures>
6125 <test name="testwave_dll"/>
6126 <toolset name="cw-9.4"/>
6127 <note author="Hartmut Kaiser" refid="2"/>
6128 </mark-expected-failures>
6130 <mark-expected-failures>
6131 <test name="test_slex_lexer"/>
6132 <toolset name="hp_cxx-65*"/>
6133 <note author="Hartmut Kaiser" refid="2"/>
6134 </mark-expected-failures>
6139 <library name="xpressive">
6142 <toolset name="gcc-2.95.3*"/>
6143 <toolset name="msvc-6.5*"/>
6144 <toolset name="msvc-7.0"/>
6145 <note author="Eric Niebler">
6146 These compilers do not support class template partial
6151 <toolset name="borland-*"/>
6152 <note author="Eric Niebler">
6153 Boost.Proto doesn't work on this compiler.
6157 <toolset name="cw-8.3"/>
6158 <note author="Eric Niebler">
6159 This compiler doesn't support SFINAE / enable_if
6163 <toolset name="dmc*"/>
6164 <note author="Eric Niebler">
6165 Digital Mars cannot seem to handle dependent default template parameters,
6166 such as "template < class T, bool B = is_foo < T > ::value >"
6170 <toolset name="sunpro-5_3-sunos"/>
6171 <toolset name="sun-5.7"/>
6172 <toolset name="sun-5.8"/>
6173 <toolset name="sun-5.9"/>
6174 <toolset name="sun-5.10"/>
6178 <mark-expected-failures>
6179 <test name="test_actions"/>
6180 <toolset name="acc"/>
6181 <note author="Eric Niebler" refid="43"/>
6182 </mark-expected-failures>
6184 <mark-expected-failures>
6185 <test name="test_symbols"/>
6186 <toolset name="acc"/>
6187 <note author="Eric Niebler" refid="43"/>
6188 </mark-expected-failures>
6192 <!-- /////////////// Standard note definitions /////////////// -->
6195 This test fails only intermittently.
6199 The failure is caused by a problem in Boost code. The Boost developers are aware of
6200 the problem and plan to fix it.
6204 The failure is caused by a compiler bug.
6208 The failure is caused by a compiler bug, which has been reported to the compiler
6209 supplier (or is already known to them).
6213 The failure is caused by a standard library bug.
6217 The failure is caused by a standard library bug, which has been reported to the
6218 standard library supplier (or is already known to them).
6222 The failure is probably caused by the test code, harness, or configuration. Thus,
6223 it may not affect users of the library.
6227 The failure is serious and likely to prevent all use of this Boost library with this compiler.
6231 The failure is serious and likely to prevent all use of this Boost library with this
6232 compiler. The failure is caused by a compiler bug, which has been reported to the compiler
6233 supplier (or is already known to them).
6237 The failure is caused by a platform API bug.
6241 The failure is caused by a platform API bug, which has been reported to the platform API
6242 supplier (or is already known to them).
6246 The failure is not serious and will not affect most users. The library degrades gracefully.
6250 This compiler's bugs are not supported by the library.
6254 Locales missing or adequately supported by this compiler.
6258 Missing or inadequate wchar/wstring/wstream support for this compiler.
6262 No std iterator traits for this compiler.
6266 Library has limited input/output support due to compiler inadequacies.
6270 No high precision clock for this platform.
6274 A bug in standard library prevents passing std::set from DLL to
6275 application. A fixed <tree> header is available from
6276 http://www.dinkumware.com/vc_fixes.html.
6280 Although the documentation from the Comeau website would make it appear
6281 that windows DLL's are supported using the --windows option, after some
6282 experimentation we have been unsuccessful in making dll configurations
6287 The failure is caused by a runtime limitation. Locale support is only
6288 available with the static linked variant of the runtime. Generally
6289 the dynamic linked variant is required when building dynamic modules,
6290 DLL, <code>so</code>, etc.
6294 This failure is caused by a compiler bug with no known workaround.
6295 Patches are welcome!
6299 This failure is caused by bugs in the standard library implementation and/or
6300 bugs in the compiler.
6304 Unresearched failure -- please contact library developers for more
6305 information about possible causes.
6309 The test fails due to unresearched issues. The library
6310 developers are aware of this failure, but need help with
6311 investigating/addressing it for future releases.
6315 The support for this deficient compiler will be dropped starting
6316 from Boost 1.33.0 release. Please use one of the previous Boost
6317 releases if you need the library to work on this compiler.
6321 This failure is caused by compiler bugs or limitations. Some advanced
6322 or esoteric library features may be unavailable or only partially available.
6323 This does not impact most common uses of the library.
6327 This failure is caused by a compiler bug. Certain code constructs that should
6328 fail compilation are accepted by the compiler. This can mask some programming
6329 errors, but does not impact the usability of the library.
6333 The failures are caused by the wrong handling of the
6334 <code>std::internal</code> flag in the iostreams implementation of the
6335 standard library used on that compiler/platform combo. Apart from that,
6336 the format library works as expected.
6340 The failures are caused by the fact that the iword and pword arrays seem
6341 to share the same memory area in the iostreams implementation of the
6342 standard library used on that compiler/platform combo. As long as you
6343 stay clear of iword and pword, the library should work ok.
6347 This failure occurs only when using shared libraries for this
6348 compiler and platform, although the same programs should work
6349 properly when using static libraries. This problem has not
6354 Wide character support is disabled in the GNU Standard C++ library as
6355 supplied on the QNX Neutrino version 6.3.0 distribution.
6359 This problem is due to the non-conforming STLport
6360 implementation of vector's swap: it can be easily
6361 reproduced with the following code snippet:
6363 typedef std::vector<int> vector_type;
6364 typedef vector_type::reference reference_type;
6366 vector_type v1(4u, 1);
6367 vector_type v2(7u, 0);
6369 reference_type ref = v1[2];
6373 BOOST_CHECK(v2[2] == x); // ok
6375 BOOST_CHECK(ref != x); // oops
6379 When compiling this test, aCC6 runs out of memory. The HP
6380 compiler group is aware of this issue and is working on the fix.
6384 This test assumes native typeof support.
6388 This test assumes compiler support for rvalue references.
6392 These tests rely on the ability of an std::deque container to be
6393 constructed off two input iterators. Unfortunately, the Rogue Wave
6394 library version 2.2 and higher assumes iterator which has + and -
6395 operators which only random access iterator is required to provide.
6399 Internal compiler error: GCC Bugzilla Bug 33580.
6400 This is a regression in the gcc 4.2 series.
6404 These test failures are reported to be
6405 under investigation at HP's compiler lab.
6409 This compiler does not support gcc stdcall function attribute.
6413 The Rogue Wave standard library version used by this compiler provides
6414 a faulty vector<bool> iterator, which is not symmetric. There is an
6415 associated bug report in the Rogue Wave bug tracking system for this
6420 The test does not compile, most likely because of new version of EDG Front End
6421 implementing Core Issue 574. In the HP bug tracking system, it is tracked as
6422 QuIX ID: QXCR1000804484.
6426 Depending on system load, the compile time may exceed specified timeout value.
6427 The test passes when the timeout value is increased.
6431 This test fails when BOOST_UBLAS_NO_NESTED_CLASS_RELATION is defined.
6435 This test fails because MinGW apparently does not always catch exceptions properly.
6438 </explicit-failures-markup>