[1.41.0] Beta 1 release candidate
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/ Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems. Thanks, --Beman
Beman Dawes escribió:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
boost-docs.7z contains two subdirs named boost_propertytree and property_tree. Is this intentional? Also, a handful of libs have folders prefixed with boost_, which seems not to be the norm. Joaquín M López Muñoz Telefónica, Investigación y Desarrollo
On Mon, Nov 2, 2009 at 12:18 PM, <joaquin@tid.es> wrote:
boost-docs.7z contains two subdirs named boost_propertytree and property_tree. Is this intentional? Also, a handful of libs have folders prefixed with boost_, which seems not to be the norm.
'boost-docs.7z' contains just the generated files for the 'doc/html' folder, it's not meant for general use. Daniel
On Mon, Nov 2, 2009 at 8:37 AM, Daniel James <daniel_james@fmail.co.uk>wrote:
On Mon, Nov 2, 2009 at 12:18 PM, <joaquin@tid.es> wrote:
boost-docs.7z contains two subdirs named boost_propertytree and property_tree. Is this intentional? Also, a handful of libs have folders prefixed with boost_, which seems not to be the norm.
'boost-docs.7z' contains just the generated files for the 'doc/html' folder, it's not meant for general use.
My boo boo. I should have pointed to just the release specific files. For Windows line endings: http://boost.cowic.de/rc/boost_1_41_1_beta1.7z <http://boost.cowic.de/rc/> http://boost.cowic.de/rc/boost_1_41_1_beta1.zip For Unix line endings: http://boost.cowic.de/rc/boost_1_41_1_beta1.tar.bz2 http://boost.cowic.de/rc/boost_1_41_1_beta1.tar.gz --Beman
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
linux: straight Ubuntu 9.04 compiler: gcc-4.3.3 commands used: ./bootstrap.sh ./bjam stage All builds fine *except* Boost.Python which fails to compile (nearly?) all files with errors such as: ./boost/python/detail/wrap_python.hpp:75:24: error: patchlevel.h: No such file or directory -and- error: ‘PyObject’ was not declared in this scope I am not a normal user of Boost.Python so I suspect I don't have the correct dependencies installed. michael -- ---------------------------------- Michael Caisse Object Modeling Designs www.objectmodelingdesigns.com
Michael Caisse wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
linux: straight Ubuntu 9.04 compiler: gcc-4.3.3
commands used:
./bootstrap.sh ./bjam stage
All builds fine *except* Boost.Python which fails to compile (nearly?) all files with errors such as:
./boost/python/detail/wrap_python.hpp:75:24: error: patchlevel.h: No such file or directory
-and-
error: ‘PyObject’ was not declared in this scope
I am not a normal user of Boost.Python so I suspect I don't have the correct dependencies installed.
You need python-dev package installed, I think. I plan to have an automatic check in 1.42. - Volodya
Vladimir Prus wrote:
Michael Caisse wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
linux: straight Ubuntu 9.04 compiler: gcc-4.3.3
commands used:
./bootstrap.sh ./bjam stage
All builds fine *except* Boost.Python which fails to compile (nearly?) all files with errors such as:
./boost/python/detail/wrap_python.hpp:75:24: error: patchlevel.h: No such file or directory
-and-
error: ‘PyObject’ was not declared in this scope
I am not a normal user of Boost.Python so I suspect I don't have the correct dependencies installed.
You need python-dev package installed, I think. I plan to have an automatic check in 1.42.
- Volodya
That did the trick. -- ---------------------------------- Michael Caisse Object Modeling Designs www.objectmodelingdesigns.com
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http:// boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Hi Beman, A couple of minor documentation issues. I noticed that these libraries are not listed in the Libraries Listed Alphabetically, though they are listed in the Libraries by Category. Perhaps someone could add them to the alphabetical listing? conversion/lexical_cast - lexical_cast class template, from Kevlin Henney result_of - determines the type of a function call expression. base-from-member - Idiom to initialize a base class with a member, from Daryle Walker. -- Noel
On Mon, Nov 2, 2009 at 1:38 PM, K. Noel Belcourt <kbelco@sandia.gov> wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http:// boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Hi Beman,
A couple of minor documentation issues. I noticed that these libraries are not listed in the Libraries Listed Alphabetically, though they are listed in the Libraries by Category. Perhaps someone could add them to the alphabetical listing?
conversion/lexical_cast - lexical_cast class template, from Kevlin Henney
Added. Listed under the lexical_cast name, as that's where people will look for it.
result_of - determines the type of a function call expression.
Added.
base-from-member - Idiom to initialize a base class with a member, from Daryle Walker.
That's one of several listed as "Utilities". While they probably should be broken out separately, I don't have the patience at the moment. These changes will go in the release, but not the beta. Thanks, --Beman
On 02.11.2009 17:46, Michael Caisse wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
linux: straight Ubuntu 9.04 compiler: gcc-4.3.3
commands used:
./bootstrap.sh ./bjam stage
All builds fine *except* Boost.Python which fails to compile (nearly?) all files with errors such as:
./boost/python/detail/wrap_python.hpp:75:24: error: patchlevel.h: No such file or directory
-and-
error: ‘PyObject’ was not declared in this scope
I am not a normal user of Boost.Python so I suspect I don't have the correct dependencies installed.
michael
Same configuration but with MPI enabled. Everything builds fine except for BGL parallel. In boost/graph/distributed/detail/mpi_process_group.ipp needs <cstdio> to be included. Regards Henning
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
windows compiler: vc-80 commands used: .\bootstrap.bat .\bjam stage The bootstrap results with: ----------------------- Building Boost.Jam build engine The system cannot find the file specified. ----------------------- but then continues fine. All builds as expected. michael -- ---------------------------------- Michael Caisse Object Modeling Designs www.objectmodelingdesigns.com
Michael Caisse wrote:
windows compiler: vc-80
commands used:
.\bootstrap.bat .\bjam stage
The bootstrap results with:
----------------------- Building Boost.Jam build engine The system cannot find the file specified. -----------------------
This is related to creating the tools\jam\src\bootstrap directory. If it starts off with an empty dir there is no message. HTH michael -- ---------------------------------- Michael Caisse Object Modeling Designs www.objectmodelingdesigns.com
Michael Caisse wrote:
Michael Caisse wrote:
windows compiler: vc-80
commands used:
.\bootstrap.bat .\bjam stage
The bootstrap results with:
----------------------- Building Boost.Jam build engine The system cannot find the file specified. -----------------------
This is related to creating the tools\jam\src\bootstrap directory. If it starts off with an empty dir there is no message.
Michael, thanks for diagnosing. I've checked in a patch (trunk, release) that should fix the issue. Could you give it a try? - Volodya
Vladimir Prus wrote:
Michael Caisse wrote:
This is related to creating the tools\jam\src\bootstrap directory. If it starts off with an empty dir there is no message.
Michael,
thanks for diagnosing. I've checked in a patch (trunk, release) that should fix the issue. Could you give it a try?
- Volodya
Checked out boost/trunk/tools/jam (57403) into the zip version of 1.41.0 beta 1 from sourceforge. .\boostrap.bat now looks clean. Thanks michael -- ---------------------------------- Michael Caisse Object Modeling Designs www.objectmodelingdesigns.com
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
Platform: Ubuntu 9.10 (x86) Compiler: gcc 4.4.1 Built successfully. (Quite a lot of strict-aliasing rules warnings thrown up.) However, I did notice that tools/regression/doc/library_status.html refers to the script tools/regression/library_test_all This file doesn't exist: it is actually tools/regression/src/library_test_all.sh Tried to run tests, but ended up with lots of "bad variable name" errors for LD_LIBRARY_PATH thrown up by dash (Ubuntu's /bin/sh). Anybody come across this before? Phil -- Phil Richards, <news@derived-software.ltd.uk>
On 11/02/2009 04:47 PM, Phil Richards wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
Platform: Ubuntu 9.10 (x86) Compiler: gcc 4.4.1
Built successfully. (Quite a lot of strict-aliasing rules warnings thrown up.)
I take it these are all related to boost.python ? Python 3 reportedly has fixed its C API so it can be used with strict aliasing turned on. And, coincidentally, this boost release is the first one to support Python 3. :-) FWIW. Stefan -- ...ich hab' noch einen Koffer in Berlin...
On Mon, 2009-11-02 at 18:45 -0500, Stefan Seefeld wrote:
On 11/02/2009 04:47 PM, Phil Richards wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
Platform: Ubuntu 9.10 (x86) Compiler: gcc 4.4.1
Built successfully. (Quite a lot of strict-aliasing rules warnings thrown up.)
I take it these are all related to boost.python ?
Some are in python, e.g.: gcc.compile.c++ bin.v2/libs/python/build/gcc-4.4.1/release/threading-multi/dict.o ./boost/python/detail/destroy.hpp: In member function ‘bool boost::python::detail::dict_base::has_key(const boost::python::api::object&) const’: ./boost/python/detail/destroy.hpp:33: warning: dereferencing pointer ‘p.190’ does break strict-aliasing rules ./boost/python/detail/destroy.hpp:90: note: initialized from here [...] A number pop up from wave, e.g.: gcc.compile.c++ bin.v2/libs/wave/build/gcc-4.4.1/release/threading-multi/instantiate_cpp_exprgrammar.o In file included from ./boost/wave/wave_config.hpp:217, from libs/wave/src/instantiate_cpp_exprgrammar.cpp:11: ./boost/wave/util/flex_string.hpp: In member function ‘Storage& boost::wave::util::CowString<Storage, Align>::Data() const [with Storage = boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, Align = char*]’: ./boost/wave/util/flex_string.hpp:1312: instantiated from ‘boost::wave::util::CowString<Storage, Align>::~CowString() [with Storage = boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, Align = char*]’ ./boost/wave/util/flex_string.hpp:1519: instantiated from ‘boost::wave::util::flex_string<E, T, A, Storage>::~flex_string() [with E = char, T = std::char_traits<char>, A = std::allocator<char>, Storage = boost::wave::util::CowString<boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, char*>]’ ./boost/wave/grammars/cpp_expression_grammar.hpp:769: instantiated from ‘static bool boost::wave::grammars::expression_grammar_gen<TokenT>::evaluate(const typename std::list<TokenT, boost::fast_pool_allocator<TokenT, boost::default_user_allocator_new_delete, boost::details::pool::pthread_mutex, 32u> >::const_iterator&, const typename std::list<TokenT, boost::fast_pool_allocator<TokenT, boost::default_user_allocator_new_delete, boost::details::pool::pthread_mutex, 32u> >::const_iterator&, const typename TokenT::position_type&, bool, boost::wave::grammars::value_error&) [with TokenT = boost::wave::cpplexer::lex_token<boost::wave::util::file_position<boost::wave::util::flex_string<char, std::char_traits<char>, std::allocator<char>, boost::wave::util::CowString<boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, char*> > > >]’ libs/wave/src/instantiate_cpp_exprgrammar.cpp:41: instantiated from here ./boost/wave/util/flex_string.hpp:1214: warning: dereferencing type-punned pointer will break strict-aliasing rules ./boost/wave/util/flex_string.hpp: In member function ‘void boost::wave::util::CowString<Storage, Align>::MakeUnique() const [with Storage = boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, Align = char*]’: ./boost/wave/util/flex_string.hpp:653: warning: dereferencing pointer ‘<anonymous>’ does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:584: warning: dereferencing pointer ‘<anonymous>’ does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:583: warning: dereferencing pointer ‘<anonymous>’ does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:581: warning: dereferencing pointer ‘<anonymous>’ does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:576: warning: dereferencing pointer ‘<anonymous>’ does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:1243: note: initialized from here ./boost/wave/util/flex_string.hpp:611: warning: dereferencing pointer ‘<anonymous>’ does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:1243: note: initialized from here [... and so on...] This is one of shorter warning messages: one of them is a little over 8000 characters. I have no idea whether these warnings are bogus or not, or whether they actually indicate a real or potential problem... Phil -- Phil Richards, <news@derived-software.ltd.uk>
On Tue, 2009-11-03 at 20:29 +0000, Phil Richards wrote:
On Mon, 2009-11-02 at 18:45 -0500, Stefan Seefeld wrote:
On 11/02/2009 04:47 PM, Phil Richards wrote:
(Quite a lot of strict-aliasing rules warnings thrown up.) I take it these are all related to boost.python ? Some are in python, e.g.: [...] A number pop up from wave, e.g.: [...]
One library that I missed out that I think might be annoying is Boost.Optional. There are lots of warnings like: ./boost/optional/optional.hpp:315: warning: dereferencing pointer ‘<anonymous>’ does break strict-aliasing rules ./boost/optional/optional.hpp:422: note: initialized from here The first line can vary, the initialization point is always the same. Phil -- Phil Richards, <news@derived-software.ltd.uk>
Phil Richards wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
Platform: Ubuntu 9.10 (x86) Compiler: gcc 4.4.1
Built successfully. (Quite a lot of strict-aliasing rules warnings thrown up.)
However, I did notice that tools/regression/doc/library_status.html refers to the script tools/regression/library_test_all
This file doesn't exist: it is actually tools/regression/src/library_test_all.sh
Tried to run tests, but ended up with lots of "bad variable name" errors for LD_LIBRARY_PATH thrown up by dash (Ubuntu's /bin/sh). Anybody come across this before?
Please use --debug-configuration and post the initial output. I did fix such an error before -- it could be that you're actually using system-wide Boost.Build. Or, alternatively, there could be a further bug -- while I am on Kubuntu, I don't use that incompatible-with-everything-brand-new-shell ;-) If that's a further bug, could you run with -n and post a single test run command? - Volodya
On Tue, Nov 3, 2009 at 12:47 AM, Vladimir Prus <vladimir@codesourcery.com>wrote:
Phil Richards wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
Platform: Ubuntu 9.10 (x86) Compiler: gcc 4.4.1
Built successfully. (Quite a lot of strict-aliasing rules warnings thrown up.)
However, I did notice that tools/regression/doc/library_status.html refers to the script tools/regression/library_test_all
This file doesn't exist: it is actually tools/regression/src/library_test_all.sh
Tried to run tests, but ended up with lots of "bad variable name" errors for LD_LIBRARY_PATH thrown up by dash (Ubuntu's /bin/sh). Anybody come across this before?
Please use --debug-configuration and post the initial output. I did fix such an error before -- it could be that you're actually using system-wide Boost.Build. Or, alternatively, there could be a further bug -- while I am on Kubuntu, I don't use that incompatible-with-everything-brand-new-shell ;-) If that's a further bug, could you run with -n and post a single test run command?
Volodya, is this something we should hold the beta for, or can it wait until the release? Thanks, --Beman
Beman Dawes wrote:
On Tue, Nov 3, 2009 at 12:47 AM, Vladimir Prus <vladimir@codesourcery.com>wrote:
Phil Richards wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Thanks,
--Beman
Platform: Ubuntu 9.10 (x86) Compiler: gcc 4.4.1
Built successfully. (Quite a lot of strict-aliasing rules warnings thrown up.)
However, I did notice that tools/regression/doc/library_status.html refers to the script tools/regression/library_test_all
This file doesn't exist: it is actually tools/regression/src/library_test_all.sh
Tried to run tests, but ended up with lots of "bad variable name" errors for LD_LIBRARY_PATH thrown up by dash (Ubuntu's /bin/sh). Anybody come across this before?
Please use --debug-configuration and post the initial output. I did fix such an error before -- it could be that you're actually using system-wide Boost.Build. Or, alternatively, there could be a further bug -- while I am on Kubuntu, I don't use that incompatible-with-everything-brand-new-shell ;-) If that's a further bug, could you run with -n and post a single test run command?
Volodya, is this something we should hold the beta for, or can it wait until the release?
I have just tried on Kubuntu 8.04, using dash as shell, and tests still do work. So, I still think local misconfiguration is a likely reason, and that we should not delay beta. We should be able to figure this for the final release. - Volodya
On Tue, 2009-11-03 at 18:57 +0300, Vladimir Prus wrote:
Beman Dawes wrote:
Volodya, is this something we should hold the beta for, or can it wait until the release? I have just tried on Kubuntu 8.04, using dash as shell, and tests still do work. So, I still think local misconfiguration is a likely reason, and that we should not delay beta. We should be able to figure this for the final release.
Yep - it was me being dumb - I'd managed to lose PATH update that put the freshly built bjam on the path, so I was picking up the system one. The tests are now running - I'll check over the results and post if there are any unexpected failures, but that isn't going to be anytime soon :-) There are a couple of minor updates that should be put in (probably for 1.42) regarding the documentation for building and running the tests in tools/regression/doc/library_status.html: * the wrong path for library_test and library_test_all * mangled HTML in the formatted pages: * ../bin.v2/libs/test/<library%gt;/... * ../libs/%lt;library%gt;/example * the "preliminaries" suggest that the tools are built by just invoking bjam in the tools/regression/build directory: this only builds process_jam_log. It didn't build library_status (had to explicitly build it). Apologies for the misleading error report! Phil -- Phil Richards, <news@derived-software.ltd.uk>
The most worrying looking warnings are from Boost.Function IMO: ./boost/function/function_base.hpp:319: warning: dereferencing type-punned pointer will break strict-aliasing rules ./boost/function/function_base.hpp:323: warning: dereferencing type-punned pointer will break strict-aliasing rules Along with several other more verbose warnings to the same effect (included from program_options). Normally these warnings do genuinely indicate issues that will cause real bugs when optimisations are turned on, but in this case I don't see an obvious fix :-( This is with gcc-4.4.1 on Ubuntu BTW. This one looks like it could be easily fixed too: libs/program_options/src/options_description.cpp:74: warning: suggest explicit braces to avoid ambiguous 'else' as could: libs/program_options/src/value_semantic.cpp:175: warning: suggest parentheses around '&&' within '||' libs/program_options/src/value_semantic.cpp: In function 'void boost::program_options::validate(boost::any&, const std::vector<std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> >, std::allocator<std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > > >&, std::string*, int)': libs/program_options/src/value_semantic.cpp:190: warning: suggest parentheses around '&&' within '||' Cheers, John.
More strict aliasing warnings, this time from spirit: ./boost/spirit/home/classic/core/scanner/scanner.hpp:132: warning: dereferencing pointer 'val' does break strict-aliasing rules Again, this is normally a bug once optimizations are turned on. There are also a lot of warnings from Serialization relating to use of deprecated spirit headers - is there a #define that can be set to suppress these - or a better fix? We don't really want to see them in the library build IMO: ./boost/spirit/core/non_terminal/rule.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_rule.hpp" etc. Cheers, John.
More strict aliasing warnings, this time from spirit:
./boost/spirit/home/classic/core/scanner/scanner.hpp:132: warning: dereferencing pointer 'val' does break strict-aliasing rules
Again, this is normally a bug once optimizations are turned on.
It's the first time I'm seeing this problem in Spirit. We need to investigate this more closely as the warning comes from a template function (and 'val' is a template parameter), so I assume the issue is in the code using Spirit. What file produces this diagnostic?
There are also a lot of warnings from Serialization relating to use of deprecated spirit headers - is there a #define that can be set to suppress these - or a better fix? We don't really want to see them in the library build IMO:
./boost/spirit/core/non_terminal/rule.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_rule.hpp"
This one is easy, just change the included header to the one suggested by the warning. If things need to be unchanged you will need to #define BOOST_SPIRIT_USE_OLD_NAMESPACE as well before including any headers from Spirit. But it would be better to touch the code using spirit. Everything referenced from boost::spirit needs to be referenced from boost::spirit::classic after including the new file (and if the #define is _not_ used). But I guess it's too late for this change now. And just for the records, this warning is in place since Boost V1.38, so I really wonder why the code has not been changed yet. Regards Hartmut ------------------- Meet me at BoostCon http://boostcon.com
./boost/spirit/home/classic/core/scanner/scanner.hpp:132: warning: dereferencing pointer 'val' does break strict-aliasing rules
Again, this is normally a bug once optimizations are turned on.
It's the first time I'm seeing this problem in Spirit. We need to investigate this more closely as the warning comes from a template function (and 'val' is a template parameter), so I assume the issue is in the code using Spirit. What file produces this diagnostic?
It's from the serialization build: gcc.compile.c++ /home/john/bin/boost/bin.v2/libs/serialization/build/gcc-4.4.1/release/link-static/threading-multi/xml_wgrammar.o In file included from ./boost/archive/impl/basic_xml_grammar.hpp:59, from libs/serialization/src/xml_wgrammar.cpp:18: ./boost/spirit/core/non_terminal/rule.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_rule.hpp" In file included from libs/serialization/src/basic_xml_grammar.ipp:20, from libs/serialization/src/xml_wgrammar.cpp:146: ./boost/spirit/core/composite/operators.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_operators.hpp" In file included from libs/serialization/src/basic_xml_grammar.ipp:21, from libs/serialization/src/xml_wgrammar.cpp:146: ./boost/spirit/core/composite/actions.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_actions.hpp" In file included from libs/serialization/src/basic_xml_grammar.ipp:22, from libs/serialization/src/xml_wgrammar.cpp:146: ./boost/spirit/core/primitives/numerics.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_numerics.hpp" In file included from ./boost/spirit/home/classic/core/non_terminal/rule.hpp:33, from ./boost/spirit/include/classic_rule.hpp:11, from ./boost/spirit/core/non_terminal/rule.hpp:25, from ./boost/archive/impl/basic_xml_grammar.hpp:59, from libs/serialization/src/xml_wgrammar.cpp:18: ./boost/spirit/home/classic/core/scanner/scanner.hpp: In member function 'typename boost::spirit::match_result<ScannerT, AttrT>::type boost::spirit::impl::concrete_parser<ParserT, ScannerT, AttrT>::do_parse_virtual(const ScannerT&) const [with ParserT = boost::spirit::alternative<boost::spirit::alternative<boost::spirit::alternative<boost::spirit::rule<boost::spirit::scanner<__gnu_cxx::__normal_iterator<wchar_t*, std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > >, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> >, boost::spirit::nil_t, boost::spirit::nil_t>, boost::spirit::rule<boost::spirit::scanner<__gnu_cxx::__normal_iterator<wchar_t*, std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > >, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> >, boost::spirit::nil_t, boost::spirit::nil_t> >, boost::spirit::rule<boost::spirit::scanner<__gnu_cxx::__normal_iterator<wchar_t*, std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > >, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> >, boost::spirit::nil_t, boost::spirit::nil_t> >, boost::spirit::action<boost::spirit::chset<wchar_t>, boost::archive::xml::append_char<std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > > > >, ScannerT = boost::spirit::scanner<__gnu_cxx::__normal_iterator<wchar_t*, std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > >, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> >, AttrT = boost::spirit::nil_t]': ./boost/spirit/home/classic/core/scanner/scanner.hpp:132: warning: dereferencing pointer 'val' does break strict-aliasing rules ./boost/optional/optional.hpp:422: note: initialized from here Note that this warning only shows up in release mode. Harmut, while I have your attention, maybe the largest slew of warnings (in terms of lines of output) are coming from wave, the first one reads: gcc.compile.c++ /home/john/bin/boost/bin.v2/libs/wave/build/gcc-4.4.1/release/threading-multi/instantiate_cpp_exprgrammar.o In file included from ./boost/wave/wave_config.hpp:217, from libs/wave/src/instantiate_cpp_exprgrammar.cpp:11: ./boost/wave/util/flex_string.hpp: In member function 'Storage& boost::wave::util::CowString<Storage, Align>::Data() const [with Storage = boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, Align = char*]': ./boost/wave/util/flex_string.hpp:1312: instantiated from 'boost::wave::util::CowString<Storage, Align>::~CowString() [with Storage = boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, Align = char*]' ./boost/wave/util/flex_string.hpp:1519: instantiated from 'boost::wave::util::flex_string<E, T, A, Storage>::~flex_string() [with E = char, T = std::char_traits<char>, A = std::allocator<char>, Storage = boost::wave::util::CowString<boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, char*>]' ./boost/wave/grammars/cpp_expression_grammar.hpp:769: instantiated from 'static bool boost::wave::grammars::expression_grammar_gen<TokenT>::evaluate(const typename std::list<TokenT, boost::fast_pool_allocator<TokenT, boost::default_user_allocator_new_delete, boost::details::pool::pthread_mutex, 32u> >::const_iterator&, const typename std::list<TokenT, boost::fast_pool_allocator<TokenT, boost::default_user_allocator_new_delete, boost::details::pool::pthread_mutex, 32u> >::const_iterator&, const typename TokenT::position_type&, bool, boost::wave::grammars::value_error&) [with TokenT = boost::wave::cpplexer::lex_token<boost::wave::util::file_position<boost::wave::util::flex_string<char, std::char_traits<char>, std::allocator<char>, boost::wave::util::CowString<boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, char*> > > >]' libs/wave/src/instantiate_cpp_exprgrammar.cpp:41: instantiated from here ./boost/wave/util/flex_string.hpp:1214: warning: dereferencing type-punned pointer will break strict-aliasing rules ./boost/wave/util/flex_string.hpp: In member function 'void boost::wave::util::CowString<Storage, Align>::MakeUnique() const [with Storage = boost::wave::util::AllocatorStringStorage<char, std::allocator<char> >, Align = char*]': ./boost/wave/util/flex_string.hpp:653: warning: dereferencing pointer '<anonymous>' does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:584: warning: dereferencing pointer '<anonymous>' does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:583: warning: dereferencing pointer '<anonymous>' does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:581: warning: dereferencing pointer '<anonymous>' does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:576: warning: dereferencing pointer '<anonymous>' does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:1243: note: initialized from here ./boost/wave/util/flex_string.hpp:611: warning: dereferencing pointer '<anonymous>' does break strict-aliasing rules ./boost/wave/util/flex_string.hpp:1243: note: initialized from here I'm not going to attempt to post the rest as there are 100's of K of them :-( HTH, John.
./boost/spirit/home/classic/core/scanner/scanner.hpp:132: warning: dereferencing pointer 'val' does break strict-aliasing rules
Again, this is normally a bug once optimizations are turned on.
It's the first time I'm seeing this problem in Spirit. We need to investigate this more closely as the warning comes from a template function (and 'val' is a template parameter), so I assume the issue is in the code using Spirit. What file produces this diagnostic?
It's from the serialization build:
[snipped warnings] Will investigate.
Note that this warning only shows up in release mode.
Harmut, while I have your attention, maybe the largest slew of warnings (in terms of lines of output) are coming from wave, the first one reads:
gcc.compile.c++
[snipped warnings]
I'm not going to attempt to post the rest as there are 100's of K of them :-(
Hmm, I'm not seeing these warnings (Windows, gcc 4.4.1, MingW) while compiling with -O3. What other command line options do I need to specify? Thanks! Regards Hartmut ------------------- Meet me at BoostCon http://boostcon.com
Hmm, I'm not seeing these warnings (Windows, gcc 4.4.1, MingW) while compiling with -O3. What other command line options do I need to specify?
"g++" -ftemplate-depth-128 -O3 -finline-functions -Wno-inline -Wall -pthread -DBOOST_ALL_NO_LIB=1 -DNDEBUG -I"." -c -o "/home/john/bin/boost/bin.v2/libs/wave/build/gcc-4.4.1/release/link-static/threading-multi/instantiate_cpp_exprgrammar.o" "libs/wave/src/instantiate_cpp_exprgrammar.cpp" With gcc-4.4.1. I see so many pages of warnings from this that I gave up counting :-( Looks to be an issue new to 4.4.1 and/or the new Ubuntu-9.1 release? HTH, John.
Hmm, I'm not seeing these warnings (Windows, gcc 4.4.1, MingW) while compiling with -O3. What other command line options do I need to specify?
"g++" -ftemplate-depth-128 -O3 -finline-functions -Wno-inline - Wall -pthread -DBOOST_ALL_NO_LIB=1 -DNDEBUG -I"." -c -o "/home/john/bin/boost/bin.v2/libs/wave/build/gcc-4.4.1/release/link- static/threading-multi/instantiate_cpp_exprgrammar.o" "libs/wave/src/instantiate_cpp_exprgrammar.cpp"
With gcc-4.4.1.
I see so many pages of warnings from this that I gave up counting :-(
Ok, got it. I started to address those but there are still a couple of the strict aliasing warnings to fix. I'll work on that asap, even if I'm not sure how to address the warnings yet. I just was not aware these warnings get generated... Regards Hartmut ------------------- Meet me at BoostCon http://boostcon.com
./boost/spirit/home/classic/core/scanner/scanner.hpp:132: warning: dereferencing pointer 'val' does break strict-aliasing rules
Again, this is normally a bug once optimizations are turned on.
It's the first time I'm seeing this problem in Spirit. We need to investigate this more closely as the warning comes from a template function (and 'val' is a template parameter), so I assume the issue is in the code using Spirit. What file produces this diagnostic?
It's from the serialization build:
gcc.compile.c++ /home/john/bin/boost/bin.v2/libs/serialization/build/gcc- 4.4.1/release/link-static/threading-multi/xml_wgrammar.o In file included from ./boost/archive/impl/basic_xml_grammar.hpp:59, from libs/serialization/src/xml_wgrammar.cpp:18: ./boost/spirit/core/non_terminal/rule.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_rule.hpp" In file included from libs/serialization/src/basic_xml_grammar.ipp:20, from libs/serialization/src/xml_wgrammar.cpp:146: ./boost/spirit/core/composite/operators.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_operators.hpp" In file included from libs/serialization/src/basic_xml_grammar.ipp:21, from libs/serialization/src/xml_wgrammar.cpp:146: ./boost/spirit/core/composite/actions.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_actions.hpp" In file included from libs/serialization/src/basic_xml_grammar.ipp:22, from libs/serialization/src/xml_wgrammar.cpp:146: ./boost/spirit/core/primitives/numerics.hpp:18:4: warning: #warning "This header is deprecated. Please use: boost/spirit/include/classic_numerics.hpp" In file included from ./boost/spirit/home/classic/core/non_terminal/rule.hpp:33, from ./boost/spirit/include/classic_rule.hpp:11, from ./boost/spirit/core/non_terminal/rule.hpp:25, from ./boost/archive/impl/basic_xml_grammar.hpp:59, from libs/serialization/src/xml_wgrammar.cpp:18: ./boost/spirit/home/classic/core/scanner/scanner.hpp: In member function 'typename boost::spirit::match_result<ScannerT, AttrT>::type boost::spirit::impl::concrete_parser<ParserT, ScannerT, AttrT>::do_parse_virtual(const ScannerT&) const [with ParserT = boost::spirit::alternative<boost::spirit::alternative<boost::spirit::al ternative<boost::spirit::rule<boost::spirit::scanner<__gnu_cxx::__norma l_iterator<wchar_t*, std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > >, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> >, boost::spirit::nil_t, boost::spirit::nil_t>, boost::spirit::rule<boost::spirit::scanner<__gnu_cxx::__normal_iterator <wchar_t*, std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > >, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> >, boost::spirit::nil_t, boost::spirit::nil_t> >, boost::spirit::rule<boost::spirit::scanner<__gnu_cxx::__normal_iterator <wchar_t*, std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > >, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> >, boost::spirit::nil_t, boost::spirit::nil_t> >, boost::spirit::action<boost::spirit::chset<wchar_t>, boost::archive::xml::append_char<std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > > > >, ScannerT = boost::spirit::scanner<__gnu_cxx::__normal_iterator<wchar_t*, std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > >, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> >, AttrT = boost::spirit::nil_t]': ./boost/spirit/home/classic/core/scanner/scanner.hpp:132: warning: dereferencing pointer 'val' does break strict-aliasing rules ./boost/optional/optional.hpp:422: note: initialized from here
This one is caused by the code in optional.hpp (I'm seeing similar warnings from optional in Wave, BTW), so it needs to be fixed there... Regards Hartmut ------------------- Meet me at BoostCon http://boostcon.com
John Maddock wrote:
The most worrying looking warnings are from Boost.Function IMO:
./boost/function/function_base.hpp:319: warning: dereferencing type-punned pointer will break strict-aliasing rules ./boost/function/function_base.hpp:323: warning: dereferencing type-punned pointer will break strict-aliasing rules
Along with several other more verbose warnings to the same effect (included from program_options).
Normally these warnings do genuinely indicate issues that will cause real bugs when optimisations are turned on, but in this case I don't see an obvious fix :-(
This is with gcc-4.4.1 on Ubuntu BTW.
This one looks like it could be easily fixed too:
libs/program_options/src/options_description.cpp:74: warning: suggest explicit braces to avoid ambiguous 'else'
This is actually fixed in trunk.
as could:
libs/program_options/src/value_semantic.cpp:175: warning: suggest parentheses around '&&' within '||' libs/program_options/src/value_semantic.cpp: In function 'void boost::program_options::validate(boost::any&, const std::vector<std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> >, std::allocator<std::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > > >&, std::string*, int)': libs/program_options/src/value_semantic.cpp:190: warning: suggest parentheses around '&&' within '||'
I've just fixed this on trunk (and discovered and fixed a real bug in the process). Thanks, Volodya
Cheers, John.
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
On Tue, 2009-11-03 at 20:10 +0000, Phil Richards wrote:
The tests are now running - I'll check over the results and post if there are any unexpected failures, but that isn't going to be anytime soon :-)
My test run has completed (gcc 4.4.1, Ubuntu 9.10, x86, release - no MPI - I've not managed to get bjam to recognise that MPI is available yet - more configuration problems on my part, I suspect). Every test in spirit/phoenix was marked as a failure: looking at the failure report, none of them actually failed. (All had a success report for the test.) There are a large number of warnings (from the compiler warnings already posted by myself and John Maddock), but there also seem to be quite a lot of errors in maths related tests. I suspect that most of these problems are purely related to the fact that this is a release build with optimization turned on. I wouldn't have thought that any of these are ones that would affect the release. For example, in accumulators: ...libs-accumulators-test-pot_quantile.test-gcc-4.4.1-release-link-static-threading-multi Run output: Running 1 test case... pot_quantile.cpp(82): error in "test_stat": difference{1.0227%} between quantile(acc5, quantile_probability = 0.999){6.9786481796869042} and 6.908{6.9080000000000004} exceeds 1% *** 1 failure detected in test suite "Master Test Suite" EXIT STATUS: 201 ...libs-accumulators-test-weighted_tail_mean.test-gcc-4.4.1-release-link-static-threading-multi Run output: Running 1 test case... weighted_tail_mean.cpp(61): error in "test_stat": difference{1.11208%} between non_coherent_weighted_tail_mean(acc1, quantile_probability = 0.025){0.0123609897219755} and 0.0125{0.012500000000000001} exceeds 1% weighted_tail_mean.cpp(62): error in "test_stat": difference{2.44744%} between non_coherent_weighted_tail_mean(acc1, quantile_probability = 0.01){0.0048776278773278257} and 0.005{0.0050000000000000001} exceeds 1% *** 2 failures detected in test suite "Master Test Suite" EXIT STATUS: 201 (Plus more) And in maths: libs-math-test-test_normal.test-gcc-4.4.1-release-link-static-threading-multi Run output: Running 1 test case... Tolerance for type f is 0.01 % test_normal.cpp(233): error in "test_main_caller( argc, argv )": difference{1.09314e-05%} between hazard(dist, x){0.00425984897} and pdf(dist, x) / cdf(complement(dist, x)){0.00425984943} exceeds 5.96046448e-07% Tolerance for type d is 0.01 % Tolerance for type e is 0.01 % Tolerance for type N5boost4math8concepts12real_conceptE is 0.01 % *** 1 failure detected in test suite "Test Program" EXIT STATUS: 201 (Plus more) And in numeric/ublas: libs-numeric-ublas-test-test_lu.test-gcc-4.4.1-release-link-static-threading-multi Run output: factorization completed: 1 LU factors are correct: 0 permutation is correct: 1 inverse is correct: 0 EXIT STATUS: 1 Hope this is of some use, Phil -- Phil Richards, <news@derived-software.ltd.uk>
Phil Richards wrote:
On Tue, 2009-11-03 at 20:10 +0000, Phil Richards wrote:
The tests are now running - I'll check over the results and post if there are any unexpected failures, but that isn't going to be anytime soon :-)
My test run has completed (gcc 4.4.1, Ubuntu 9.10, x86, release - no MPI - I've not managed to get bjam to recognise that MPI is available yet - more configuration problems on my part, I suspect).
Every test in spirit/phoenix was marked as a failure: looking at the failure report, none of them actually failed. (All had a success report for the test.)
You lost me there. I don't know what the problem is. Regards, -- Joel de Guzman http://www.boostpro.com http://spirit.sf.net http://www.facebook.com/djowel Meet me at BoostCon http://www.boostcon.com/home http://www.facebook.com/boostcon
On Thu, 2009-11-05 at 07:56 +0800, Joel de Guzman wrote:
Phil Richards wrote:
Every test in spirit/phoenix was marked as a failure: looking at the failure report, none of them actually failed. (All had a success report for the test.) You lost me there. I don't know what the problem is.
Not exactly my clearest written description, but basically, the summary report from the test runs gives a table which looks a bit like: Test Name gcc-4.4.1 release link-static threading-multi threading-multi algorithm bind container core detail function object operator scope statement algorithm bind container core detail function object operator scope statement arithmetic_tests Fail Missing Missing Missing Missing Missing Missing Missing Fail Missing Missing Fail Missing Missing Missing Missing Missing Missing Missing Fail Missing Missing bind_function_object_tests Fail Missing Fail Missing Missing Missing Missing Missing Missing Missing Missing Fail Missing Fail Missing Missing Missing Missing Missing Missing Missing Missing (I suspect this will be horribly mangled by email, but basically there are lots of columns, most with "Missing", and every other entry is "Fail".) Clicking on a "Fail" (for example the first one), leads to: libs-spirit-phoenix-test-arithmetic_tests.test-gcc-4.4.1-release-link-static-threading-multi Run output: No errors detected. EXIT STATUS: 0 So the library_status.html for phoenix is wrongly generated: there are no errors. Hope this makes slightly more sense, Phil -- Phil Richards, <news@derived-software.ltd.uk>
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
How do I get to a list of which libraries have been added or modified in this release? Thanks, Michael Goldshteyn
Download and unpack fine - Vista, MSVC 9.0. It seems to have grown ;-) Rebuilt a few things... NO showstoppers :-)) but a few minor observations... As usual was confused by instructions on building libraries. No warning that it will take quite a long time. I suspect many Windows users may not need all the versions of all the libraries. Regex and Test and Datetime most popular? I repeat my suggestion that a few well-chosen command files would be *much* more immediately useful and reduce the steady stream of cries for help on the mailing list. Not helped by missing link to Boost.Build documentation: File "boost-build/doc/html/bbv2/advanced/invocation.html"not found. From: "For a complete description of these and other invocation options, please see the Boost.Build documentation." As usual a slew of warnings :-( "warning C4702: unreachable code" - in Boost.Test Shall I submit Trak for these? "warning C4180: qualifier applied to function type has no meaning; ignored" in math library (mea culpa - submit Trak to self!) and C4180 in Fusion too. This C4180 seems such an unhelpful warning that I wonder if it should be 'globally' suppressed in Boost config. Does anyone consider C4180 a useful warning to them? It is a problem on other platforms? Paul --- Paul A. Bristow Prizet Farmhouse Kendal, UK LA8 8AB +44 1539 561830, mobile +44 7714330204 pbristow@hetp.u-net.com
-----Original Message----- From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Beman Dawes Sent: Monday, November 02, 2009 12:11 PM To: Boost Developers List Subject: [boost] [1.41.0] Beta 1 release candidate
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
Paul A. Bristow wrote:
Download and unpack fine - Vista, MSVC 9.0. It seems to have grown ;-)
Rebuilt a few things...
NO showstoppers :-))
but a few minor observations...
As usual was confused by instructions on building libraries.
What exactly was confused? Did you follow "Simplified Build from Source", or something else?
No warning that it will take quite a long time.
Where do you suggest to print/show such a warning?
I suspect many Windows users may not need all the versions of all the libraries. Regex and Test and Datetime most popular?
I repeat my suggestion that a few well-chosen command files would be *much* more immediately useful and reduce the steady stream of cries for help on the mailing list.
Can you elaborate?
Not helped by missing link to Boost.Build documentation:
File "boost-build/doc/html/bbv2/advanced/invocation.html"not found.
From: "For a complete description of these and other invocation options, please see the Boost.Build documentation."
This is a strange hiccup of the web site. The nightly build of Boost.Build, which is pulled by the site and then served, is just fine. - Volodya
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
There are lots of warnings, but these look a bit scary: gcc.compile.c++ bin.v2/libs/test/build/gcc-4.4.1/release/link-static/debug.o In file included from libs/test/src/debug.cpp:17: /usr/lib/gcc/x86_64-redhat- linux/4.4.1/../../../../include/c++/4.4.1/bits/stl_algo.h: In function ‘char* boost::debug::<unnamed>::prepare_window_title(const boost::debug::dbg_startup_info&)’: /usr/lib/gcc/x86_64-redhat- linux/4.4.1/../../../../include/c++/4.4.1/bits/stl_algo.h:186: warning: array subscript is above array bounds
AMDG Neal Becker wrote:
Beman Dawes wrote:
1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems.
There are lots of warnings, but these look a bit scary:
gcc.compile.c++ bin.v2/libs/test/build/gcc-4.4.1/release/link-static/debug.o In file included from libs/test/src/debug.cpp:17: /usr/lib/gcc/x86_64-redhat- linux/4.4.1/../../../../include/c++/4.4.1/bits/stl_algo.h: In function ‘char* boost::debug::<unnamed>::prepare_window_title(const boost::debug::dbg_startup_info&)’: /usr/lib/gcc/x86_64-redhat- linux/4.4.1/../../../../include/c++/4.4.1/bits/stl_algo.h:186: warning: array subscript is above array bounds
It looks like the compiler is confused by the unrolling in std::find, but I'm not absolutely sure as I don't get this warning with mingw. In Christ, Steven Watanabe
participants (16)
-
Beman Dawes -
Daniel James -
gmane -
Hartmut Kaiser -
Henning Basold -
joaquin＠tid.es -
Joel de Guzman -
John Maddock -
K. Noel Belcourt -
Michael Caisse -
Neal Becker -
Paul A. Bristow -
Phil Richards -
Stefan Seefeld -
Steven Watanabe -
Vladimir Prus