July 2012 - Boost - lists.preview.boost.org

Macro fo rBOOST_NO_NUMERIC_LIMITS_MAX_DIGITS10 or use BOOST_NO_NUMERIC_LIMITS_LOWEST
by Paul A. Bristow 13 Jul '12

13 Jul '12

I have recently used BOOST_NO_NUMERIC_LIMITS_LOWEST to act as a proxy for support of numeric_limits<T>:: max_digits10 (as correctly suggested by John Maddock). Do we need yet another new macro for this, say BOOST_NO_NUMERIC_LIMITS_MAX_DIGITS10 or continue to use BOOST_NO_NUMERIC_LIMITS_LOWEST as a proxy. I think the latter will suffice (despite its unfortunate name for this task). But I think an addition to the configuration macro reference docs would be useful. Perhaps in macro_reference.qbk [[`BOOST_NO_NUMERIC_LIMITS_LOWEST`][Standard library][ Static function `numeric_limits<T>::lowest()` is not available for use. Also acts as a proxy for static const int `numeric_limits<T>:: max_digits10;` is not available for use. ]] I suspect there are also other additions that went with these two items, but I'm having a senior moment about what they are ;-( Views? Paul --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow(a)hetp.u-net.com

4 12

Running the regression tests
by Sergey Cheban 12 Jul '12

12 Jul '12

On 09.07.2012 10:26, Tim Blechmann wrote: > .... there are still some test failures on win32, but i think i found > the reason and fixed the tests (btw it is rather time consuming to > work with the test farm as the win32 tests are only run about once a > week). I'm trying to setup the regression tests on our build server and I have some problems with specifying the toolset(s). The command line is: C:\Python27\python.exe run.py --runner=MSVC9_x64_DoctorWeb --toolsets=msvc-9.0 --bjam-options=architecture=x86 --bjam-options=address-model=64 --tag=branches/release --force-update --bjam-options="-j 10" In the bjam.log, I see: warning: No toolsets are configured. warning: Configuring default toolset "msvc". warning: If the default is wrong, your build may not work correctly. warning: Use the "toolset=xxxxx" option to override our guess. warning: For more configuration options, please consult warning: http://boost.org/boost-build2/doc/html/bbv2/advanced/configuration.html notice: [msvc-cfg] msvc-10.0 detected, command: 'C:\program files (x86)\microsoft visual studio 10.0\VC\bin\cl.exe' notice: [msvc-cfg] msvc-9.0 detected, command: 'C:\program files (x86)\microsoft visual studio 9.0\VC\bin\cl.exe' notice: will use 'C:\program files (x86)\microsoft visual studio 10.0\VC\bin\cl.exe' for msvc, condition <toolset>msvc-10.0 notice: [msvc-cfg] condition: '<toolset>msvc-10.0/<architecture>/<address-model>', setup: 'call "C:\program files (x86)\microsoft visual studio 10.0\VC\vcvarsall.bat" x86 >nul ' notice: [msvc-cfg] condition: '<toolset>msvc-10.0/<architecture>/<address-model>32', setup: 'call "C:\program files (x86)\microsoft visual studio 10.0\VC\vcvarsall.bat" x86 >nul ' notice: [msvc-cfg] condition: '<toolset>msvc-10.0/<architecture>x86/<address-model>', setup: 'call "C:\program files (x86)\microsoft visual studio 10.0\VC\vcvarsall.bat" x86 >nul ' notice: [msvc-cfg] condition: '<toolset>msvc-10.0/<architecture>x86/<address-model>32', setup: 'call "C:\program files (x86)\microsoft visual studio 10.0\VC\vcvarsall.bat" x86 >nul ' notice: [msvc-cfg] condition: '<toolset>msvc-10.0/<architecture>/<address-model>64', setup: 'call "C:\program files (x86)\microsoft visual studio 10.0\VC\vcvarsall.bat" x86_amd64 >nul ' notice: [msvc-cfg] condition: '<toolset>msvc-10.0/<architecture>x86/<address-model>64', setup: 'call "C:\program files (x86)\microsoft visual studio 10.0\VC\vcvarsall.bat" x86_amd64 >nul ' notice: [msvc-cfg] condition: '<toolset>msvc-10.0/<architecture>ia64/<address-model>', setup: 'call "C:\program files (x86)\microsoft visual studio 10.0\VC\vcvarsall.bat" x86_ia64 >nul ' notice: [msvc-cfg] condition: '<toolset>msvc-10.0/<architecture>ia64/<address-model>64', setup: 'call "C:\program files (x86)\microsoft visual studio 10.0\VC\vcvarsall.bat" x86_ia64 >nul ' C:/Users/BoostTest/Release/tools_bb/build\feature.jam:308: in implied-subfeature from module feature error: "9.0" is not a known subfeature value of <toolset>msvc C:/Users/BoostTest/Release/tools_bb/build\feature.jam:491: in validate-value-string from module feature C:/Users/BoostTest/Release/tools_bb/build\feature.jam:358: in expand-subfeatures-aux from module feature C:/Users/BoostTest/Release/tools_bb/build\feature.jam:419: in feature.expand-subfeatures from module feature C:/Users/BoostTest/Release/tools_bb/build\build-request.jam:20: in apply-to-property-set from module build-request C:/Users/BoostTest/Release/tools_bb/kernel\modules.jam:103: in modules.call-in from module build-request C:/Users/BoostTest/Release/tools_bb/util\sequence.jam:48: in sequence.transform from module sequence C:/Users/BoostTest/Release/tools_bb/build\build-request.jam:32: in build-request.expand-no-defaults from module build-request C:/Users/BoostTest/Release/tools_bb\build-system.jam:630: in load from module build-system C:\Users\BoostTest\Release\tools_bb/kernel\modules.jam:283: in import from module modules C:\Users\BoostTest\Release\tools_bb/kernel/bootstrap.jam:142: in boost-build from module C:\Users\BoostTest\Release\boost\boost-build.jam:17: in module scope from module What am I doing wrong? -- Sergey Cheban

3 3

[trac] Version defaults to 1.51
by Daniel James 12 Jul '12

12 Jul '12

Hi, When you create a new ticket, the version is set to 1.51 by default, which hasn't been released yet. It should be set to 1.50, can someone change it please? thanks, Daniel

1 0

[Algorithm] Incorrect behavior of is_sorted_until (predicate version)
by Michel Morin 12 Jul '12

12 Jul '12

Hi Marshall, Suppose we have an array int ar[5] = {1, 2, 2, 2, 2}; Then, `is_sorted_until(ar, ar + 5, std::less<int>())` should return `ar`'s end pointer. But it gives `ar + 2`. To fix this, change line 49 of is_sorted.hpp if ( !p ( *first, *next )) return next; to if ( p ( *next, *first )) return next; After fixing it, don't forget to change the predicate used in non-predicate version of is_sorted_until ;) Spotted by "clang-cxx11-r159746" test runner. Regards, Michel

2 1

[lockfree] include into 1.51?
by Tim Blechmann 11 Jul '12

11 Jul '12

hi all, especially release managers, boost.lockfree is now in trunk and it passes the tests on most supported platforms. atm it includes a private copy of boost.atomic, which will be removed once boost.atomic is merged (helge wanted to do it quite soon) ... there are still some test failures on win32, but i think i found the reason and fixed the tests (btw it is rather time consuming to work with the test farm as the win32 tests are only run about once a week). iac, the deadline for new libraries in 1.51 is today, but from my side, boost.lockfree could be included as it is now. however if the release managers consider it as too late, i'm don't have a problem with waiting for another release cycle ... thoughts? tim

5 29

Who runs the Cray test runner?
by Paul A. Bristow 11 Jul '12

11 Jul '12

I note that the Cray tests are all failing. Concerned that this is possibly (though I think not entirely from look of error messages) due to changes I have made to Boost.Test recently, I'd like to contact whoever manages the Sandia Cray test runner. Does anyone know who to contact? Thanks Paul --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow(a)hetp.u-net.com

2 2

[container] Forward declarations for scoped allocator traits
by Daniel James 11 Jul '12

11 Jul '12

Hi Ion, I was just looking at the scoped allocator support. A container currently needs to include all of scoped_allocator.hpp just to customize the appropriate traits which seems a bit much. Could you declare (not define) constructible_with_allocator_suffix and constructible_with_allocator_prefix in a separate header so that they can be specialised without including all the other things? I've also attached a patch that fixes a couple of typos that I noticed. Daniel

2 2

[config] Add BOOST_NO_RANGE_BASED_FOR macro?
by Beman Dawes 10 Jul '12

10 Jul '12

I'm about to add Boost.Filesystem support for C ++11 range-based for statements, and would like to include direct test cases for compilers that have the feature. Is it OK to add a BOOST_NO_RANGE_BASED_FOR macro to trunk? --Beman

7 19

[xpressive] Compiler error on VC7.1
by Robert Dailey 10 Jul '12

10 Jul '12

Hi, I'm using Boost 1.49 Xpressive on VC7.1. When I compile the following code, I get compiler errors (these do not appear on VC9): sregex const& rex = bos >> _serverName >> !as_xpr("_GWDomino") >> ".diagnostics" >> *(as_xpr(".base")|".renamed"); 'rex' is really a function parameter in my real world use case. I'm passing the entire static regex into a function that takes a 'sregex const&'. I get the compiler error below. Anyone know why this is happening? C:\Code\work\cmake-decomp\build-vc7\third_party\boost\1.49.0-stlport-5.2.1\include\boost\proto\traits.hpp(691) : error C2664: 'boost::proto::exprns_::expr<Tag,Args,Arity>::expr(const boost::proto::exprns_::expr<Tag,Args,Arity> &)' : cannot convert parameter 1 from 'boost::proto::detail::as_expr<T,Generator,WantsBasicExpr>::result_type' to 'const boost::proto::exprns_::expr<Tag,Args,Arity> &' with [ Tag=boost::proto::tagns_::tag::terminal, Args=boost::proto::argsns_::term<boost::proto::detail::as_expr<const char [10],boost::proto::detail::base_generator<boost::proto::default_generator>::type,false>::value_type>, Arity=0 ] and [ T=const char [10], Generator=boost::proto::detail::base_generator<boost::proto::default_generator>::type, WantsBasicExpr=false ] and [ Tag=boost::proto::tagns_::tag::terminal, Args=boost::proto::argsns_::term<boost::proto::detail::as_expr<const char [10],boost::proto::detail::base_generator<boost::proto::default_generator>::type,false>::value_type>, Arity=0 ] Reason: cannot convert from 'boost::proto::detail::as_expr<T,Generator,WantsBasicExpr>::result_type' to 'const boost::proto::exprns_::expr<Tag,Args,Arity>' with [ T=const char [10], Generator=boost::proto::detail::base_generator<boost::proto::default_generator>::type, WantsBasicExpr=false ] and [ Tag=boost::proto::tagns_::tag::terminal, Args=boost::proto::argsns_::term<boost::proto::detail::as_expr<const char [10],boost::proto::detail::base_generator<boost::proto::default_generator>::type,false>::value_type>, Arity=0 ] No constructor could take the source type, or constructor overload resolution was ambiguous \Code\work\cmake-decomp\server\exchange\tools\uploadlog\uploader.cpp(1372) : see reference to function template instantiation 'const boost::add_const<T>::type boost::proto::functional::as_expr<>::operator ()<char,10>(const char (&)[10]) const' being compiled with [ T=boost::proto::detail::as_expr<const char [10],boost::proto::detail::base_generator<boost::proto::default_generator>::type,false>::result_type ]

3 18

[thread][chrono][system] fails to compile with -fno-exceptions
by Vicente J. Botet Escriba 10 Jul '12

10 Jul '12

Hi, I'm trying to make Boost.Thread work when -fno-exceptions is defined and I getting a lot of errors on Boost.System cd libs/system/build bjam toolset=darwin-4.7.0 -j4 cxxflags="-fno-exceptions" define=BOOST_NO_EXCEPTION darwin.compile.c++ ../../../bin.v2/libs/system/build/darwin-4.7.0/debug/error_code.o ../../../libs/system/src/error_code.cpp: In member function ‘virtual std::string {anonymous}::generic_error_category::message(int) const’: ../../../libs/system/src/error_code.cpp:149:10: error: expected ‘catch’ before ‘(’ token ../../../libs/system/src/error_code.cpp:149:12: error: expected type-specifier before ‘sz’ ../../../libs/system/src/error_code.cpp:149:12: error: exception handling disabled, use -fexceptions to enable ../../../libs/system/src/error_code.cpp:149:15: error: expected ‘)’ before ‘>’ token ../../../libs/system/src/error_code.cpp:149:15: error: expected ‘{’ before ‘>’ token ../../../libs/system/src/error_code.cpp:149:15: error: expected primary-expression before ‘>’ token ../../../libs/system/src/error_code.cpp:149:29: error: expected ‘;’ before ‘)’ token In file included from ../../../libs/system/src/error_code.cpp:19:0: ../../../boost/system/error_code.hpp: At global scope: ../../../boost/system/error_code.hpp:214:36: warning: ‘boost::system::posix_category’ defined but not used [-Wunused-variable] ../../../boost/system/error_code.hpp:215:36: warning: ‘boost::system::errno_ecat’ defined but not used [-Wunused-variable] ../../../boost/system/error_code.hpp:216:36: warning: ‘boost::system::native_ecat’ defined but not used [-Wunused-variable] "/usr/local/gcc-4-7-svn/bin/g++" -ftemplate-depth-128 -O0 -fno-inline -Wall -g -dynamic -gdwarf-2 -fexceptions -fPIC -fno-exceptions -DBOOST_ALL_NO_LIB=1 -DBOOST_NO_EXCEPTION -DBOOST_SYSTEM_DYN_LINK=1 -I"../../.." -c -o "../../../bin.v2/libs/system/build/darwin-4.7.0/debug/error_code.o" "../../../libs/system/src/error_code.cpp" ...failed darwin.compile.c++ ../../../bin.v2/libs/system/build/darwin-4.7.0/debug/error_code.o... ...skipped <p../../../bin.v2/libs/system/build/darwin-4.7.0/debug>libboost_system.dylib for lack of <p../../../bin.v2/libs/system/build/darwin-4.7.0/debug>error_code.o... ...failed updating 1 target... After applying this patch the library builds Index: ../../../libs/system/src/error_code.cpp =================================================================== --- ../../../libs/system/src/error_code.cpp (revision 79245) +++ ../../../libs/system/src/error_code.cpp (working copy) @@ -133,7 +133,9 @@ } } std::string msg; +# ifndef BOOST_NO_EXCEPTIONS try +# endif { msg = ( ( result == invalid_argument ) ? "Unknown error" : bp ); } cd libs/chrono/build bjam toolset=darwin-4.7.0 -j4 cxxflags="-fno-exceptions" define=BOOST_NO_EXCEPTION darwin.compile.c++ ../../../bin.v2/libs/chrono/build/darwin-4.7.0/debug/threading-multi/process_cpu_clocks.o In file included from ../../../boost/chrono/detail/inlined/process_cpu_clocks.hpp:21:0, from ../../../libs/chrono/src/process_cpu_clocks.cpp:17: ../../../boost/system/system_error.hpp: In member function ‘virtual const char* boost::system::system_error::what() const’: ../../../boost/system/system_error.hpp:73:16: error: exception handling disabled, use -fexceptions to enable After applying this patch Index: ../../../boost/system/system_error.hpp =================================================================== --- ../../../boost/system/system_error.hpp (revision 79245) +++ ../../../boost/system/system_error.hpp (working copy) @@ -21,7 +21,7 @@ class BOOST_SYMBOL_VISIBLE system_error : public std::runtime_error // BOOST_SYMBOL_VISIBLE is needed by GCC to ensure system_error thrown from a shared - // library can be caught. See svn.boost.org/trac/boost/ticket/3697 + // library can be caught. See svn.boost.org/trac/boost/ticket/3697 { public: system_error( error_code ec ) @@ -61,13 +61,17 @@ { if ( m_what.empty() ) { +#ifndef BOOST_NO_EXCEPTIONS try +#endif { m_what = this->std::runtime_error::what(); if ( !m_what.empty() ) m_what += ": "; m_what += m_error_code.message(); } +#ifndef BOOST_NO_EXCEPTIONS catch (...) { return std::runtime_error::what(); } +#endif } return m_what.c_str(); } I get darwin.link.dll ../../../bin.v2/libs/chrono/build/darwin-4.7.0/debug/threading-multi/libboost_chrono.dylib Undefined symbols for architecture x86_64: "boost::throw_exception(std::exception const&)", referenced from: boost::chrono::chrono_detail::steady_full_ec(boost::system::error_code&) in chrono.o boost::chrono::steady_clock::now(boost::system::error_code&) in chrono.o boost::chrono::process_real_cpu_clock::now(boost::system::error_code&) in process_cpu_clocks.o boost::chrono::process_user_cpu_clock::now(boost::system::error_code&) in process_cpu_clocks.o boost::chrono::process_system_cpu_clock::now(boost::system::error_code&) in process_cpu_clocks.o boost::chrono::process_cpu_clock::now(boost::system::error_code&) in process_cpu_clocks.o Does this mean that the user needs to link with a static library? Is there a way to provide a dynamic library for Boost.Chrono if it is up to the user to define boost::throw_exception? Is there a possibility that Boost.System support this mode? BTW, an alternative to redefine the the what function is to pass the string at the construction of the exception. Best, Vicente

1 1