- Boost-users - lists.preview.boost.org

Lexical_cast and VC ExpresEdition 2010
by Claude 23 Nov '10

23 Nov '10

Hi! I use the line: uint32_t& c; string str("10); [...] c = lexical_cast<uint32_t>(str); I obtaing link error with VC 2010 (I use the code::blocks IDE): error LNK2019: riferimento al simbolo esterno "void __cdecl boost::throw_exception(class std::exception const &)" (?throw_exception@boost@@YAXABVexception@std@@@Z) non risolto nella funzione "unsigned int __cdecl boost::detail::lexical_cast<unsigned int,class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> >,0,char>(class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > const &,char *,unsigned int)" (??$lexical_cast@IV?$basic_string@DU?$char_LIBCMT.lib(wincrt0.obj)||error LNK2019: riferimento al simbolo esterno _WinMain@16 non risolto nella funzione ___tmainCRTStartup| LockFile.exe||fatal error LNK1120: 2 esterni non risolti But the same code in GCC under Linux work well. Why? -- View this message in context: http://boost.2283326.n4.nabble.com/Lexical-cast-and-VC-ExpresEdition-2010-t… Sent from the Boost - Users mailing list archive at Nabble.com.

4 4

icc ULONG_MAX/UINT_MAX vs msvc ULONG_MAX/UINT_MAX
by Hai Nguyen 23 Nov '10

23 Nov '10

Hi, I'm running into a case where the Intel compiler is seeing mt19937 as: typedef random::mersenne_twister< unsigned long, 32, 624, 397, 31, 0x9908b0df, 11, 7, 0x9d2c5680, 15, 0xefc60000, 18, 3346425566U > mt19937; and MSVC is seeing it as: typedef random::mersenne_twister< unsigned int, 32, 624, 397, 31, 0x9908b0df, 11, 7, 0x9d2c5680, 15, 0xefc60000, 18, 3346425566U > mt19937; ...it's defined as: typedef random::mersenne_twister< uint32_t, 32, 624, 397, 31, 0x9908b0df, 11, 7, 0x9d2c5680, 15, 0xefc60000, 18, 3346425566U > mt19937; This is leading to a link time problem. I'm a bit lost as to what to do since the lib I'm depending on is compiled with the Intel compiler and my client program is using MSVC. >From my very brief look - it seems that uint32_t is controlled via ULONG_MAX and UINT_MAX - I was wondering if someone could shed some light on how the Intel compiler handles ULONG_MAX and UINT_MAX. I couldn't find its limits.h to figure out what it was doing. Also - is there a work around for cases like this? Thanks, Hai

1 0

BOOST_PROTO_EXTENDS_USING_ASSIGN
by Hossein Haeri 23 Nov '10

23 Nov '10

Dear all, When I compile the example right after http://beta.boost.org/doc/libs/1_45_0/doc/html/proto/users_guide.html#boost… I get the following error in GCC 4.5.0 (MinGW32, WinXPSP2): error: ISO C++ forbids declaration of 'BOOST_PROTO_EXTENDS_USING_ASSIGN' with no type error: expected ';' before 'typedef' I first thought that that's because I haven't included the right set of headers. Replaced them all with #include <boost/proto/proto.hpp> and the problem remains intact. Any suggestions? TIA, --Hossein

2 5

[Tuple] augment_tuple, patch, or whatever... maybe MPL?
by Hossein Haeri 23 Nov '10

23 Nov '10

Dear all, Is there any function in Boost.Tuple which works like this for n-tuples for any given n? template<typename T1, typename T2, typename T3> tuples::tuple<T1, T2, T3> augment_tuple(tuples::tuple<T1, T2> tup, T3 t3) { return tuples::make_tuple(tup.get<0>(), tup.get<1>(), t3); } (Obviously, here n == 2.) The purpose of this augment_tuple (or, patch, or whatever) is to take an n-tuple and a new element and patch the new element to the end of the n-tuple in order to get a (n + 1)-tuple. Repetitive code which can produce enough overloads for a large enough n is of course trivial. I'm wondering whether MPL has a utility which can automated this though? TIA, --Hossein

2 3

Re: [Boost-users] [Property-tree] Boost 1.45 with Intel Compiler + VS 2010 compile error
by Edward Rankin 23 Nov '10

23 Nov '10

Thanks for the reply, Eric. > On 11/22/2010 8:57 PM, Edward Rankin wrote: >> Hi, >> >> I am trying to compile boost 1.45 using the recently released Intel >> Composer XE 2011 for Windows. I am running into the same error as >> described in ticker 4092 >> (https://svn.boost.org/trac/boost/ticket/4092) I have changed the >> check for BOOST_MSVC to include BOOST_INTEL_WIN (line 19 of >> boost/property_tree/detail/ptree_implementation.hpp) and can compile >> successfully. > <snip> > > Sounds like good information to the attach to the bug report. I have added a comment to the ticket. Is that sufficient? Sorry, I am not too familiar with how things work around here. Thanks, Edward

2 1

[Property-tree] Boost 1.45 with Intel Compiler + VS 2010 compile error
by Edward Rankin 23 Nov '10

23 Nov '10

Hi, I am trying to compile boost 1.45 using the recently released Intel Composer XE 2011 for Windows. I am running into the same error as described in ticker 4092 (https://svn.boost.org/trac/boost/ticket/4092) I have changed the check for BOOST_MSVC to include BOOST_INTEL_WIN (line 19 of boost/property_tree/detail/ptree_implementation.hpp) and can compile successfully. Not sure if this is the proper fix, so I hope the author/maintainer can look into this and fix this. Thanks! Edward

4 3

Conditional thread join
by Chris 23 Nov '10

23 Nov '10

Hello. I have following small program (pseudocode): void Read() { do { // ... } while (...); } // Read() void Write() { do { // ... } while (...); } // Write() boost::thread thrd1(Read); boost::thread thrd2(Write); // I don't want the following thrd1.join(); thrd2.join(); No, I want to stop when either thread finishes something like a join_any(). Does that exist? greetings Chris } while (...); } // Write() boost::thread thrd1(Read); boost::thread thrd2(Write); // I don't want the following thrd1.join(); thrd2.join(); No, I want to stop when either thread finishes something like a conditional join. Does that exist? greetings Chris

2 2

[MPI] MPI_Issend() support
by Riccardo Murri 23 Nov '10

23 Nov '10

Hello, I read in the manual that there's no equivalent of MPI_Issend() (noblocking synchronous send) in Boost.MPI. Is there any architectural obstacle to implementing it, or is it just that no-one has asked for it yet? Best regards, Riccardo

2 2

Boost.MPI library binary
by Hicham Mouline 22 Nov '10

22 Nov '10

Hello if I find a library binary of Boost.MPI, does it mean it must have been built against 1 of the many MPI implementations? Or is there some sort of link-level compatibility ?

3 3

boost.mpi: some MPI implementation that mixes mpi process from both windows and linux
by Hicham Mouline 22 Nov '10

22 Nov '10

Hello, Does such an implementation exist? and can it be used with Boost.MPI? Failing that, I am looking at having the GUI of my application in windows and the mpi processes running on a set of linux boxes I have. The question is: How can the GUI (not an mpi process) control the mpi environment (the bunch of mpi processes that run on the linux boxes)? regards,

1 0