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

Re: [Boost-users] Remove edge
by Gianni Loiacono 18 Feb '09

18 Feb '09

Hi Mr. Michael Olea, thanks for the replay, for the example code and for the help :-) I have tested your code and it works perfectly. But I have a problem, my graph isn't static but dynamic. In fact, the arcs weights are assigned dynamicaly and I don't understand how to adapt your code to my code. Can you help me? Sorry for my poor english, Best regards. This is an example of my code that I want to adapt with remove_edge_if(): #include <iostream> #include <ctime> #include <utility> #include <boost/graph/graph_traits.hpp> #include <boost/graph/adjacency_list.hpp> #include <boost/graph/graph_utility.hpp> using namespace std; using namespace boost; HINSTANCE hInst; namespace boost { enum vertex_idNodo_t { vertex_idNodo = 1111 }; enum edge_arcoProprieta_t { edge_arcoProprieta = 2222 }; BOOST_INSTALL_PROPERTY(vertex, idNodo); BOOST_INSTALL_PROPERTY(edge, arcoProprieta); } typedef property<vertex_idNodo_t, int> NodoProperty; typedef property<edge_capacity_t, long> ArcoProperty; typedef adjacency_list<vecS, vecS, undirectedS, NodoProperty, ArcoProperty> Graph; Graph g; property_map<Graph, vertex_idNodo_t>::type idNodo = get(vertex_idNodo, g); property_map <Graph, edge_capacity_t >::type capacityCut = get(edge_capacity, g); graph_traits < Graph >::edge_iterator ei, ei_end; void deleteEdges() { for (tie(ei, ei_end) = edges(g); ei != ei_end; ++ei) { if (capacityCut[*ei] < 10) { remove_edge(source(*ei,g),target(*ei,g),g); //Error invalidates the edge iterators :( } } } void main() { for (tie(ei, ei_end) = edges(g); ei != ei_end; ++ei) { srand(static_cast<unsigned int>(time(NULL))); capacityCut[*ei] = rand() % 8 + 4; } } Passa a Yahoo! Mail. La webmail che ti offre GRATIS spazio illimitato, antispam e messenger integrato. http://it.mail.yahoo.com/

2 2

Clarification on Array, UNION serialization
by niranjan bangera 18 Feb '09

18 Feb '09

Hi all, I want to clarify myself in serialization of union and array, Please clarify my doubts. Boost Version:1.36.0 Compiler :/_TOOLS_/dist/gnu-gcc-/4.1.1-binutils-2.16.1/x86_64-pc-linux2.4/bin/gcc class A { union my_union { uint64_t array1[ 4]; int64_t array2[4]; double dbl [4]; uint32_t array3 [ 8 ]; int32_t array4 [ 8 ]; float flt [ 8]; uint16_t array5[ 16 ]; int16_t array6[ 16 ]; uint8_t array7 [ 32 ]; int8_t arrray8[ 32 ]; } my_union my_obj; friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int file_version) { ar & my_obj.array1; } }; Is it safe to do it, or any othe optimised method is there to do it? how to serialize a dynamicaly allocated array struc my_struct{ int x; } struct new_struct { { my_struct **my_var; ///my_struct is a structure defined before this structure. } In this case do i need to serialize each element int the array manually, Pointed By my_var Is there other some other way to do it... Thanks, Niranjan

2 1

[bccl] Compilation fails due to BOOST_CONCEPT_ASSERT
by Adam Badura 17 Feb '09

17 Feb '09

In my project I have two header files both introducing four BOOST_CONCEPT_ASSERT macros at namespace level. It happened that both those files used those macros at the same lines (same line numbers). The compilation of the one which is included as second fails (on MSVC 2008) with error: d:\studies\thesis\project\prover\formulas\terms\variable.h(83) : error C2365: 'boost_concept_check83' : redefinition; previous definition was 'enumerator' d:\studies\thesis\project\prover\formulas\terms\constant.h (83) : see declaration of 'boost_concept_check83' (repeated for each o those four uses with indexes 84, 85, 86 - clearly line numbers). It seems the line number is used to build some inner use name. Reasonable. But failable as well. Maybe file name macro could be used here however this could give string inappropriate for the identifier. I don't know how to solve this (a temporal workaround is obviously easy - just move the macros in on of files to a different place). I haven't found any info on that neither in the documentation nor in the groups. Adam Badura

1 0

unzip iterator?
by Neal Becker 17 Feb '09

17 Feb '09

I use an iterator adapator that converts a pair of scalar iterators into a complex iterator. But how to go the other direction? Take a complex iterator to a pair of scalar iterators. This is something like 'unzip'. Any thoughts?

2 1

Re: [Boost-users] boost geometry lib
by Barend Gehrels 17 Feb '09

17 Feb '09

Hi, Sorry for my late reply, was not on this list earlier. Thanks for your interest. As Bruno said we're currently busy with preview 4. There are substantial changes internally, but from user perspective the changes are not so large. Two things I'll mention: 1) > > http://geometrylibrary.geodan.nl/6__transformation__example_8cpp-exampl > > e.html > > shows the geometry library and uBLAS working together The transformation example will be changed (and much more simpler, from user perspective). The new version has internally transformation capacities and strategies can be used to finetune this (still using uBLAS) 2) if you plan to use the linestrings, note that the algorithms currently take line.begin() and line.end(). This is changed in the new version to just "line", using the boost range library. Preview 4 is expected this month or the beginning of February. Barend Gehrels Geodan, Amsterdam

2 3

Problem compiling a statemachine with "class" instead of "struct"
by Rudolf Leitgeb 17 Feb '09

17 Feb '09

Hi folks, we have run into an interesting problem when compiling one of our state machines using the boost state chart templates. If we instantiate a state like this: struct Active : public boost::statechart::simple_state< Active, PRX5StateMachine > { public: Active() : m_data("") {} ~Active() {} std::string & Data() { return m_data;} const std::string& Data() const { return m_data; } private: std::string m_data; }; and use it in our code, everything compiles fine. As soon as we change the declaration to class Active : public boost::statechart::simple_state< Active, PRX5StateMachine > we get massive compiler errors, which look like this: /usr/include/boost/statechart/simple_state.hpp:832: error: within this context /usr/include/boost/statechart/state_machine.hpp:243: error: ‘typedef struct boost::statechart::detail::rtti_policy boost::statechart::state_machine<PRX5StateMachine, Active, std::allocator<void>, boost::statechart::null_exception_translator>::rtti_policy_type’ is inaccessible Is it mandatory to inherit from simple_state via a struct and not from a class? Thanks for any inputs on this. On request I post two source directories which show the problem. Cheers, Rudi

3 8

[Interprocess] shared_ptr in shared memory and inheritance
by Gaetan Gaumer 17 Feb '09

17 Feb '09

Hello, we would like to store collections of differents objects derived from a same base class in a shared memory segment, and use polymorphism to manipulate them using shared_ptr<base>. But as we can't have virtual functions in shared memory we are exploring this solution : We designed 2 parallels hierarchies of classes - A hierarchy of empty objects in process memory with virtual functions (polymorphism) (ex : class A) - A hierarchy of data objects in shared memory without virtual functions (class A_data) A "process memory" object stores a weak_ptr to the "shared memory" data object. Use of data are encapsuled in accessors which take care of weak_ptr locking and concurrent access. I don't know if this is a good design (any opinions ?) but it seems to work well... until the destruction of "shared memory" objects. The wrong destructor is used (the base one) instead of the destructor of the derived class, probably because the destructor is not virtual... So I'm currently stuck and I'm looking for some advices. Here is a little code demonstrating the problem : #include <boost/interprocess/managed_shared_memory.hpp> #include <boost/interprocess/smart_ptr/shared_ptr.hpp> #include <iostream> using namespace boost::interprocess; using namespace std; class base { protected : int id; public: base(int i=0):id(i){}; ~base(){cout << "base dtor" << endl;} int getId() const {return id;} }; class derived : public base { protected: double data; public: derived(int i=0, double d=0): base(i), data(d){}; ~derived(){cout << "derived dtor" << endl;} double getData(){return data;} }; typedef managed_shared_ptr<base, fixed_managed_shared_memory>::type base_shared_ptr; int main (){ fixed_managed_shared_memory segment(open_or_create,"sharedPtrSharedMemoryTest", 2048); // does not compile but that's what I would like //base_shared_ptr baseSP = make_managed_shared_ptr(segment.construct<derived>(anonymous_instance)(2,3.0), segment); // turnaround base_shared_ptr baseSP = make_managed_shared_ptr((base*)segment.construct<derived>(anonymous_instance)(2,3.0), segment); cout << "baseSP->getId() =" << baseSP->getId(); cout <<", baseSP->getData() = " << static_pointer_cast<derived>(baseSP)->getData() << endl; baseSP.reset(); // don't know why but base dtor is called 3 times !!! } I can live with static_pointer_cast<> because they'll be hidden in accessors. But I really need that the right destructor (here ~derived()) will be called as it's cause some asserts in real case : include/boost/interprocess/detail/segment_manager_helper.hpp:176: static boost::interprocess::detail::block_header* boost::interprocess::detail::block_header::block_header_from_value(const void*, size_t, size_t): Assertion `hdr->m_value_alignment == algn' failed. The commented call to make_managed_shared_ptr makes gcc complains with (thanks gfilt for formatting) : sharedPtrSharedMemoryTest.cc:30: error: conversion from `boost::interprocess::shared_ptr< derived , boost::interprocess::allocator< void , boost::interprocess::segment_manager< char , boost::interprocess::rbtree_best_fit< boost::interprocess::mutex_family, void *, 0u >, boost::interprocess::iset_index > >, boost::interprocess::deleter< derived , boost::interprocess::segment_manager< char , boost::interprocess::rbtree_best_fit< boost::interprocess::mutex_family, void *, 0u >, boost::interprocess::iset_index > > >' to non-scalar type `boost::interprocess::shared_ptr< base , boost::interprocess::allocator< void , boost::interprocess::segment_manager< char , boost::interprocess::rbtree_best_fit< boost::interprocess::mutex_family, void *, 0u >, boost::interprocess::iset_index > >, boost::interprocess::deleter< base , boost::interprocess::segment_manager< char , boost::interprocess::rbtree_best_fit< boost::interprocess::mutex_family, void *, 0u >, boost::interprocess::iset_index > > >' requested Any help will be welcome. Gaetan

3 8

uBlas: compiler error (Borland 2007)
by Hendrik Fuß 17 Feb '09

17 Feb '09

Hi folks, I'm trying my luck with the Boost.uBlas library, but I'm running into problems. This little snippet: #include <boost/numeric/ublas/matrix.hpp> using namespace boost::numeric::ublas; int main(int argc, char* argv[]) { identity_matrix<double> i(3); matrix<double> m(i); return 0; } should compile alright, shouldn't it? The compiler says: [BCC32 Error] matrix.hpp(2671): E2247 'identity_matrix<double,std::allocator<double> >::one_' is not accessible The same happens if I use the assignment operator instead of copy construction. Can anyone help me? As you've probably seen from the error message, I'm using the CodeGear (alias Borland) 2007 compiler. I'm using a BoostPro 1.37.0 installation. I haven't tried boost 1.38, but the change log doesn't explicitly state any uBlas-related changes. thanks Hendrik

2 2

C1001 ice with date/time and Boost 1.38
by gast128 17 Feb '09

17 Feb '09

Dear all, I just downloaded Boost 1.38 and got ice with VStudio 2003 and sp1: c:\Work Sdk\boost_1_38_0\boost\date_time\gregorian\conversion.hpp(43) : fatal error C1001: INTERNAL COMPILER ERROR (compiler file 'msc1.cpp', line 2708) Please choose the Technical Support command on the Visual C++ Help menu, or open the Technical Support help file for more information It took me 3 hours to solve but it seems that if the /FR (Include All Browse Information) is on, this error is generated. Maybe this helps other people. Sometimes one get aggressive feelings about those boys from Redmond... wkr, me

3 3

[serialization]const member when using boost serialization
by Ray Zhang 17 Feb '09

17 Feb '09

Hi, I got a problem when i use boost:: serialization lib --- whether a const data member could be serialized? For example, class A { const int a; } how can i serialize 'a' by using boost? Best Regards. Ray Zhang. Tsinghua Univ. P.R.China

2 1

[boost-users][bind]
by Igor R 17 Feb '09

17 Feb '09

Hello, Could you please explain what causes the following warning and how to avoid it (boost 1.36, MSVC9.0SP1): typedef boost::weak_ptr<Task> TaskWeakPtr; typedef std::vector<TaskWeakPtr> TaskWeakPtrs; TaskWeakPtrs tasks_; TaskWeakPtrs::iterator toRemove = std::remove_if(tasks_.begin(), tasks_.end(), !boost::bind(&TaskWeakPtr::lock, _1)); 1>../boost\boost/bind.hpp(1643) : warning C4180: qualifier applied to function type has no meaning; ignored 1> ../boost\boost/bind.hpp(1677) : see reference to class template instantiation 'boost::_bi::add_cref<Pm,I>' being compiled 1> with 1> [ 1> Pm=boost::shared_ptr<Task> (__thiscall boost::weak_ptr<Task>::* )(void), 1> I=1 1> ] 1> .\TaskMaker.cpp(90) : see reference to class template instantiation 'boost::_bi::dm_result<Pm,A1>' being compiled 1> with 1> [ 1> Pm=boost::shared_ptr<Task> (__thiscall boost::weak_ptr<Task>::* )(void), 1> A1=boost::arg<1> 1> ] Thank you!

2 1

[boost.propertymap] queries
by Sean Farrow 17 Feb '09

17 Feb '09

Hi: I have two queries about boost.Property map: 1. I am in the middle of writing a class, and have an array with strings in. I would like to be able to have a series of properties in the class, and have these retrieved from different indexes of the array, so for example Comments [array index 1]. Is this possible with boost.Property Map 2 if the answer is yes to the first question, how would this be achieved? Any help appreciated Cheers Sean.

1 0

Re: [Boost-users] Remove edge
by Gianni Loiacono 16 Feb '09

16 Feb '09

Hi Michael, thanks for the answer. Can you write me an example for to use the remove_edge_if() function? I have read the docs on the site, but there isn't an example :( Thanks a lot Passa a Yahoo! Mail. La webmail che ti offre GRATIS spazio illimitato, antispam e messenger integrato. http://it.mail.yahoo.com/

2 1

[Serialize] Problem creating boost::archive::text_iarchive
by Javier Sanz 16 Feb '09

16 Feb '09

Hello, I'm getting problems creating a boost::archive::text_iarchive with boosts version 1.35.0 and Visual C++ Studio 2005 (Version 8). When I do so I get a std::bad_alloc exception. The code is as follows: void Serialize::deserialize(std::ifstream &modelFile, CoreModel &c) { try { boost::archive::text_iarchive ia(modelFile); // This line throws the exception (I've seen it with the debugger) ia >> c; } catch (std::exception& ex) { // I catch the bad_alloc here } } And this code si called this way: ... for (...) { modelFile.open(model_name.c_str(), std::ios::binary); if (modelFile.is_open() == false) { // ... We deal with the error but it does not get into here } CoreModel model; Serialize::deserialize(modelFile, model); .... } With the debugger I have seen that the exception is thrown into the method text_iarchive_impl<Archive>::load. The method basic_text_iprimitive::load is not setting the size of the string and when the string resize is done the value is huge (is garbage actually) and it throws the std::bad_alloc. Digging into the mailing list I have seen a thread that explains the same problem: http://lists.boost.org/boost-users/2006/03/18032.php But the solution proposed there does not fit me (As far as I can see). /Thanks Francisco Javier Sanz Olviera

3 2

Remove edge
by Gianni Loiacono 16 Feb '09

16 Feb '09

Hi all, I have defined my graph like this: typedef adjacency_list<vecS, vecS, undirectedS,NodoProperty, ArcoProperty> Graph2; And also: graph_traits < Graph2 >::edge_iterator ei, ei_end; Now, I want to delete all edges where the capacity is less than 10. Than I have write this code: for (tie(ei, ei_end) = edges(g); ei != ei_end; ++ei) { if (capacity2[*ei] < 10) { remove_edge(source(*ei,g),target(*ei,g),g); } } But the program after 2 cicles....it crashs :(....Why? Thanks a lot Passa a Yahoo! Mail. La webmail che ti offre GRATIS spazio illimitato, antispam e messenger integrato. http://it.mail.yahoo.com/

2 1

boost::asio handler based on received data
by Abdo Haji-Ali 16 Feb '09

16 Feb '09

Greetings, A server that I'm communicating with sends two types of messages identified by an ID. The first message type is sent automatically on specific events. The second type is sent in response to commands. I want to provide an interface that notifies the user when events occur (asynchronously) and at the same time enable performing commands. My first implementation did this (pseudocode): // In initialization socket.async_receive(buffer, event_handler); // In send_command socket.send(command_buffer); socket.receive_async(response_buffer, user_response_handler); event_handler() { user_event_handler(); socket.async_receive(buffer, event_handler); } The problem here is that 'event_handler' might (or will?) receive the response message. An additional problem is that these messages have no specific order. In other words, the following sequence might occur: - Send command (A) - Send command (B) - Receive event notification - Receive response (A) - Receive response (B) A possible solution might be the following: // In initialization socket.async_receive(buffer, msg_handler); // In send_command socket.send(command_buffer); handlers_queue.enqueue(user_command_handler); socket.receive_async(response_buffer, msg_handler); // Line(3) msg_handler() { if (event) user_event_handler(); else handlers_queue.dequeue() (); // Call first user handler socket.async_receive(buffer, msg_handler); } I want to know if boost::asio provides a better way to solve this problem. I'm particularly interested if writing a custom io service can help me in this situation. Thanks you, -- Abdo Haji-Ali Programmer In|Framez

1 0

[Lambda] Misplaced "const" causes runtime error?
by Paul Baker 16 Feb '09

16 Feb '09

Hi, The following program, as expected, outputs "1 1 1": #include <algorithm> #include <iostream> #include <vector> #include <boost/mem_fn.hpp> #include <boost/bind.hpp> #include <boost/lambda/bind.hpp> int main() { std::vector<std::pair<int, int> > pairs; pairs.push_back(std::make_pair(1, 1)); std::vector<int> result; std::transform(pairs.begin(), pairs.end(), std::back_inserter(result), boost::mem_fn(&std::pair<int, int>::first)); std::transform(pairs.begin(), pairs.end(), std::back_inserter(result), boost::bind(&std::pair<int, int>::first, _1)); std::transform(pairs.begin(), pairs.end(), std::back_inserter(result), boost::lambda::bind(&std::pair<int, int>::first, boost::lambda::_1)); std::cout << result[0] << ' ' << result[1] << ' ' << result[2]; } Changing "std::pair<int, int>" to "std::pair<const int, int>" on the line using boost::mem_fn or the line using boost::bind causes a compile-time error. However, performing the same change on the line using boost::lambda::bind leads to a successful compile - yet at run-time result[2] is set to a seemingly random value. This has been tested on Visual C++ 7.1 SP1, 8.0 SP1 and 9.0 SP1. Is this expected behaviour? I was expecting to see either a compile-time error or run-time success. Thanks

2 1

[boost-users][signals]
by Germán Diago 16 Feb '09

16 Feb '09

Hello. I'm trying to implement a similar system to signals/slots found in qt, with connections at runtime. Now I need to know the argument types for a given signal, like boost::function<>::first_arg_type and so on. Is this possible, and how? thank you very much.

3 2

[asio] read with transfer_at_least(300) return after 256 characters with no error
by tomasz jankowski 16 Feb '09

16 Feb '09

hello, how it is possible that size = boost::asio::read(socket_, boost::asio::buffer(serverReply_,300),boost::asio::transfer_at_least(300)); return with no error after reading 256? thanks, tom ---------------------------------------------------- Inżynierskie Targi Pracy 26.02.2009 http://klik.wp.pl/?adr=http%3A%2F%2Fwww.itp.BEST.krakow.pl&sid=643

1 0

Importing namespaced templates
by Esben Mose Hansen 16 Feb '09

16 Feb '09

Hi, from the C++0x standard, 7.3.3 (5) I gather that it is not possible to write a using statement to import a templated function into the global namespace. E.g. #include <boost/units/quantity.hpp> using template<typename X, typename Y> boost::units::quantity_cast<X,Y>; will be malformed. The only alternatives I have been able to come up with are 1) Use boost::units::quantity_cast<X>(Y) which is a bit longwinded 2) Write a new quantity_cast in the global namespace, and forward it to the boost::units one. Is there a better way? (Sorry that this is only borderline a boost question) -- kind regards, Esben

2 1

FindBoost.cmake
by Ruediger Berlich 16 Feb '09

16 Feb '09

Hi there, I'm using CMake as my build system -- not to build boost, but build my own project, which heavily depends on Boost. I habe in the past used the FindBoost.cmake macro to locate the most suitable Boost version. However, there does not seem to be a centrally maintained version of this file (I believe it originated from the KDE project ?), and I was wondering whether "anyone" in Boost was working on a replacement. After all, if CMake is considered as an alternative build system for Boost itself, allowing users to find the installed libraries would be the most logical step. Thanks and Best Regards, Ruediger

2 3

Seeking writers to develop quarterly digests of Boost newsgroups
by Brian Wood 16 Feb '09

16 Feb '09

I'm planning to start an on line magazine called C++ Quarterly. *http://tinyurl.com/df6hn2 or **http://preview.tinyurl.com/df6hn2 * If you are interested in writing a quarterly digest of Boost Users and/or Boost Developers, please let me know. Alternatively, if you're interested in writing a non-digest article, I'd be happy to hear from you. Just to clarify something from the above link. I want to set things up so that advertisers send money directly to authors. Advertisers would send ads to us but money to authors. Brian Wood Ebenezer Enterprises www.webEbenezer.net <http://www.webebenezer.net/>

1 0

[Regex]static compilation problems
by Moritz Beber 15 Feb '09

15 Feb '09

Dear users, to start using regular expressions I tried a very simple program only declaring one regular expression. #include <boost/regex.hpp> int main(int , char* []) { boost::regex re("\\d+"); return 0; } Linking and compiling this with the command: g++ test.cpp -o test -L ~/libs/boost/lib/ -l boost_regex-gcc43-mt works without problems, but running it then leads to the following: error while loading shared libraries: libboost_system-gcc43-mt-1_37.so.1.37.0: cannot open shared object file: No such file or directory But the directory clearly shows its existence: -rw-rw-r-- 1 engineer engineer 1587006 2008-11-21 07:53 /home/engineer/libs/boost/lib/libboost_regex-gcc43-mt-1_37.a lrwxrwxrwx 1 engineer engineer 38 2008-11-21 07:50 /home/engineer/libs/boost/lib/libboost_regex-gcc43-mt-1_37.so -> libboost_regex-gcc43-mt-1_37.so.1.37.0 -rwxrwxr-x 1 engineer engineer 722184 2008-11-21 07:50 /home/engineer/libs/boost/lib/libboost_regex-gcc43-mt-1_37.so.1.37.0 lrwxrwxrwx 1 engineer engineer 30 2008-11-21 07:53 /home/engineer/libs/boost/lib/libboost_regex-gcc43-mt.a -> libboost_regex-gcc43-mt-1_37.a lrwxrwxrwx 1 engineer engineer 38 2008-11-21 07:52 /home/engineer/libs/boost/lib/libboost_regex-gcc43-mt.so -> libboost_regex-gcc43-mt-1_37.so.1.37.0 Then I tried doing the same using static libraries: g++ -static test.cpp -o test -L ~/libs/boost/lib/ -l boost_regex-gcc43-mt but got an extremely long list of errors: /home/engineer/libs/boost/lib//libboost_regex-gcc43-mt.a(static_mutex.o): In function `boost::scoped_static_mutex_lock::unlock()': static_mutex.cpp:(.text+0x16): undefined reference to `pthread_mutex_unlock' /home/engineer/libs/boost/lib//libboost_regex-gcc43-mt.a(static_mutex.o): In function `boost::scoped_static_mutex_lock::lock()': static_mutex.cpp:(.text+0x46): undefined reference to `pthread_mutex_lock' /home/engineer/libs/boost/lib//libboost_regex-gcc43-mt.a(static_mutex.o): In function `boost::scoped_static_mutex_lock::~scoped_static_mutex_lock()': static_mutex.cpp:(.text+0x76): undefined reference to `pthread_mutex_unlock' /home/engineer/libs/boost/lib//libboost_regex-gcc43-mt.a(static_mutex.o): In function `boost::scoped_static_mutex_lock::~scoped_static_mutex_lock()': static_mutex.cpp:(.text+0xa6): undefined reference to `pthread_mutex_unlock' /home/engineer/libs/boost/lib//libboost_regex-gcc43-mt.a(static_mutex.o): In function `boost::scoped_static_mutex_lock::scoped_static_mutex_lock(boost::static_mutex&, bool)': static_mutex.cpp:(.text+0xdd): undefined reference to `pthread_mutex_lock' /home/engineer/libs/boost/lib//libboost_regex-gcc43-mt.a(static_mutex.o): In function `boost::scoped_static_mutex_lock::scoped_static_mutex_lock(boost::static_mutex&, bool)': static_mutex.cpp:(.text+0x10d): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale.o): In function `__gnu_cxx::__scoped_lock::~__scoped_lock()': (.text._ZN9__gnu_cxx13__scoped_lockD1Ev[__gnu_cxx::__scoped_lock::~__scoped_lock()]+0x18): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale.o): In function `std::locale::_Impl::_M_install_cache(std::locale::facet const*, unsigned int)': (.text._ZNSt6locale5_Impl16_M_install_cacheEPKNS_5facetEj+0x2d): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale.o): In function `std::locale::_Impl::_M_install_cache(std::locale::facet const*, unsigned int)': (.text._ZNSt6locale5_Impl16_M_install_cacheEPKNS_5facetEj+0x59): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale_init.o): In function `std::locale::locale()': (.text._ZNSt6localeC1Ev+0x2a): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale_init.o): In function `std::locale::locale()': (.text._ZNSt6localeC1Ev+0x52): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale_init.o): In function `std::locale::global(std::locale const&)': (.text._ZNSt6locale6globalERKS_+0x2a): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale_init.o): In function `std::locale::global(std::locale const&)': (.text._ZNSt6locale6globalERKS_+0x9a): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale_init.o): In function `std::locale::locale()': (.text._ZNSt6localeC2Ev+0x2a): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(locale_init.o): In function `std::locale::locale()': (.text._ZNSt6localeC2Ev+0x52): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(eh_alloc.o): In function `__cxa_free_exception': (.text.__cxa_free_exception+0x6b): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(eh_alloc.o): In function `__cxa_free_exception': (.text.__cxa_free_exception+0x86): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(eh_alloc.o): In function `__cxa_allocate_exception': (.text.__cxa_allocate_exception+0x6f): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libstdc++.a(eh_alloc.o): In function `__cxa_allocate_exception': (.text.__cxa_allocate_exception+0xd7): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libgcc_eh.a(unwind-dw2-fde-glibc.o): In function `__register_frame_info_bases': (.text+0x60): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libgcc_eh.a(unwind-dw2-fde-glibc.o): In function `__register_frame_info_bases': (.text+0x77): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libgcc_eh.a(unwind-dw2-fde-glibc.o): In function `__register_frame_info_table_bases': (.text+0x12a): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libgcc_eh.a(unwind-dw2-fde-glibc.o): In function `__register_frame_info_table_bases': (.text+0x141): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libgcc_eh.a(unwind-dw2-fde-glibc.o): In function `_Unwind_Find_FDE': (.text+0x18a8): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libgcc_eh.a(unwind-dw2-fde-glibc.o): In function `_Unwind_Find_FDE': (.text+0x18f8): undefined reference to `pthread_mutex_unlock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libgcc_eh.a(unwind-dw2-fde-glibc.o): In function `__deregister_frame_info_bases': (.text+0x1aaa): undefined reference to `pthread_mutex_lock' /usr/lib/gcc/i386-redhat-linux/4.3.0/libgcc_eh.a(unwind-dw2-fde-glibc.o): In function `__deregister_frame_info_bases': (.text+0x1b2f): undefined reference to `pthread_mutex_unlock' collect2: ld returned 1 exit status If anyone could point me to a solution of this problem, or give one here I'd very much appreciate it. Thank you. Best, Moritz

2 1

[Preprocessor] Questiona bout reentrant call and file iteration performance
by Joel Falcou 15 Feb '09

15 Feb '09

I've been playing with Boost PreProcessor for some times now and it really helped. Two corners still looks obscure for me as the doc doesn't provide clear-cut enough use-cases for my taste. 1/ Concerning the _D version of some macro, what's their use case ? I can see the example in the doc but I can't see why useing the normal version of the _D function should fail or behave badly 2/ Concerning file iteration vs BOOST_PP_REPEAT, what's the trade-off of one vs the other ? Is iterating faster at compile time than REPEATing ? Is there features of iteration that can't be done using REPEAT (or the reverse) or is the only bonus of ITERATE the fact that error actually send you in code and not on macro call site ? Thanks in advance -- ___________________________________________ Joel Falcou - Assistant Professor PARALL Team - LRI - Universite Paris Sud XI Tel : (+33)1 69 15 66 35

2 1

boost interprocess / shared memory question
by Naveen Santhanam 14 Feb '09

14 Feb '09

Hello, We have a windows application (a process) that requires a large data file (approx. 100 MB) to perform a task. We have a master that launches several instances of this application. We would like to share the large data file across the processes such that, there is just one instance of the data file in physical memory without any disc access. Hence each application will have a small footprint in memory during execution. We have tried using a managed shared memory object from the boost::interprocess library to achieve this. However, this doesn't help us as the memory footprint of the application goes up upon access of the data file. Is there a way to achieve the desired single footprint memory sharing? Does this work as expected with other OSes? Naveen

2 1