- Boost-users - lists.boost.org

Using BOOST_*_EQUAL on bit-fields?
by Chris Cleeland 07 Mar '13

07 Mar '13

I'm using the Boost Unit Test Framework and have a struct with bitfields declared in it. When I compile a unit test with code like this StructWithBitfields mystruct; StructWithBitfields otherstruct; BOOST_CHECK_EQUAL(mystruct.bitfield, otherstruct.bitfield); I get a compiler errors error: non-const reference cannot bind to bit-field 'bitfield' referring to both arguments to the macro. Is there a preferred way for comparing bitfield values via unit test macros? -- Chris Cleeland

1 0

Boost.Function compile error
by gast128 07 Mar '13

07 Mar '13

Dear all, i got a compilation error with vstudio 2010 and Boost.Function: 1) pass functor in constructor 2) pass functor as assignment 3) use Boost.Ref 4) temporary function void foo() { std::identity<int> bla; boost::function<int (int)> fc(bla); boost::function<int (int)> fc2 = bla; boost::function<int (int)> fc(boost::ref(bla)); boost::function<int (int)>(bla); //<-- } first 3 are ok, but the last one vstudio sees this as redefintion of bla of type 'boost::function<int (int)>. What (and why) do I wrong or should use an explicit cast in the last case?

3 8

Re: [Boost-users] asio: async ops in multithreading environment
by Marat Abrarov 06 Mar '13

06 Mar '13

> 1) One thread issues scatter-gather boost::asio::async_write() for > boost::array<boost::asio::const_buffer, 2> on socket A. > 2) Another thread closes socket A after that. > 3) Meantime (simultaneously with 2), first buffer write is completed and > one of the threads running io_service::run() starts to execute second part > of async_write from 1). > > The problem is: second write which is issued internally by boost cannot be > protected by the same mutex which is used for protected access to the > socket in my code. At first, RTFM: 1. http://www.boost.org/doc/libs/1_53_0/doc/html/boost_asio/reference/Handler.h tml ~~~~~~~~~~~~~~ asio_handler_invoke(f, &h); ~~~~~~~~~~~~~~ 2. http://www.boost.org/doc/libs/1_53_0/doc/html/boost_asio/overview/core/stran ds.html ~~~~~~~~~~~~~~ In the case of composed asynchronous operations, such as async_read() or async_read_until(), if a completion handler goes through a strand, then all intermediate handlers should also go through the same strand. This is needed to ensure thread safe access for any objects that are shared between the caller and the composed operation (in the case of async_read() it's the socket, which the caller can close() to cancel the operation). This is done by having hook functions for all intermediate handlers which forward the calls to the customisable hook associated with the final handler: ~~~~~~~~~~~~~~ Than google some more: 3. https://github.com/boostcon/2011_presentations/raw/master/mon/thinking_async hronously.pdf See template <class Handler> struct mutex_wrapper at page 95. 4. http://asio-samples.blogspot.ru/2011/06/design-journeys-with-asio.html Sorry, it's in Russian. Regards, Marat Abrarov.

1 0

icu libraries renamed in icu-50-*
by koala01 06 Mar '13

06 Mar '13

Hello, I just have compiled a personnal build of mingw-w64 (gcc 4.8) and different extern libraries, as icu-50.1.2 (the last release of that library) on windows vista When trying to compile boost (1-53-0), it appears that icu is not found, due to following name changes in icu: * icudata library is now named icudt * icui18n library is now named icuin (i dont know if those change are recent or not :S) We can, of course, provide a copy of those libraries with the correct / needed names, but shouldn't it better to make a conditionnal choice on the icu version to know if we have to use one or other name, to ensure that icu would be correctly detected ? I expect that such test should simply take place in a jamfile or in another, am i wrong?

3 2

1.53, Bjam fails when adding arguments [Beginner]
by wespaugh 06 Mar '13

06 Mar '13

I'm trying to build 1.53 for use with an android platform, using some configuration code provided kindly provided by one of the community. (Android shouldn't be relevant to the problem, but it seemed noteworthy). My user-config.jam is: import os ; if [ os.name ] = CYGWIN || [ os.name ] = NT { androidPlatform = windows ; } androidNDKRoot = /cygdrive/c/Android/android-ndk-r8d ; using gcc:android:$(androidNDKRoot)/toolchains/arm-linux-androideabi/prebuilt/$(androidPlatform)/bin/arm-linux-androideabi-g++ <compileflags>--sysroot=$(androidNDKRoot)/platforms/android-8/arch-arm/ <compileflags>-I$(androidNDKRoot)/sources/cxx-stl/gnu-libstdc++/4.7/include <compileflags>-I$(androidNDKRoot)/sources/cxx-stl/gnu-libstdc++/4.7/libs/armeabi/include <archiver>$(androidNDKRoot)/toolchains/arm-linux-androideabi-4.7/prebuilt/$(androidPlatform)/bin/arm-linux-androideabi/ar <ranlib>$(androidNDKRoot)/toolchains/arm-linux-androideabi-4.7/prebuilt/$(androidPlatform)/bin/arm-linux-androideabi-ranlib ; And I am trying to build with the following command: b2 release --with-serialization --with-thread --with-date_time --with-system --with-filesystem --layout=system threading=multi link=static runtime-link=static toolset=gcc-android target-os=linux threadapi=pthread stage When I try to run, however, I get .../user-config.jam:109: in modules.load *** argument error * rule using ( toolset-module : * ) * called with: ( gcc:android:/cygdrive/c/Android/android-ndk-r8d/toolchains/arm-linux-androideabi/prebuilt/windows/bin/arm-linux-androideabi <compileflags>--sysroot=/cygdrive/c...{Rest of my compiler flags go here} * extra argument <compileflags>--sysroot=/cygdrive/c/Android/android-ndk-r8d/...{Rest of my compiler flags go here} /{...}/build/project.jam:886: see definition of rule 'using' being called /{...}/build-system.jam:257: in load-config /{...}/build-system.jam:399: in load-configuration-files /{...}/build-system.jam:555: in load /{...}/kernel/modules.jam:284: in import /{...}/kernel/bootstrap.jam:142: in boost-build /{...}/boost-build.jam:8: in module scope >From what I can tell, it doesn't like something about how my compile flags are passed in. I found some documentation on boost.org suggesting that the proper way to define them is like this: compileflags="--sysroot=/cygdrive/c/Android/{...}/ -I$(androidNDKRoot)/toolchains/{...} etc" But that gave me the same error. There's a strong chance that this is in TFM somewhere I just need to R. If so, I apologize, because I cannot find the manual. I would greatly appreciate any assistance in finding where to learn what I need to know about these config files. I would even more greatly appreciate some assistance with this specific problem. -- View this message in context: http://boost.2283326.n4.nabble.com/1-53-Bjam-fails-when-adding-arguments-Be… Sent from the Boost - Users mailing list archive at Nabble.com.

2 3

boost::filesystem create directory fail
by albert 06 Mar '13

06 Mar '13

I use the code below to create directories, but it fails. What is the correct way to do it? Thanks in advance. try { string logdirname("./notexistdir/logdir/"); boost::filesystem::path logdir(logdirname.c_str()); boost::filesystem::file_status s = boost::filesystem::status(logdir); if( !boost::filesystem::is_directory(s) ) { cout<< "not a directory, mkdir" <<endl; if(boost::filesystem::create_directory(logdir)) { cout<<"directory create ok" <<endl; } else { cout<<"directory create fail" <<endl; } } else { cout<<"is a directory" <<endl; } } catch(exception& ex) { cout<< "exception caught[" << ex.what() <<"]" <<endl; }

2 1

asio: async ops in multithreading environment
by Max Dmitrichenko 06 Mar '13

06 Mar '13

Hi! I have strange crashes of my application that uses async ops in multiple threads. Before I describe my vision of the problem I point the following: 1) I'm aware that shared access to the same socket from multiple threads is unsafe. 2) All operations which my code issues on the same socket are made under mutex. As far as I understand, the following happens: 1) One thread issues scatter-gather boost::asio::async_write() for boost::array<boost::asio::const_buffer, 2> on socket A. 2) Another thread closes socket A after that. 3) Meantime (simultaneously with 2), first buffer write is completed and one of the threads running io_service::run() starts to execute second part of async_write from 1). In this place application segfaults: #0 0x00007fa3a3a2c4eb in boost::asio::detail::epoll_reactor::start_op (this=0x19c3410, op_type=1, descriptor=27, descriptor_data=@0x7f9dfc013260, op=0x7f9dfc2d2130, allow_speculative=true) at boost-1_51/boost/asio/detail/impl/epoll_reactor.ipp:219 #1 0x00007fa3a3a2bf63 in boost::asio::detail::reactive_socket_service_base::start_op (this=0x19c4338, impl=..., op_type=1, op=0x7f9dfc2d2130, is_non_blocking=true, noop=false) at boost-1_51/boost/asio/detail/impl/reactive_socket_service_base.ipp:211 #2 0x00007fa3a3af4ec5 in boost::asio::detail::reactive_socket_service_base::async_send (this=0x19c4338, impl=..., buffers=..., flags=0, handler=...) at boost-1_51/boost/asio/detail/reactive_socket_service_base.hpp:212 #3 0x00007fa3a3af90bf in boost::asio::stream_socket_service<boost::asio::ip::tcp>::async_send (this=0x19c4310, impl=..., buffers=..., flags=0, handler=...) at boost-1_51/boost/asio/stream_socket_service.hpp:298 #4 0x00007fa3a3b05992 in boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> >::async_write_some (this=0x7f9dfc013250, buffers=..., handler=...) at boost-1_51/boost/asio/basic_stream_socket.hpp:676 because descriptor_data equals null-pointer here. The problem is: second write which is issued internally by boost cannot be protected by the same mutex which is used for protected access to the socket in my code. So, the question is: how can I safely close the socket? -- Max

2 4

Problem with the parallel graph
by Yeni Lora 06 Mar '13

06 Mar '13

I am student of Computer Science in Universidad de Oriente from Cuba, and I have a problem with the parallel graph, when I am adding edge betwen two procesors and one of them change any property of my edge, the other processor not update the property . for example my code is : #include "mpi.h" #include <boost/config.hpp> #include <boost/graph/use_mpi.hpp> #include <boost/graph/distributed/mpi_process_group.hpp> #include <boost/graph/distributed/adjacency_list.hpp> #include <boost/property_map/parallel/distributed_property_map.hpp> #include <boost/serialization/list.hpp> #include <list> using namespace std; //===============================Vertex================================ struct Vertex{ int id; // Serialization support is required! template<typename Archiver> void serialize(Archiver& ar, const unsigned int ) { ar & id ; } Vertex(int id = 0): id(id){} }; //===============================Edge================================ struct Edge{ int type; Edge( int type = 0): type(type){ } // Serialization support is required! template<typename Archiver> void serialize(Archiver& ar,const unsigned int ) { ar & type; } }; using namespace boost; using boost::graph::distributed::mpi_process_group; typedef adjacency_list < listS, distributedS<mpi_process_group, listS>, undirectedS, Vertex, Edge > Graph; typedef typename Graph::edge_descriptor edge_descriptor; typedef typename Graph::vertex_descriptor vertex_descriptor; int main (int argc, char ** argv){ boost::mpi::environment env(argc,argv); mpi::communicator world; Graph g; vertex_descriptor vi,vj; property_map<Graph, int Edge::*>::type type1 = get(&Edge::type, g); if (g.processor()==0) vi = add_vertex(Vertex(1),g); if (g.processor()==1) vj = add_vertex(Vertex(2),g); broadcast(world, vj, 1); broadcast(world, vi, 0); if (g.processor()==0){ pair<edge_descriptor,bool> p = add_edge(vi, vj, Edge(1), g); std::cout<<"Process "<<g.processor()<<std::endl; std::cout<<"Process owner of edge "<<owner(p.first)<<std::endl; std::cout<<"Property type of edge :"<<get(type1,p.first)<<std::endl; } synchronize(g); if (g.processor()==1){ pair<edge_descriptor,bool> p = edge(vi, vj, g); if(p.second){ std::cout<<"\nProcess "<<g.processor()<<std::endl; std::cout<<"Process owner of edge "<<owner(p.first)<<std::endl; put(type1,p.first,3); std::cout<<"Change property type of edge :"<<get(type1,p.first)<<std::endl; } } synchronize(g); if (g.processor()==0){ pair<edge_descriptor,bool> p = edge(vi, vj, g); std::cout<<"\nProcess "<<g.processor()<<std::endl; std::cout<<"Property type of edge :"<<get(type1,p.first)<<std::endl; } return 0; } The output: Process 0 Process owner of edge 0 Property type of edge :1 Process 1 Process owner of edge 1 Change property type of edge :3 Process 0 Property type of edge :1 What can I do for resolve the problem?

2 1

boost.MSM: No member function 'process_any_event' declared
by RaRi 05 Mar '13

05 Mar '13

Hi Christophe, In case a class is derived from the msm::back::state_machine class and used with the favor_compile_time policy I get the following error: *No member function 'process_any_event' declared* Please consider the following code snippets to get more details: *// MasterStateMachine.hpp* struct Master_: mgp_state_machine_def<Master_> { // ... } *class MasterStateMachine : public msm::back::state_machine<Master_, msm::back::favor_compile_time>{ public: MasterStateMachine () { // some stufff }; } *// MasterStateMachine.cpp* #include "MasterStateMachine.hpp" BOOST_MSM_BACK_GENERATE_PROCESS_EVENT(MasterStateMachine) Why does the compiler complains although the class MasterStateMachine inherits the "process_any_event" method from the *msm::back::state_machine<Master_, msm::back::favor_compile_time>* back-end? The point of matter is the *BOOST_MSM_BACK_GENERATE_PROCESS_EVENT(MasterStateMachine)* macro. In case it is uncomment, everythin works fine, but unfortunately without the support of the favor_compile_time policy. Background: I need to derive the class *MasterStateMachine * from the the mentioned back-end as singleton semantic has to be added to the *MasterStateMachine * class. *Not using* the favor_compile_policy is not a option! I have already tried to overload the *process_any_event* method in the derived "MasterStateMachine" class to comply with the compilers needs: class MasterStateMachine : public msm::back::state_machine<Master_, msm::back::favor_compile_time>{ private: static MasterStateMachine *m_Instance; // Constructor to avoid instantiation from the outside MasterStateMachine() {}; // Constructor to avoid copy-constructing MasterStateMachine(const MasterStateMachine&) {} // Constructor to avoid assignation MasterStateMachine& operator = (const MasterStateMachine&) { return *this; } ~MasterStateMachine() {}; public: static MasterStateMachine* getInstance(); static void release(); msm::back::HandledEnum process_any_event(const boost::any & evt) { return this->process_any_event(evt); } };* Unfortunately the error remains! Does anybody ;) have a clue? Thanks in advanced! Regards, Rafael -- View this message in context: http://boost.2283326.n4.nabble.com/boost-MSM-No-member-function-process-any… Sent from the Boost - Users mailing list archive at Nabble.com.

2 2

[thread] Using lambda with unique_future
by Christian Henning 05 Mar '13

05 Mar '13

Hi all, I'm toying around with futures and I'm stumbling with the Visual Studio 2010 compiler. The following doesn't compile: #include <iostream> #include <boost/thread/future.hpp> int main() { boost::unique_future<int> the_answer = boost::async( [] () ->int { return 42; } ); std::cout << the_answer.get() << std::endl; return 0; } the error is: 1>c:\gil_contributions\sdl_test\async_test.cpp(9): error C2440: 'initializing' : cannot convert from 'boost::unique_future<R>' to 'boost::unique_future<R>' 1> with 1> [ 1> R=void 1> ] 1> and 1> [ 1> R=int 1> ] 1> No constructor could take the source type, or constructor overload resolution was ambiguous clang works just fine: http://liveworkspace.org/code/1gOYwa$2 Is this a compiler error or am I doing something wrong? Thanks, Christian

2 2