August 2015 - Boost-users - lists.boost.org

[interprocess] Determine size of managed_shared_memory
by Kai Benndorf 12 Oct '19

12 Oct '19

Hi all, I want to create a managed_shared_memory in exactly the size I needed. Therefore I do something like this: size = sizeof( common_managed_shared_memory ); size += sizeof( MyClassInSharedMem ); managed_shared_memory segment(create_only, "name", size ); MyClassInSharedMem allocates itself some memory by segment->allocate(). I added this also to the size, but get a boost::interprocess::bad_alloc exception on allocating this additional memory. Are there some magic bytes I forgot to add? Many thanks in advance Kai

3 2

[Range, Phoenix] What am I missing...
by Robert Jones 25 Apr '17

25 Apr '17

Hi Folks Can anyone explain the output of this program... #include <iostream> #include <boost/range/algorithm/copy.hpp> #include <boost/range/counting_range.hpp> #include <boost/range/adaptor/reversed.hpp> #include <boost/range/adaptor/transformed.hpp> #include <boost/phoenix/bind/bind_function.hpp> #include <boost/phoenix/core/argument.hpp> int f( int i ) { return i * 2; } int main( ) { using boost::adaptors::reversed; using boost::adaptors::transformed; using boost::counting_range; using boost::copy; using boost::phoenix::arg_names::_1; copy( counting_range( 1, 5 ) | reversed | transformed( f ), std::ostream_iterator<int>( std::cout, " " ) ); std::cout << std::endl; copy( counting_range( 1, 5 ) | reversed | transformed( bind( f, _1 ) ), std::ostream_iterator<int>( std::cout, " " ) ); std::cout << std::endl; } Which is... 8 6 4 2 1768406272 1768406272 1768406272 1768406272 I'm using Gcc 4.6.2 & Boost 1.53 Thx All.

3 4

[statechart] Passing parameters from event to State Local Storage
by Christophe Bourez 16 Nov '16

16 Nov '16

I need to maintain a parameter into the state local storage. This parameter is initially contained in the event that causes the transition to the target state. I give hereafter a small illustrative example. typename sc = boost::statechart; class EvTest : public sc::event<EvTest> { public: EvTest(std::size_t val) : value(val) {} std::size_t getValue() const { return this->value; } private: std::size_t value; }; class MyStateMachine; class StSource; class StTarget; class StSource : public sc::simple_state<StSource, MyStateMachine> { public: typedef sc::custom_reaction<EvTest> reactions; sc::result react(const EvTest &event) { // How to pass event.getValue() to the target state ? transit<StTarget>(); } }; class StTarget : public sc::simple_state<StTarget, MyStateMachine> { public: StTarget(std::size_t val) : val(value) {} private: std::size_t value; }; Thanks, Christophe

5 6

[Interprocess] Communication between x86 and x64 issue
by Toni Neubert 19 Feb '16

19 Feb '16

I want to communicate between different platforms using boost interprocess. I am using vc12 and boost 1.58 on windows 7. My code below is a very simple example, that should work. But it doesn't for communications between different platforms... If I create it in x64 and open in win32, the process stuck at a lock at function try_based_lock in boost/int/sync/detail/common_algorithms.hpp In the other way around: win32 create, x64 open: the process crashes at name_length in segment_manager_helper.hpp while trying to find the name in index (priv_generic_find in segment_manager). #include <iostream> #include <boost/interprocess/managed_shared_memory.hpp> int main() { namespace bip = boost::interprocess; // open in WIN32, create in x64 #ifdef _WIN32 bip::managed_shared_memory msm(bip::open_only, "TestIPC"); #elsif X64 bip::shared_memory_object::remove("TestIPC"); bip::managed_shared_memory msm(bip::create_only, "TestIPC", 4096); msm.construct<uint32_t>("Data")[1](10); #endif // Get Data and print it auto data = msm.find<uint32_t>("Data"); if (data.second == 1) { std::cout << *data.first << std::endl; } std::cin.ignore(); return 0; } Does anybody have experience in this? Thank you in advance!

3 2

Re: [Boost-users] Crashing Issue
by swati upadhyaya 04 Nov '15

04 Nov '15

hi I am using like this where RunningThreads_ is vector containing boost thread object pointer(handler).I am using interrupt() for the interruption of thread interruption_point() catching that interruption then joining.everytime its crashing in join() for me but for others who have the same configuration its working fine all of us has windos 8. For me also it was working last week but this week the same piece of code is crashing. :( RunningThreads_[i].handle->interrupt(); RunningThreads_[i].handle->join(); delete RunningThreads_[i].handle; On Thu, Feb 5, 2015 at 9:31 AM, swati upadhyaya <swatiupadhyaya(a)gmail.com> wrote: > RunningThreads_[i].handle->interrupt(); > LOG4CPLUS_INFO(root, "Thread Interrupted for " << > RunningThreads_[i].app_name << " successful"); > RunningThreads_[i].handle->join(); > LOG4CPLUS_INFO(root, "Waiting for thread to exit " << > RunningThreads_[i].app_name << " successful"); > delete RunningThreads_[i].handle; > LOG4CPLUS_INFO(root, "Thread handle for " << > RunningThreads_[i].app_name << " successfully deleted"); > //RunningThreads_.erase(i); > > On Wed, Feb 4, 2015 at 12:10 AM, Vicente J. Botet Escriba < > vicente.botet(a)wanadoo.fr> wrote: > >> Le 03/02/15 09:10, swati upadhyaya a écrit : >> >> In my code join() timed_joined() is crashing, I am using these API >> after interrupt(),Same piece of code is working for others but in my system >> its crashing, >> Is there some dependency problem I am using >> Boost version: 1.55.0 >> OS: Windows 8 >> IDE: Visual Studio 2010 >> >> Hi, >> >> it could help if you post the example and the backtrace. >> >> Which version/OS/compiler use others for which this works? >> >> Vicente >> >> _______________________________________________ >> Boost-users mailing list >> Boost-users(a)lists.boost.org >> http://lists.boost.org/mailman/listinfo.cgi/boost-users >> > >

3 2

[MPI] Fail in compiling MPI and related libraries
by Yuan Yao 16 Sep '15

16 Sep '15

3 3

Include line number in boost logs
by Florian Lindner 03 Sep '15

03 Sep '15

Hello, I try to implement that a log entry also logs its line number with boost::log. If any developer of it reads here, please also consider this a wishlist item... I think the best way to add an additional attribute for it core::get()->add_global_attribute( "Line", attributes::mutable_constant<size_t>(0) ); in my logger config that is called once at startup. #define logInfo(methodname, message) \ boost::log::core::get()->get_global_attributes()["Line"].set(__LINE__); \ BOOST_LOG_SEV(_log, boost::log::trivial::severity_level::trace) << message \ Is my idea, but boostlog.cpp:126:5: error: no member named 'set' in 'boost::log::v2_mt_posix::aux::attribute_set_reference_proxy' getting the attribute somehow seems to work, but how can I set the value of the attribute? Thanks! Florian

2 5

[Type Erasure] Ambiguous call with vs 2015 community
by Yuan Yao 02 Sep '15

02 Sep '15

Dear all, I try to use type erasure on the free function foo. However, VS 2015 failed to compile the following code, which is OK with GCC and CLang. ========================================================= #include <boost/type_erasure/any.hpp> #include <boost/type_erasure/free.hpp> #include <typeindex> std::type_index foo(int); BOOST_TYPE_ERASURE_FREE((HasGetField), foo, 1) using Type = boost::type_erasure::any< HasGetField<std::type_index(boost::type_erasure::_self const&)>, boost::type_erasure::_self const& >; std::type_index foo(int) { return typeid(int); } int main() { Type x{ 1 }; foo(x); return 0; } ========================================================= Source.cpp C:\Program Files\Microsoft Visual Studio 14.0\VC\INCLUDE\xlocale(341): warning C4530: C++ exception handler used, but unwind semantics are not enabled. Specify /EHsc Source.cpp(20): error C2668: 'boost::type_erasure::injectHasGetField<R (const boost::type_erasure::_self &),Base,boost::type_erasure::index_list<0>>::foo': ambiguous call to overloaded function with [ R=std::type_index, Base=boost::type_erasure::any_base<boost::type_erasure::any<HasGetField<std::type_index (const boost::type_erasure::_self &)>,boost::type_erasure::_self &&>> ] Source.cpp(7): note: could be 'std::type_index boost::type_erasure::injectHasGetField<R (const boost::type_erasure::_self &),Base,boost::type_erasure::index_list<0>>::foo(eval_if_c<0,boost::type_erasure::detail::maybe_const_this_param<const boost::type_erasure::_self &,Base>,boost::type_erasure::as_param<Base,const boost::type_erasure::_self &>>::type)' [found using argument-dependent lookup] with [ R=std::type_index, Base=boost::type_erasure::any_base<boost::type_erasure::any<HasGetField<std::type_index (const boost::type_erasure::_self &)>,boost::type_erasure::_self &&>> ] Source.cpp(7): note: or 'std::type_index boost::type_erasure::injectHasGetField<R (const boost::type_erasure::_self &),Base,boost::type_erasure::index_list<0>>::foo(eval_if_c<0,boost::type_erasure::detail::maybe_const_this_param<const boost::type_erasure::_self &,Base>,boost::type_erasure::as_param<Base,const boost::type_erasure::_self &>>::type)' [found using argument-dependent lookup] with [ R=std::type_index, Base=boost::type_erasure::any_base<boost::type_erasure::any<HasGetField<std::type_index (const boost::type_erasure::_self &)>,boost::type_erasure::_self &>> ] Source.cpp(7): note: or 'std::type_index boost::type_erasure::injectHasGetField<R (const boost::type_erasure::_self &),Base,boost::type_erasure::index_list<0>>::foo(eval_if_c<0,boost::type_erasure::detail::maybe_const_this_param<const boost::type_erasure::_self &,Base>,boost::type_erasure::as_param<Base,const boost::type_erasure::_self &>>::type)' [found using argument-dependent lookup] with [ R=std::type_index, Base=boost::type_erasure::any_base<boost::type_erasure::any<HasGetField<std::type_index (const boost::type_erasure::_self &)>,boost::type_erasure::_self>> ] Source.cpp(7): note: or 'std::type_index boost::type_erasure::injectHasGetField<R (const boost::type_erasure::_self &),Base,boost::type_erasure::index_list<0>>::foo(eval_if_c<0,boost::type_erasure::detail::maybe_const_this_param<const boost::type_erasure::_self &,Base>,boost::type_erasure::as_param<Base,const boost::type_erasure::_self &>>::type)' [found using argument-dependent lookup] with [ R=std::type_index, Base=boost::type_erasure::any_base<boost::type_erasure::any<HasGetField<std::type_index (const boost::type_erasure::_self &)>,const boost::type_erasure::_self &>> ] Source.cpp(14): note: or 'std::type_index foo(int)' Source.cpp(20): note: while trying to match the argument list '(Type)' ========================================================= Moreover, I wonder whether the declaration is a must? Thank you!

2 2

Re: [Boost-users] is it possible to group values with Boost's rtree?
by Andy Edwards 01 Sep '15

01 Sep '15

Noel wrote: > I had to do something similar packing 100k 3d objects into the smallest > number of groups, where each group contains only non-overlapping 3d > objects. I used the Boost Graph library with a smallest least ordering > graph coloring algorithm. Since this is a quasi-np hard problem, I know > the packing density is sub-optimal but still quite good for our > application. Perhaps that approach would work for you? Unfortunately > I?m not familiar with Boost rtree. > > ? Noel Belcourt Thanks for the tip Noel, but it sounds like I'll be able to accomplish it using Adam's suggestions (I don't need completely optimal packing density). Since I'll be using the rtree for spatial queries as well it's the best approach for me. Adam wrote: > It is possible though it isn't officially supported. Have a look at the > utilities in this directory: > > https://github.com/boostorg/geometry/tree/master/include/boost/geometry/ind… > > To traverse the rtree you must write a visitor similar to the > Boost.Variant visitor, defining operator() overloads taking nodes. There > is a boost::geometry::index::detail::rtree::utilities::view<> providing > read-only access to the rtree's internals, in particular to the > apply_visitor() method applying the visitor to the root node of the > rtree and depth() function returning the number of tree levels. To > access the node's elements container you may use > boost::geometry::index::detail::rtree::elements() function and to apply > the visitor to the child node you may call > boost::geometry::index::detail::rtree::apply_visitor() function. The > elements of an internal node are of type similar to std::pair<> where > the first member is a Box and the second member is a pointer to the > child node. The elements of a leaf node are of type value_type. This way > you may recursively traverse the rtree in depth-first order counting the > current level, create containers of values when some level is met and > then gather the values from leafs. See e.g. how the utilities in the > directory mentioned above are implemented. > > Regards, > Adam Thanks Adam! That sounds ideal. Glad I'll be able to use this! I wonder if a per-node predicate is planned for the public API in the future (rather than the value-only satisfies predicate)?

2 1

[boost] [afio] Formal review of Boost.AFIO
by Ahmed Charles 01 Sep '15

01 Sep '15

The formal review of the Boost.AFIO library starts, August 21st and ends on Monday August 31st. Boost.AFIO provides a portable API implementing synchronous and asynchronous race-free filesystem and scatter-gather file i/o. It requires a minimum of C++ 11. It has minimum dependencies on a Filesystem TS and a Networking TS implementation (e.g. Boost.Filesystem and Boost.ASIO). Backends are provided for the Windows NT kernel and POSIX. The utility of a portable library providing strong concurrent race guarantees about filesystem and file i/o can be seen in its tutorial where a transactional ACID key-value store is built from first principles. Boost.AFIO was brought to Boost as part of GSoC 2013 and Niall Douglas has continued to develop and improve the library since then, generating two internal implementation libraries Boost.APIBind and Boost.Monad which are expected to be separately brought for Boost review in 2016. The documentation can be found here: https://boostgsoc13.github.io/boost.afio/doc/html/afio.html The source code can be found here: https://github.com/BoostGSoC13/boost.afio/tree/boost-peer-review Online web compiler playpen can be found here: http://melpon.org/wandbox/permlink/DR8wCpu5Rl20GMdM? Please answer the following questions: 1. Should Boost.AFIO be accepted into Boost? Please state all conditions for acceptance explicity. 2. What is your evaluation of the design? 3. What is your evaluation of the implementation? 4. What is your evaluation of the documentation? 5. What is your evaluation of the potential usefulness of the library? 6. Did you try to use the library? With what compiler? Did you have any problems? 7. How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? 8. Are you knowledgeable about the problem domain?

9 17

[Interprocess] Problem with file_lock on Linux
by Jan Niklas Hasse 31 Aug '15

31 Aug '15

The following code #include <boost/interprocess/sync/file_lock.hpp> #include <iostream> #include <fstream> int main() { std::ofstream file("test.lock"); boost::interprocess::file_lock flock("test.lock"); std::cout << flock.try_lock(); std::cout << flock.try_lock() << std::endl; } outputs 10 on Windows and 11 on Linux. Shouldn't it also print 10 on Linux or am I missunderstanding file_lock? I've also tried locking from different threads, but no luck.

2 3

[SPAM] from: giridhar
by giridhar 30 Aug '15

30 Aug '15

Sup boost http://arheatingcooling.com/unlikely.php?rule=g94y8vfvqvgeu2 giridharms(a)gmail.com giridhar

1 0

[variant][mpl] silly mistake in make_variant_over usage?
by Nat Goodspeed 29 Aug '15

29 Aug '15

The following program fails to compile (148 lines of errors, will paste on request) with g++ (Ubuntu 4.9.2-0ubuntu1~14.04) 4.9.2 in -std=c++11 mode. This is using Boost 1.59.0. Command line: g++ -std=c++11 -I ~/boost_1_59_0 -o make_variant make_variant.cpp I assume I'm making a simple error. I would appreciate your pointing it out. Thank you! (Is there a simpler way to obtain a variant of the result types of passed functions? I was hoping I could just write: boost::variant< std::result_of<Fns>::type... >.) #include <boost/variant/variant.hpp> #include <boost/mpl/vector.hpp> #include <boost/mpl/transform.hpp> #include <type_traits> // std::result_of template < typename... Fns > typename boost::make_variant_over< boost::mpl::transform< boost::mpl::vector<Fns...>, std::result_of<boost::mpl::_1> > >::type multiplex(Fns && ... functions) { typedef typename boost::make_variant_over< boost::mpl::transform< boost::mpl::vector<Fns...>, std::result_of<boost::mpl::_1> > >::type return_t; // dummy for now return return_t(); } int main(int argc, char *argv[]) { boost::variant<int, char*> var = multiplex([argc](){ return argc; }, [argv](){ return argv[0]; }); return 0; }

2 2

is it possible to group values with Boost's rtree?
by Andy Edwards 27 Aug '15

27 Aug '15

Hi everyone, I need a way to take a collection of 3D objects and put them into disjoint minimally overlapping groups no bigger than some N. Can this be done with Boost's rtree or any other Boost libraries? I've done this by putting the objects in an R* tree and using the nodes at a certain level of the tree (for example, 2 levels up from the leaf level) to determine the groups. However, that was in my own Java R* tree implementation that provided read-only access to its internal nodes. I'd like to port this program to C++ using Boost's rtree, but it doesn't provide read-only access to its nodes, does it? Does anyone know if it would be possible to extend the rtree to provide this? Or would it be possible to write a custom predicate to get the groupings? Thanks! Andy

3 2

[filesystem] Relative Draft Proposal RFC
by Beman Dawes 27 Aug '15

27 Aug '15

There have been requests for a Filesystem library relative function for at least ten years. The requested functionality seems simple - given two paths with a common prefix, return the non-common suffix portion of one of the paths such that it is relative to the other path. In terms of the Filesystem library, path p("/a/b/c"); path base("/a/b"); path rel = relative(p, base); // the requested function cout << rel << endl; // outputs "c" assert(absolute(rel, base) == p); If that was all there was to it, the Filesystem library would have had a relative function years ago. Critical issues: Clashing requirements, symlinks, directory placeholders (dot, dot-dot), user-expectations, corner cases. A paper by Jamie Allsop, Additions to Filesystem supporting Relative Paths, is what broke my mental logjam. Much of [my proposal] is based directly on Jamie's analysis and proposal. The weakly_canonical function and aspects of the semantic specifications are my contributions. Mistakes, of course, are mine. The full proposal is at http://boostorg.github.io/filesystem/relative_proposal.html A preliminary implementation is available in the feature/relative2 branch of the Boost Filesystem Git repository. See github.com/boostorg/filesystem/tree/feature/relative2 The overall plan is to get feedback from boost developers, then merge to develop for boost 1.60, then propose to the C++ committee for addition to the Filesystem TS. Comments welcome. If you are one of the many people who have requested "relative" functionality, does this proposal meet your needs? If not, why? Thanks, --Beman

5 7

Avoid including path to BOOST into exe
by Pavel Seleznev 27 Aug '15

27 Aug '15

Hello, I have the same question as described here http://stackoverflow.com/questions/32141335/avoid-including-path-to-boost-i… I have built BOOST with the following command line: b2 --toolset=msvc-10.0 --stagedir=./stage/2010/win32 address-model=32 variant=release threading=multi link=static runtime-link=static then compiled a project with statically linked runtime and release configuration with NDEBUG, BOOST_DISABLE_ASSERTS defined. Compiler options: c++\advanced\use full paths (/FC) is off c++\general\debug information format (/ZI /Zi /Z7) is empty ...and still having the entries in my binary file: c:\work\boost/xpressive/basic_regex.hpp void __cdecl boost::xpressive::detail::throw_on_stack_error(bool) Regex stack space exhausted c:\work\boost/xpressive/detail/dynamic/parse_charset.hpp void __thiscall boost::xpressive::detail::char_overflow_handler::operator () How can I avoid this? Regards, Pavel

1 0

Using boost modified bessel functions with complex arguments
by Ross Kynch 27 Aug '15

27 Aug '15

Dear all, Does anyone know how to call boost's modified bessel function of the first kind using a complex argument? Specifically I want the positive half integer order with complex argument so I use: #include <boost/math/special_functions/bessel.hpp> std::complex<double> output = boost::math::cyl_bessel_i(double nhalf, std::complex<double> x); but this gives me a huge list of compiler errors which seems to suggest the std::complex<double> template isn't valid. I am able to call other boost math library functions so I don't believe my install is to blame here. Thanks for any help Ross

2 2

Asio io_service and Thread scheduler/executor
by Mikael Olenfalk 26 Aug '15

26 Aug '15

Hi, I am currently thinking about how to move our project at work from a set of specific threads (1 thread per module e.g.) and our own network library abstraction to Boost.Asio and Boost.Thread's scheduler/executors. Before I read about schedulers/executors I considered just using asio::io_service (with a pool of threads calling .run()) for quick-shot background offloading, but after looking at scheduler/executor I would prefer using it for background work. Does anybody have any experience how to combine boost::asio with boost::thread's scheduler/executors? In my preferred design I would only have one thread pool backing both asio work and background computational tasks (e.g. using boost::async(Executor&, F&&, ...) ) My initial gut feeling was to have a thread-pool calling boost::asio::io_service.run() and then wrap that in an executor. I guess what I am asking for if somebody has any previous experience to share with this or similar approaches. Kind regards, /M

2 1

Fw: Use of QMutex not as expected
by Don El Ron 26 Aug '15

26 Aug '15

3 2

Boost Social Event at CppCon
by Jon Kalb 25 Aug '15

25 Aug '15

Boosters, Because a lot of Boosters are going to be attending CppCon, we¹d like to arrange a social event for Boosters/C++Now regulars. http://cppcon.org/ If you are planning to attend CppCon or even If you are just going to be in or around Bellevue, Washington between September 20th and 25 and you¹d like to join us for dinner, please let us know so we can get a count. Please send mail to: social(a)boost.org Once we settle on a date, etc, we¹ll let everyone who responded know when and where, so please let us know, even if you are a ³maybe.² I hope to see you at CppCon. Jon

1 0

[Graph] Distributed edge insertion
by Sensei 25 Aug '15

25 Aug '15

Dear all, I am finding difficulties in understanding how distributed graphs work. My graph is simple, a string-valued property, with some edges, however, right now I have only one edge... if it works. Since each process will add its own nodes, is it possible to avoid communication when adding nodes, given that each node owns its own vertices, and no overlap is possible between processes? This would save some communication avoiding re-distribution of nodes. Can I add edges concurrently? Meaning, each process adds some edges. I've only seen examples with root adding edges, but that's limiting. My last question is maybe easy. I'm adding an edge that connects two nodes, owned by different processes. Doing this results in a crash. It seems I cannot access properties with operator[] when traversing processes. I'm posting the code below, along with the *crashing* output. I hope you can help getting a grasp on PBGL! Thanks! #include <iostream> #include <string> #include <vector> #include <cmath> #include <list> #include <functional> #include <boost/config.hpp> #include <boost/mpi.hpp> #include <boost/graph/use_mpi.hpp> #include <boost/graph/distributed/mpi_process_group.hpp> #include <boost/graph/distributed/adjacency_list.hpp> #include <boost/graph/iteration_macros.hpp> using namespace boost; using boost::graph::distributed::mpi_process_group; // Useless class class str { public: str(std::string s = std::string()) : s_(s) { // NOP } template<typename Archiver> void serialize(Archiver& ar, const unsigned int /*version*/) { ar & s_; } std::string s_; }; // Allow indexing and constructing namespace boost { namespace graph { // Index based on the str class template<> struct internal_vertex_name<str> { typedef multi_index::member<str, std::string, &str::s_> type; }; // Construct with str class template<> struct internal_vertex_constructor<str> { typedef vertex_from_name<str> type; }; } } // Handy typedef boost::adjacency_list<vecS, distributedS<mpi_process_group, vecS>, bidirectionalS, str> dgraph; typedef graph_traits<dgraph>::vertex_descriptor dvertex; // START ME UP int main(int argc, const char * argv[]) { boost::mpi::environment env; boost::mpi::communicator comm; std::cout << "RANK " << comm.rank() << std::endl; comm.barrier(); dgraph g; std::vector<dvertex> descriptors; // Each rank for(int i = 0; i < 10; i++) { str s(std::to_string(i + (comm.rank() + 1) * 100)); std::cout << "rank " << comm.rank() << " adding " << s.s_ << std::endl; descriptors.push_back( add_vertex(s, g) ); } BGL_FORALL_VERTICES(v, g, dgraph) { std::cout << "V @ " << comm.rank() << " " << g[v].s_ << " local " << v.local << " owner " << v.owner << std::endl; } std::cout << ">>> NUM VERTICES " << num_vertices(g) << " NUM EDGES " << num_edges(g) << std::endl; comm.barrier(); if(comm.rank() >= 0) { int from, dest; bool letsCrash = true; if (letsCrash) { from = 1; dest = 0; } else { from = 0; dest = 0; } std::cout << "rank " << comm.rank() << " adding edge owners " << descriptors[from].owner << " - " << descriptors[dest].owner << std::endl; add_edge(descriptors[from], descriptors[dest], g); } boost::synchronize(g); BGL_FORALL_EDGES(e, g, dgraph) { std::cout << "E @ " << comm.rank() << " " << g[boost::source(e, g)].s_ << " -> " << g[boost::target(e, g)].s_ << std::endl; } return 0; } RANK 0 RANK 1 rank 0 adding 100 rank 0 adding 101 rank 1 adding 200 rank 1 adding 201 rank 0 adding 102 rank 0 adding 103 rank 1 adding 202 rank 1 adding 203 rank 0 adding 104 rank 1 adding 204 rank 0 adding 105 rank 1 adding 205 rank 1 adding 206 rank 0 adding 106 rank 0 adding 107 rank 1 adding 207 rank 0 adding 108 rank 1 adding 208 rank 0 adding 109 rank 1 adding 209 V @ 0 100 local 0 owner 0 V @ 0 201 local 1 owner 0 V @ 0 102 local 2 owner 0 V @ 0 203 local 3 owner 0 V @ 0 104 local 4 owner 0 V @ 0 205 local 5 owner 0 V @ 1 200 local 0 owner 1 V @ 1 101 local 1 owner 1 V @ 1 202 local 2 owner 1 V @ 1 103 local 3 owner 1 V @ 1 204 local 4 owner 1 V @ 1 105 local 5 owner 1 V @ 1 206 local 6 owner 1 V @ 1 107 local 7 owner 1 V @ 1 208 local 8 owner 1 V @ 0 106 local 6 owner 0 V @ 0 207 local 7 owner 0 V @ 0 108 local 8 owner 0 V @ 0 209 local 9 owner 0 >>> NUM VERTICES 10 NUM EDGES 0 V @ 1 109 local 9 owner 1 >>> NUM VERTICES 10 NUM EDGES 0 rank 1 adding edge owners 0 - 1 rank 0 adding edge owners 1 - 0 Assertion failed: (v.owner == processor()), function operator[], file /usr/local/include/boost/graph/distributed/adjacency_list.hpp, line 1696. Assertion failed: (v.owner == processor()), function operator[], file /usr/local/include/boost/graph/distributed/adjacency_list.hpp, line 1696. E @ 0 201 -> E @ 1 101 -> [Senseis-MBP:40908] *** Process received signal *** [Senseis-MBP:40908] Signal: Abort trap: 6 (6)

1 1

boost::ptr_vector does not support const types
by Martin Gamwell Dawids 25 Aug '15

25 Aug '15

I would like to use a ptr_vector to hold a number of const values but it seems ptr_vector does not currently support the use of const value types. See this simple code example: #include "boost/ptr_container/ptr_vector.hpp" #include <vector> class X { const int i; public: X(int i) : i(i) {} int getI() const { return i; } }; int main() { boost::ptr_vector<X> bvx; boost::ptr_vector<const X> bvcx; std::vector<X*> svx; std::vector<const X*> svcx; bvx.push_back(new X(42)); bvcx.push_back(new X(42)); // does not compile svx.push_back(new X(42)); svcx.push_back(new X(42)); return 0; } I have tried this on MSVC 2005/8.0 and with g++ on Mac – i.e. Apple LLVM version 6.1.0 (clang-602.0.53) (based on LLVM 3.6.0svn). On both compilers, calling 'push_back()' on a 'boost::ptr_vector<const X>' does not compile and on g++/Mac it yields this error message: ../boost/boost/ptr_container/ptr_sequence_adapter.hpp:249:37: error: reference to type 'const value_type' (aka 'void *const') could not bind to an lvalue of type 'value_type' (aka 'const X *') this->base().push_back( x ); // strong, commit ^ ptr_vector_test.cpp:22:8: note: in instantiation of member function 'boost::ptr_sequence_adapter<const X, std::__1::vector<void *, std::__1::allocator<void *> >, boost::heap_clone_allocator>::push_back' requested here bvcx.push_back(x); ^ /Library/Developer/CommandLineTools/usr/bin/../include/c++/v1/vector:697:62: note: passing argument to parameter '__x' here _LIBCPP_INLINE_VISIBILITY void push_back(const_reference __x); ^ 1 error generated. So the problem occurs when the passed value of type 'const X*' needs to be pushed to the backing 'std::vector<void*>' – i.e. a non-const 'void*' element type. The same problem exists for push_front(). A simple fix for 'push_back()' (and similar for 'push_front()') is to modify the file 'ptr_sequence_adapter.hpp' adding a cast of the value before pushing is: --- a/include/boost/ptr_container/ptr_sequence_adapter.hpp +++ b/include/boost/ptr_container/ptr_sequence_adapter.hpp @@ -246,7 +246,7 @@ namespace ptr_container_detail this->enforce_null_policy( x, "Null pointer in 'push_back()'" ); auto_type ptr( x ); // notrow - this->base().push_back( x ); // strong, commit + this->base().push_back( (void*) x ); // strong, commit ptr.release(); // nothrow } I am using a C-style cast as neither 'const_cast' nor 'reinterpret_cast' can be used here. (I know the formatting of the comments should also be fixed, but that would make my diff here more difficult to read.) Is there another workaround for this or have I overlooked something? /Martin

2 3

[multi-index] multi-index type trait
by Elizabeta 22 Aug '15

22 Aug '15

Hi how to test if the first index of boost::multi-index is key based or sequence based at compile time. I need to write something like this : template<class T,class M> void f(T& object,typename boost::enable_if<first_index_is_sequence<T>>::type* dummy = 0) { M element; object.push_back(element); } template<class T,class M> void f(T& object,typename boost::enable_if<first_index_is_key<T>>::type* dummy = 0) { M element; object.insert(element); } -- View this message in context: http://boost.2283326.n4.nabble.com/multi-index-multi-index-type-trait-tp467… Sent from the Boost - Users mailing list archive at Nabble.com.

2 5

Missing thread symbols when compiling from source: join(), ~thread() and start_thread()
by Edward Kornkven 22 Aug '15

22 Aug '15

Hello, I am trying to build Boost for linking with an application. I do not have admin access and there is an existing older Boost version in the system area. My problem is that there are three thread symbols that do not get defined by my build process: undefined reference to `boost::thread::join()' undefined reference to `boost::thread::~thread()' undefined reference to `boost::thread::start_thread()' Google reveals that this problem has been around for a while but the usual remedy is to install some binary or other. I need to build from source but for the life of me I can't understand what I am doing wrong that these symbols are not being defined. I can search everything that is being built using "nm" and they just aren't there. I think I am using a pretty standard build process of first building Boost.Build, then the rest: ./bootstrap.sh --with-toolset=gcc ./b2 install --prefix=${PREFIX} ${PREFIX}/bin/b2 --prefix=${PREFIX} toolset=gcc install I have geared down to various forms of magic: stage and then install, --without-thread followed by --with-thread, but I must be missing something basic about these libraries which I do not have occasion to use routinely. The build scripts are a bit opaque but don't indicate any errors. I am a little surprised that this isn't in an FAQ somewhere but maybe not many of us build Boost from source. Thanks for any assistance. Ed

1 0

build boost with specific msvcrt.lib
by Abhay Mhatre 21 Aug '15

21 Aug '15

is it possible to provide specific msvcrt.lib while building boost. It uses default one and my build system uses the one specified in build system. this causes heap corruption at the end of program termination. thanks

5 8