- Boost-users - lists.boost.org

[Intrusive] unordered_set and thread safety
by Alexandre Hamez 22 Nov '12

22 Nov '12

Hello, I'm currently using intrusive:: unordered_set in a single threaded application. But now, I need to switch to a multi threaded application. As such, the unordered_set must be protected. I don't want to block the whole table with a single mutex. Thus, it would be fine to have a lock per bucket. Furthermore, intrusive:: unordered_set provides a bucket_traits mechanism to enable some customization. So, my question: is it possible to use bucket_traits to have a lock per bucket? Thank you! -- Alexandre Hamez

3 3

Errors compiling with Boost
by Amos Anderson 21 Nov '12

21 Nov '12

Hello -- I'm having trouble compiling our code on redhat gcc 4.4.6 (full specs below). There are a lot of errors in the output, but paring down the list a bit, some examples are pasted below. This exact same code successfully compiles on several OSX compilers, as well as redhat gcc 4.1.2. I posted this to Boost.Build, and someone informed me I had posted to the wrong list. This person mentioned that: > OSX has gcc-4.2 or clang 3.1 for 10.8. > > The gcc project has done some cleanups of standard library code, eliminating > some implicit dependencies. So you might need to explicitly include the now > missing headers. > > Yours, > > Jürgen Does anybody have any suggestions for which headers I might be missing? I do not have direct access to the machine with these compile errors, so it's hard for me to just fiddle with my code. thanks for any advice! Amos. 1) gcc.compile.c++ pymodule/bin/gcc-4.4.6/release-core/threading-multi/SingleCalculator_BP.o In file included from pymodule/SingleCalculator_BP.cc:13: ../tools/boost/boost/range/algorithm/equal.hpp: In function ‘void core::wrap_vector(const std::string&) [with T = boost::shared_ptr<core::energy::SingleCalculator>, bool noProxy = true]’: pymodule/SingleCalculator_BP.cc:33: instantiated from here ../tools/boost/boost/range/algorithm/equal.hpp:160: error: ‘boost::range’ is not a function, ../tools/boost/boost/python/iterator.hpp:92: error: conflict with ‘template<class Accessor1, class Accessor2> boost::python::api::object boost::python::range(Accessor1, Accessor2)’ pymodule/wrap_vector.hh:68: error: in call to ‘range’ "g++" -ftemplate-depth-128 -O3 -finline-functions -Wno-inline -Wall -g -pthread -fPIC -Dcore_MPI -Dkiss_fft_scalar=float -Wno-unknown-pragmas -Wstrict-aliasing=2 -fno-strict-aliasing -DBOOST_PYTHON_MAX_ARITY=30 -DBOOST_SYSTEM_DYN_LINK=1 -I"../tools/boost" -I"../tools/python/include/python2.7" -I"/import/rh_software/openmpi-1.4.3/include" -c -o "pymodule/bin/gcc-4.4.6/release-core/threading-multi/SingleCalculator_BP.o" "pymodule/SingleCalculator_BP.cc" 2) <snip> instantiated from here ../tools/boost/boost/mpi/detail/mpi_datatype_oarchive.hpp:51: error: invalid use of incomplete type ‘struct boost::is_enum<float>’ ../tools/boost/boost/type_traits/is_enum.hpp:181: error: declaration of ‘struct boost::is_enum<float>’ <snip> 3) <snip> ../tools/boost/boost/mpi/detail/mpi_datatype_oarchive.hpp: In member function ‘void boost::mpi::detail::mpi_datatype_oarchive::save_override(const T&, int) [with T = long unsigned int]’: ../tools/boost/boost/mpi/detail/mpi_datatype_oarchive.hpp:51: error: template instantiation depth exceeds maximum of 128 (use -ftemplate-depth-NN to increase the maximum) instantiating ‘struct boost::is_enum<long unsigned int>’ <snip> 4) <snip> instantiated from here ../tools/boost/boost/archive/detail/check.hpp:78: error: ‘value’ is not a member of ‘boost::archive::detail::check_object_versioning() [with T = boost::serialization::nvp<float>]::typex <snip> g++ -v Using built-in specs. Target: x86_64-redhat-linux Configured with: ../configure --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-bugurl=http://bugzilla.redhat.com/bugzilla --enable-bootstrap --enable-shared --enable-threads=posix --enable-checking=release --with-system-zlib --enable-__cxa_atexit --disable-libunwind-exceptions --enable-gnu-unique-object --enable-languages=c,c++,objc,obj-c++,java,fortran,ada --enable-java-awt=gtk --disable-dssi --with-java-home=/usr/lib/jvm/java-1.5.0-gcj-1.5.0.0/jre --enable-libgcj-multifile --enable-java-maintainer-mode --with-ecj-jar=/usr/share/java/eclipse-ecj.jar --disable-libjava-multilib --with-ppl --with-cloog --with-tune=generic --with-arch_32=i686 --build=x86_64-redhat-linux Thread model: posix gcc version 4.4.6 20110731 (Red Hat 4.4.6-3) (GCC)

1 0

SVN access problem
by Simon West 20 Nov '12

20 Nov '12

Hi, I am being asked for a username and password when attempting to check out Boost anonymously using the following command: $ svn co http://svn.boost.org/svn/boost/trunk boost-trunk This command was copy-pasted directly from the instructions for anonymous read-only access described at https://svn.boost.org/trac/boost/wiki/BoostSubversion. This is holding up my progress in creating a patch - can anyone help? Thanks, Simon.

1 0

[enable_if] Compile Time Member Overloading
by Christian Henning 20 Nov '12

20 Nov '12

Hi there, can someone tell me why the following doesn't compile with VS2010 SP1? #include <boost/mpl/and.hpp> #include <boost/type_traits/is_same.hpp> #include <boost/utility/enable_if.hpp> using namespace boost; struct foo { void bla() { read_data_24<int, int>(); } template< typename T1 , typename T2 > void read_data_24( typename enable_if< mpl::and_< mpl::true_ , mpl::bool_< is_same< T1, T2 >::type::value > > >::type* /* ptr */ = 0 ) { int i = 9; } }; int main(int argc, char* argv[]) { return 0; } I get the error: 1>gil_test_vc10.cpp(25): error C2039: 'type' : is not a member of 'boost::enable_if<Cond>' 1> with 1> [ 1> Cond=boost::mpl::and_<boost::mpl::true_,boost::mpl::bool_<false>> 1> ] I tried a couple of hours to have make is happen but I don't get it. Thanks, Christian

3 6

[ASIO] "perf" shows overhead of function calls is very high
by Junchang Wang 19 Nov '12

19 Nov '12

Hi list, I'm new to Boost.asio and am solving scaling problems of a Openflow (TCP based) server. We are using one single io_service for packets receiving and transmitting. A thread pool calling io_service::run() processes HTTP requests. A multi-core server is used. I found we can't get benefits from increasing the size of thread pool when it's size is larger than 6. In other words the system doesn't scale any more on more than 6 cores. I checked the program carefully and can make sure there's no lock contention. I profile the program with Linux perf, and it turns out up to 36% of CPU time is wasted in function calls. 36.39% : 0000000000103300 <boost::detail::function::void_function_obj_invoker3<vigil::openflow::v1::construct_and_load<vigil::openflow::v1::ofp_packet_out, vigil:: : BOOST_FUNCTION_TEMPLATE_PARMS : > : struct BOOST_FUNCTION_VOID_FUNCTION_OBJ_INVOKER : { : static BOOST_FUNCTION_VOID_RETURN_TYPE : invoke(function_buffer& function_obj_ptr BOOST_FUNCTION_COMMA 0.02 : 103300: push %rbp 99.66 : 103301: mov %rsp,%rbp 0.00 : 103304: push %r12 0.04 : 103306: mov %rcx,%r12 I printed out the call stack (attached at the tail of this email). It seems for a single packet processing there are more than 30 function calls within Boost. Can that explain why the CPU time on function calls is so high? Are there some ways to avoid this kind of cost? Thanks a lot. --Junchang ================ #0 vigil::openflow::Openflow_datapath::send (this=<optimized out>, msg=0x7ffff567f920) at ../../../../src/coreapps/openflow/openflow-datapath.cc:199 #1 0x00007ffff4c50724 in vigil::openflow::Openflow_datapath::handle_handshake (this=0x7ffff007dd80, e=...) at ../../../../src/coreapps/openflow/openflow-datapath.cc:305 #2 0x00007ffff4c5086a in vigil::openflow::Openflow_datapath::handle_message (this=0x7ffff007dd80, msg=0x7ffff567f920) at ../../../../src/coreapps/openflow/openflow-datapath.cc:228 #3 0x00007ffff4c50a39 in vigil::openflow::Openflow_datapath::recv_cb (this=0x7ffff007dd80, bytes_transferred=<optimized out>) at ../../../../src/coreapps/openflow/openflow-datapath.cc:164 #4 0x00007ffff7bad130 in operator() (a0=<optimized out>, this=0x7ffff007d4b8) at /usr/include/boost/function/function_template.hpp:1013 #5 vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >::handle_recv ( this=0x7ffff007d480, ec=..., bytes_transferred=<optimized out>) at ../../../src/lib/connection.cc:185 #6 0x00007ffff7bb5fba in call<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > >, boost::system::error_code const, unsigned long const> (b2=@0x7ffff568fa10: 8, b1=..., u=..., this=0x7ffff568f9e0) at /usr/include/boost/bind/mem_fn_template.hpp:271 #7 operator()<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > > (u=..., this=0x7ffff568f9e0, a2=@0x7ffff568fa10: 8, a1=...) at /usr/include/boost/bind/mem_fn_template.hpp:286 #8 operator()<boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, const boost::system::error_code&, const long unsigned int&>, boost::_bi::list2<const boost::system::error_code&, const long unsigned int&> > (a=<synthetic pointer>, f=..., this=0x7ffff568f9f0) at /usr/include/boost/bind/bind.hpp:392 #9 operator()<boost::system::error_code, long unsigned int> (a2=@0x7ffff568fa10: 8, a1=..., this=0x7ffff568f9e0) at /usr/include/boost/bind/bind_template.hpp:102 #10 operator() (this=0x7ffff568f9e0) at /usr/include/boost/asio/detail/bind_handler.hpp:97 #11 asio_handler_invoke<boost::asio::detail::binder2<boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> >, boost::system::error_code, unsigned long> > ( function=<error reading variable: access outside bounds of object referenced via synthetic pointer>) at /usr/include/boost/asio/handler_invoke_hook.hpp:64 #12 boost_asio_handler_invoke_helpers::invoke<boost::asio::detail::binder2<boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> >, boost::system::error_code, unsigned long>, boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> > > (function=..., context=...) at /usr/include/boost/asio/detail/handler_invoke_helpers.hpp:39 #13 0x00007ffff7bb637b in asio_handler_invoke<boost::asio::detail::binder2<boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> >, boost::system::error_code, unsigned long>, boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> >, boost::system::error_code, unsigned long> (this_handler=0x7ffff568fbf0, function=...) at /usr/include/boost/asio/detail/bind_handler.hpp:132 #14 invoke<boost::asio::detail::binder2<boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> >, boost::system::error_code, unsigned long>, boost::asio::detail::binder2<boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> >, boost::system::error_code, unsigned long> > (context=..., function=...) at /usr/include/boost/asio/detail/handler_invoke_helpers.hpp:39 #15 boost::asio::detail::strand_service::dispatch<boost::asio::detail::binder2<boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> >, boost::system::error_code, unsigned long> > (this=0x7ffff0001560, impl=@0x7ffff568fbb0: 0x7ffff0047d80, handler=...) at /usr/include/boost/asio/detail/impl/strand_service.hpp:97 #16 0x00007ffff7bb66fa in dispatch<boost::asio::detail::binder2<boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> >, boost::system::error_code, unsigned long> > ( handler=<error reading variable: access outside bounds of object referenced via synthetic pointer>, this=0x7ffff568fba8) at /usr/include/boost/asio/strand.hpp:159 #17 operator()<boost::system::error_code, long unsigned int> (arg2=<synthetic pointer>, arg1=..., this=0x7ffff568fba8) at /usr/include/boost/asio/detail/wrapped_handler.hpp:65 #18 operator()<boost::system::error_code, long unsigned int> (arg2=<optimized out>, arg1=..., this=0x7ffff568fba0) at ../../../src/lib/connection.cc:64 #19 operator() (this=0x7ffff568fba0) at /usr/include/boost/asio/detail/bind_handler.hpp:97 #20 asio_handler_invoke<boost::asio::detail::binder2<vigil::custom_alloc_handler<boost::asio::detail::wrapped_handler<boost::asio::io_service::strand, boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> > > >, boost::system::error_code, unsigned long> > (function=<error reading variable: access outside bounds of object referenced via synthetic pointer>) at /usr/include/boost/asio/handler_invoke_hook.hpp:64 #21 invoke<boost::asio::detail::binder2<vigil::custom_alloc_handler<boost::asio::detail::wrapped_handler<boost::asio::io_service::strand, boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> > > >, boost::system::error_code, unsigned long>, vigil::custom_alloc_handler<boost::asio::detail::wrapped_handler<boost::asio::io_service::strand, boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp> > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> > > > > (function=..., context=...) at /usr/include/boost/asio/detail/handler_invoke_helpers.hpp:39 #22 boost::asio::detail::reactive_socket_recv_op<boost::asio::mutable_buffers_1, vigil::custom_alloc_handler<boost::asio::detail::wrapped_handler<boost::asio::io_service::strand, boost::_bi::bind_t<void, boost::_mfi::mf2<void, vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > >, boost::system::error_code const&, unsigned long const&>, boost::_bi::list3<boost::_bi::value<boost::shared_ptr<vigil::Stream_connection<boost::asio::basic_stream_socket<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp> > > > >, boost::arg<1> (*)(), boost::arg<2> (*)()> > > > >::do_complete (owner=0x6d6b30, base=<optimized out>) at /usr/include/boost/asio/detail/reactive_socket_recv_op.hpp:106 #23 0x0000000000487343 in complete (owner=..., this=0x7ffff007d510) at /usr/include/boost/asio/detail/task_io_service_operation.hpp:35 #24 boost::asio::detail::task_io_service::do_one (this=0x6d6b30, lock=..., this_idle_thread=0x7ffff568fdd0) at /usr/include/boost/asio/detail/impl/task_io_service.ipp:278 #25 0x00000000004874b0 in boost::asio::detail::task_io_service::run (this=0x6d6b30, ec=...) at /usr/include/boost/asio/detail/impl/task_io_service.ipp:130 #26 0x00000000004875c9 in boost::asio::io_service::run (this=0x6d68f8) at /usr/include/boost/asio/impl/io_service.ipp:57 #27 0x00007ffff7971ce9 in thread_proxy () from /usr/lib/libboost_thread.so.1.46.1 #28 0x00007ffff5f2ee9a in start_thread () from /lib/x86_64-linux-gnu/libpthread.so.0 #29 0x00007ffff6237cbd in clone () from /lib/x86_64-linux-gnu/libc.so.6 #30 0x0000000000000000 in ?? ()

2 2

code review please? [boost::asio]
by Littlefield, Tyler 18 Nov '12

18 Nov '12

hello all: I was curious if some boost::asio guru would mind checking this out. I just posted my code in, I'd like to know if there's anything that could be done better. Thanks, //client.h #pragma once #include <queue> #include <boost/asio.hpp> #include <boost/bind.hpp> #include "ClientBuffer.h" class Client { protected: boost::asio::ip::tcp::socket _socket; boost::asio::io_service &_service; std::queue<ClientBuffer*> _buffers; bool _writing; void StartWrite(); void HandleWrite(const boost::system::error_code& error); public: Client(boost::asio::io_service &service); ~Client(); virtual boost::asio::ip::tcp::socket& GetSocket(); virtual void Start() = 0; void Write(const char* data); void Write(void* data, unsigned int size); void Write(const std::string &data); }; //client.cpp #include <boost/asio.hpp> #include <boost/bind.hpp> #include "client.h" #include "ClientBuffer.h" Client::Client(boost::asio::io_service &service): _service(service), _socket(service), _writing(false) { } Client::~Client() { } void Client::StartWrite() { //we don't write if we're already writing. We just content ourselves with having pushed the message to the queue. if (_writing) { return; } //we're writing now. _writing = true; ClientBuffer* buffer = NULL; buffer = _buffers.front(); boost::asio::async_write(_socket, boost::asio::buffer(buffer->GetBuffer(), buffer->GetSize()), boost::bind(&Client::HandleWrite, this, boost::asio::placeholders::error)); } void Client::HandleWrite(const boost::system::error_code& error) { if (_buffers.empty()) { _writing = false; return; } ClientBuffer* buffer = _buffers.front(); _buffers.pop(); if (buffer) delete buffer; if (_buffers.empty()) _writing = false; else StartWrite(); } boost::asio::ip::tcp::socket& Client::GetSocket() { return _socket; } void Client::Write(const char* data) { ClientBuffer* buffer = new ClientBuffer(data); _buffers.push(buffer); StartWrite(); } void Client::Write(void* data, unsigned int size) { ClientBuffer* buffer = new ClientBuffer(data, size); _buffers.push(buffer); StartWrite(); } void Client::Write(const std::string &data) { ClientBuffer* buffer = new ClientBuffer(data); _buffers.push(buffer); StartWrite(); } //TelnetClient.h /* *this client is responsible for handling the telnet protocol. */ #pragma once #include <boost/asio.hpp> #include <boost/bind.hpp> #include "client.h" #include "ClientBuffer.h" class TelnetClient:public Client { int _state; boost::asio::streambuf _buffer; public: virtual void Start(); void Read(); void HandleRead(const boost::system::error_code& error); void HandleCommand(const std::string &line); }; //TelnetClient.cpp #include <string> #include <istream> #include <boost/asio.hpp> #include <boost/bind.hpp> #include "client.h" #include "TelnetClient.h" void TelnetClient::Start() { Write("Welcome to SapphireMud.\n"); Write("Use load <player> <password> to load a character.\n"); Write("Use create <player> <password> to create a new player.\n"); _state = 1; Read(); } void TelnetClient::Read() { boost::asio::async_read_until(_socket, _buffer, '\n', boost::bind(&TelnetClient::HandleRead, this, boost::asio::placeholders::error)); } void TelnetClient::HandleRead(const boost::system::error_code& error) { char buffer[2048]; if (error) { //do something here return; } std::istream input(&_buffer); input.getline(buffer, 2048); HandleCommand(buffer); Read(); } void TelnetClient::HandleCommand(const std::string &line) { Write("I don't understand that!\n"); } -- Take care, Ty http://tds-solutions.net The aspen project: a barebones light-weight mud engine: http://code.google.com/p/aspenmud He that will not reason is a bigot; he that cannot reason is a fool; he that dares not reason is a slave.

2 2

Re: [Boost-users] What is the best idiom for using call_once ?
by John M. Dlugosz 18 Nov '12

18 Nov '12

Date: Sat, 17 Nov 2012 13:51:48 +0100 From: "Vicente J. Botet Escriba" <vicente.botet(a)wanadoo.fr> > Hi, your code suffers from the problem you try to avoid "But we all know this is not > thread safe, at least for the first call" at least in c++98 as you are using static > variables. Actually, my post described the circumstances under which it does in fact work: when VarType has a built-in type that is initialized before program execution, or on C++11 with a constexpr constructor. > here they are some pointers that while not responding exactly to your > question them could help you. > > http://www.justsoftwaresolutions.co.uk/threading/multithreading-in-c++0x-pa… Interesting, the post gets it wrong. A unique_ptr has a constructor (in C++03) so still has the problem. Or, maybe the article is just not clear on how it is to be used: if you must initialize an instance of lazy_init (to the null pointer internally) _before_ multiple threads ask it to be "initialized", that just pushes the problem back to the earlier step. I can see the usefulness though, if you just want to put off the expensive "real" initialization until needed, but can arrange the pre-initialization to be done for-sure earlier. > > http://stackoverflow.com/questions/12302057/c11-safe-double-checked-locking… This points out that under C++11 rules, static locals are thread safe for the first-time initialization. That explains why there is no fancy construct for doing so; it's just there! Nice. Or it will be, some day. I looked at the code generated by Microsoft VS2010 and it does not give this feature, alas. It does tell me though that a compatibility construct like I'm looking for ought to expand into a plain static local for compilers that do work that way; just let the compiler do it. All the more reason to abstract the issue for current uses! —John

1 0

What is the best idiom for using call_once ?
by John M. Dlugosz 17 Nov '12

17 Nov '12

Consider the issue of a static variable inside a function: the idea is to initialize on first use (if any). But we all know this is not thread safe, at least for the first call. Since the C++11 standard library includes a feature just like boost::call_once, that would seem to be the proper way to address this moving forward. And, I don't see anything in the standard that would be more powerful and direct for this purpose, which leaves me to wonder if there is still more that should have been done. But, my question is, what is the "best" or slickest way to use it (for this purpose)? If Lambda functions are available (which is not the case on every platform my code needs to support) _and_ the type of variable in question has some kind of "dead" state available, it's not too bad to write something like this: VarType obtain_singleton() { static boost::once_flag f=BOOST_ONCE_INIT; static VarType retval; boost::call_once (f, [] { retval=the_real_initialization(); }); return retval; } //...later // static VarType val= the_real_initialization; // !!not thread safe on first call!! static VarType val= obtain_singleton(); // do this instead (and write it in another function for conciseness) However, it's more awkward if I can't declare VarType retval to be pre-initialized to some "dead" value, or if it has any constructor at all and would itself re-introduce the thread safety issue! In short, it only works correctly (without the availability of the constexpr feature and types having constructors of such a nature) for primitive types. I've worked around it using a pointer and having the real_init function itself contain a static variable. That makes me think that there should be a simpler way to do it. Or, once the "best way" has been determined, why can't it be done once and succinctly reused? Anyone have one to share?

3 2

Re: [Boost-users] Boost 1.52 .zip bootstrap file -- missing?
by Adam J. Cooper 17 Nov '12

17 Nov '12

I downloaded "boost_1_52_0.zip" from here on 11/16 http://www.boost.org/users/history/version_1_52_0.html I used the native unzip built into Windows 8. There were two folders in the root directory, docs and source files. There was no bootstrap file or any other tools files. - Adam ------------------------------ Message: 2 Date: Sat, 17 Nov 2012 10:13:55 +0000 From: Daniel James <dnljms(a)gmail.com> To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] Boost 1.52 .zip bootstrap file -- missing? Message-ID: <CAHOE3yct3am=xamtRZrVGjK+_BjyXESgu9ofMm_F2vyb8435DA(a)mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 On 16 November 2012 23:39, Adam J. Cooper <adam(a)adamcooper.com> wrote: > I download the .zip version of Boost 1.52 > > However, I do not see the bootstrap file when I unzip the contents. That's odd. I just downloaded and there's a 'bootstrap.sh' and 'bootstrap.bat' in the 'boost_1_52_0' directory. Where did you download it from, and what did you use to unzip it? _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users End of Boost-users Digest, Vol 3272, Issue 2 ********************************************

2 1

Boost 1.52 .zip bootstrap file -- missing?
by Adam J. Cooper 17 Nov '12

17 Nov '12

I download the .zip version of Boost 1.52 However, I do not see the bootstrap file when I unzip the contents. Please advise. Thank you, Adam

2 1