- Boost - lists.boost.org

[asio] Randomly having problems reading from the socket
by Torben Wiggerich 29 Jul '13

29 Jul '13

Hi all, I have some problems using the boost.asio.read for a TCP-socket. So far it works fine, but it randomly stops the transfer for at least 5 seconds. To resolve the error I used a simple synchronous server and client from the boost examples and modified it too send a larger amount of data. (307400 Bytes; this size comes from the program I'm trying to get run). On the client side I also changed the reading from the socket to read the complete data with boost.asio.read and transfer_all as argument. Both, server and client, measure the time for sending and reading to the socket and will give an error if it needs more than 1 second. Here is the sourcecode for the server: #include <iostream> #include <string> #include <boost/asio.hpp> #include <boost/date_time/posix_time/posix_time.hpp> using boost::asio::ip::tcp; using namespace boost::posix_time; int main() { try { boost::asio::io_service io_service; tcp::acceptor acceptor(io_service, tcp::endpoint(tcp::v4(), 8002)); tcp::socket socket(io_service); acceptor.accept(socket); unsigned char * array = new unsigned char[307400]; for (;;) { boost::system::error_code ignored_error; ptime startTime = microsec_clock::universal_time(); boost::asio::write(socket, boost::asio::buffer( array, 307400 ), boost::asio::transfer_all(), ignored_error); time_duration neededTime = microsec_clock::universal_time() - startTime; if ( neededTime.total_milliseconds() > 1000 ) { std::cout << "ERROR needed too much time: " << neededTime.total_milliseconds() << "ms" << std::endl; } } } catch (std::exception& e) { std::cerr << e.what() << std::endl; } return 0; } My client sourcecode: #include <iostream> #include <boost/array.hpp> #include <boost/asio.hpp> #include <boost/date_time/posix_time/posix_time.hpp> using boost::asio::ip::tcp; using namespace boost::posix_time; int main(int argc, char* argv[]) { try { if (argc != 2) { std::cerr << "Usage: client <host>" << std::endl; return 1; } boost::asio::io_service io_service; boost::asio::ip::tcp::endpoint endpoint( boost::asio::ip::address::from_string( argv[1] ), 8002 ); tcp::socket socket(io_service); boost::system::error_code error = boost::asio::error::host_not_found; socket.connect( endpoint, error ); if (error) throw boost::system::system_error(error); int counter = 0; unsigned char* array = new unsigned char[307400]; for (;;) { boost::system::error_code error; ptime startTime = microsec_clock::universal_time(); size_t len = boost::asio::read( socket, boost::asio::buffer( array, 307400 ), boost::asio::transfer_all(), error ); time_duration neededTime = microsec_clock::universal_time() - startTime; if ( error ) { printf( "body read error %s\n", error.message().c_str() ); } if ( counter % 5000 == 0 ) { std::cout << "passed: " << counter << "\r"; } ++counter; if ( neededTime.total_milliseconds() > 1000 ) { std::cout << "ERROR needed too much time: " << neededTime.total_milliseconds() << "ms" << std::endl; } } } catch (std::exception& e) { std::cerr << e.what() << std::endl; } return 0; } To reproduce the error just let it run for a while. It can take up to 30 minutes or even more... After a while it will produce the line: "ERROR needed too much time: 5007ms". Maybe with slightly different times, but all around 5 seconds. I tested it on different machines with Windows Vista and Seven. So am I'm doing sth. wrong? Is the reading from the socket correct? I'm using boost version 1.44. Kind regards Torben Wiggerich

6 12

New failing tests in circular buffer on some platforms
by Paul A. Bristow 29 Jul '13

29 Jul '13

I note that there are some new failing tests for circular buffer on some platforms http://www.boost.org/development/tests/trunk/developer/circular_buffer.html on MSVC noncopyable_movable_noexcept_t is causing trouble base_test.cpp C:\boost_regression\trunk\boost\boost/circular_buffer/base.hpp(694) : error C2558: struct 'noncopyable_movable_noexcept_t' : no copy constructor available or copy constructor is declared 'explicit' C:\boost_regression\trunk\boost\boost/circular_buffer/base.hpp(673) : while compiling class template member function 'noncopyable_movable_noexcept_t *boost::circular_buffer<T>::linearize(void)' with [ T=noncopyable_movable_test_t ] c:\boost_regression\trunk\boost\libs\circular_buffer\test\common.ipp(2160) : see reference to function template instantiation 'noncopyable_movable_noexcept_t *boost::circular_buffer<T>::linearize(void)' being compiled with [ T=noncopyable_movable_test_t ] c:\boost_regression\trunk\boost\libs\circular_buffer\test\common.ipp(2236) : see reference to class template instantiation 'boost::circular_buffer<T>' being compiled with [ T=noncopyable_movable_test_t ] C:\boost_regression\trunk\boost\boost/circular_buffer/base.hpp(697) : error C2558: struct 'noncopyable_movable_noexcept_t' : no copy constructor available or copy constructor is declared 'explicit' ON gcc and Clang there are warnings and failure Running 60 test cases... ../libs/circular_buffer/test/common.ipp(2303): error: in "check_containers_exception_specifications": check boost::has_nothrow_constructor<circular_buffer<int> >::value has failed Can any experts shed light on this? I do not believe that the underlying circular buffer code has changed recently. Paul --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow(a)hetp.u-net.com

2 2

[thread] Questions about broken_promise
by Klaim - Joël Lamotte 29 Jul '13

29 Jul '13

I have a few questions concerning broken promises (exception): 1. is the broken_promise exception a standard, or future standard behaviour? So far I assumed it wasn't because http://en.cppreference.com/w/cpp/thread/promise/~promise don't say anything about this and I only read about broken_promise exceptions on Boost.Thread documentation. I would like to know if I should expect the future standard library to have the same concept. I see no configuration macro to deactivate these exceptions. I don't want to actually, but if I'm correct that it's not standard or future standard, then it's surprising that I can't setup a standard-like setup. 2. I have a case that could be summarized like this: ///////////////////////////////////////////// Updated by thread A: struct X { /*... */ }; struct K { future<shared_ptr<X,U>> m_ft_x; shared_ptr<X,U> m_x; K( Foo& foo ) : m_ft_x( foo.create<X>() ) {} void update() // this is always called by a loop in the thread A { if( !m_x ) { if( m_ft_x.has_value() ) m_x = m_ft_x.get(); } else work(); } void work(); // uses m_x }; Updated by thread B: // special deleter which will just push a task in Foo work queue for deleting the object // in sync with thread B struct U { /* ... */ }; struct Foo { template< class T > future<shared_ptr<T,U>> create() // will push a creation task in a work queue updated in update() void update(); // called by a loop in the thread B, will execute all the work pushed in the work queue. }; ///////////////////////////////////////////// (this is an exageratedly simplified version of what I'm doing) It works nicely (even if I think it could be improved with future.then and other tools) but there is a minor problem appearing when an instance of K is constructed then almost immediately destructed, before the instance of Foo executed the creation of the X instance. In this case, visual studio (which is configured to detect all throw calls) detect a broken_promise exception thrown from the destructor of the promise which was in the code generated by Foo::create() and pushed in a work queue. I look at the destructor and I see: if(!future_->done && !future_->is_constructed) { future_->mark_exceptional_finish_internal(boost::copy_exception(broken_promise()), lock); } Here is the full copy_exception code: template <class T> inline exception_ptr copy_exception( T const & e ) { try { throw enable_current_exception(e); } catch( ... ) { return current_exception(); } } This code triggers actually two "first-chance" exceptions this is the first one) which are catch immediately intentionally. My question is: why is copy_exception necessary and why does it have to do it this way? It's a very minor problem as just re-setting my visual studio debugging settings to default makes the exception transparent -only visible in the console log. However each time I activate debug breaking on all c++ exception so that I can debug some potential issues, I can have a breakpoint which is actually "noise" as it is only used to implement a copy of exception. Is it possible to eliminate this noise? Or, as I'm guessing, is there a language or compiler limitation which forces this code to be written this way? Thanks for your time. Joel Lamotte

2 2

safe bool not safe enough ?
by John Maddock 28 Jul '13

28 Jul '13

Just discovered (oops!) that the so called safe bool ideom is not as safe as it appears, for example: class Testable { bool ok_; typedef void (Testable::*bool_type)() const; void this_type_does_not_support_comparisons() const {} public: explicit Testable(bool b=true):ok_(b) {} operator bool_type() const { return ok_==true ? &Testable::this_type_does_not_support_comparisons : 0; } }; Testable t(true); bool b = t; // Compiles OK - Ooops! Which is causing multiprecision to misbehave when used in conjunction with uBlas: https://svn.boost.org/trac/boost/ticket/8292#comment:2 I guess making the conversion explicit in C++11 mode would help (I still need to test that), but is there a C++03 solution? Thanks, John.

4 9

radix sort
by Sourav 28 Jul '13

28 Jul '13

Hi All, I wanted to know if someone is/was working on Radix Sort or not. I did check the SoC listings and did not see radix sort in the list. If no one is working, may someone help me provide some pointers/details as to what the expectation is with regards to Radix sort for boost ? Regards, -Sourav

7 27

boost::python help?
by David M. Cotter 26 Jul '13

26 Jul '13

i'd like my app to be "available" to python while it's running. for example, say my app is "FooBar.app". when my FooBar.app is running, now there is a python interface available to python, and the user can write python scripts to make use of it with their scripts, they can control my running application when FooBar.app is NOT running, perhaps making use of any of the python functions of "FooBar.app" would either return an error, or possibly launch "FooBar.app". can boost::python help with this? i've never worked with extending or embedding python, so any help would be super great

4 5

Patch bonanza for VS2013 Preview support
by Lars Viklund 25 Jul '13

25 Jul '13

Hi list. What's the best way to spend your vacation if not trying to get Boost to build on bleeding-edge compilers. I set out to see how well Boost 1.54.0 behaved on the newly released go-live-ready Visual Studio 2013 Preview. Below is the list of bugs filed against Boost and Microsoft. Enjoy. With these patches in place, I can get a complete build working with all libraries normally available on a fresh machine (means that I didn't test Python, MPI, ICU, compressed streams). I haven't been able to find out how to run the tests in any human-readable form, but this should not cause any regressions as it's either increasing workaround bounds or adding headers that the standard mandates already. Boost.Config VS2013 Preview version bump https://svn.boost.org/trac/boost/ticket/8753 The Visual Studio 2013 Preview toolchain has _MSC_VER 1800 and _MSC_FULL_VER 180020617. See the attached patch for the minimum changes that are needed to get anywhere at all. Boost.Build needs support for the VS2013 Preview toolset (18.00.20617.1) https://svn.boost.org/trac/boost/ticket/8754 Visual Studio 2013 Preview is out and claims to be 12.0. I copy-pasted everything related to 11.0 to refer to 12.0 and it Works On My Machine, see attached patch. Boost.Signals VS2013 Preview version bump https://svn.boost.org/trac/boost/ticket/8755 The BOOST_WORKAROUND macros at boost/signals/detail/named_slot_map.hpp:130 and libs/signals/src/named_slot_map.cpp:27 needs to be increased to <= 1800 in order to encompass the Visual Studio 2013 Preview compiler. Boost.MPL VS2013 Preview version bump https://svn.boost.org/trac/boost/ticket/8756 The BOOST_WORKAROUND macros at boost/mpl/assert.hpp lines 37 and 247 need to be changed to also consider BOOST_MSVC, == 1800, patch attached. Boost.Serialization lacks algorithm header include for std::min https://svn.boost.org/trac/boost/ticket/8757 The <algorithm> header providing std::min is not included in boost/archive/iterators/transform_width.hpp, this breaks on Visual Studio 2013 Preview due to library changes. Boost.Asio lacks algorithm header include for std::min https://svn.boost.org/trac/boost/ticket/8758 The <algorithm> header providing std::min is not included in boost/asio/detail/impl/win_iocp_io_service.hpp, this breaks on Visual Studio 2013 Preview due to library changes. std::bind broken with boost/tuple/tuple.hpp or anything else providing a ADL tie https://connect.microsoft.com/VisualStudio/feedback/details/792163/std-bind… Not our fault, I filed a bug on Connect and am hoping that they'll notice it before RTM. 'std::bind' finds our 'tie' if a type from a boost namespace is used in the 'std::function' argument type list. -- Lars Viklund | zao(a)acc.umu.se

13 35

[preprocessor] ARRAY_REPLACE bug?
by Mostafa 25 Jul '13

25 Jul '13

On Windows 7/Cygwin, using boost 1.53, preprocessing the following snippet: pp_array.cc ----------- #include <boost/preprocessor.hpp> //#define BOOST_PP_TUPLE_REM_0() // nothing #define ARRAY (3, (a, x, c)) BOOST_PP_ARRAY_REPLACE(ARRAY, 1, b) via: g++ -E -P -I /... pp_array.cc gives the following incorrect result: (3, (BOOST_PP_TUPLE_REM_0 () a, b , c)) Uncommenting the pound define in the snippet gives the correct result for the preprocessor. Is this a bug?

1 0

[dynamic_bitset] Compiler intrinsics to accelerate. whos agree?
by Jairo Andres Velasco Romero 25 Jul '13

25 Jul '13

I noted detail for dynamic_bitset<> is not using the far old ASM instruction BSF which could accelerate find_first() and find_next() methodsI checked asm generated with current table based implementation.On x64 and x86 BSF and BSR are available long time ago.Population count can be implemented with LZCNT for performance oriented ways to build.Core i7 with AVX2 enable new wide block bit manipulation but it needs to code with specific GNU g++ intrinsics and MSVC intrinsics. Someone agree? Jairo.

2 1

semi lock-free Signals/Slots
by Andrew Ho 24 Jul '13

24 Jul '13

Hi, I've been doing some testing with implementing a semi lock-free Signals/Slots library. Basically: 1. emits are lock-free thanks to atomic shared_ptr's/other atomic operations. 2. Only a single writer is allowed at a time, uses a mutex for locking. The primary goal is to improve low and no slot-count signal emit performance. In tests, I have been able to get substantial performance boosts (~2 to 5 times) on low slot-count signals, approaching the same performance as signals2 in low-contention (basically, single thread testing). I haven't tested high emit contention performance, though I suspect this library will excel quite well here because there is no locking on emit. The test implementation is still incomplete, and notably uses many C++11 features, though I don't think any of features I'm using can't be ported one way or another for C++03 support. Even if the library isn't used in full, some of the ideas might be implemented into the existing Signals2 library. link: https://github.com/helloworld922/signals3

2 2