August 2009 - Boost-users - lists.boost.org

Feedback on units use
by Noah Roberts 27 Aug '09

27 Aug '09

Is there a better way to do the following? The purpose is to provide a conversion from a pressure to a head given the density of the fluid the head should be in. Head is "hydrostatic pressure" and is a measure of the length a fluid can be pushed up a column at pressure X. The value is a function of the fluid and of the pressure acted upon that fluid. Object is to do this in any system. Part of the equation is the constant "g", which is acceleration due to gravity. We're modeling objects on earth so the value of g is 9.80665 m/s^2. The conversion equation is very basic: h = p / (d * g) where h is the headloss in length, p is pressure, d is mass density, and g is the gravitational constant. #include <boost/units/quantity.hpp> #include <boost/units/physical_dimensions.hpp> template < typename T, typename System > struct acceleration_due_to_gravity; template < typename T, typename System > boost::units::quantity< boost::units::unit<boost::units::length_dimension, System>, T > pressure_to_head( boost::units::quantity< boost::units::unit<boost::units::pressure_dimension, System>, T > press , boost::units::quantity< boost::units::unit<boost::units::mass_density_dimension, System>, T> density) { return press / (density * acceleration_due_to_gravity<T,System>::value()); } #include <boost/units/systems/si.hpp> template < typename T > struct acceleration_due_to_gravity<T, boost::units::si::system> { typedef boost::units::si::system System; static boost::units::quantity< boost::units::unit< boost::units::acceleration_dimension, System>, T> value() { return boost::units::quantity<boost::units::unit< boost::units::acceleration_dimension, System>, T>::from_value(9.80665); } };

2 1

Re: [Boost-users] Compiling python extensions examples fail
by Dan Sanduleac 27 Aug '09

27 Aug '09

I had the same issue, trying to compile latest svn trunk of boost. When I bootstrap.sh-ed, it created project-config.jam which had my python configuration in it (I provided --with-python-root=). Now, I also had ~/user-config.jam which had the exact same line ("using python : 2.6 : /home/dan/stackless ;") and bjam wasn't sure which one of those to choose, even though they were identical. The solution for me was to remove the superfluous user-config.jam from my home directory. Sean Snyders wrote: > > Hi, > > I've compiled Boost 1.39 locally (to a folder > '/local1/installs/boost_1_39_0') such that I can try out building python > extensions without disrupting the current system setup. I compiled it with > gcc 4.3.3 64bit and with python 2.5.2 64bit. > > My system details are: > Ubuntu 7.04 > Linux 2.6.25.18-20081021 x86_64 GNU/Linux > > I've setup my jam configs as follows: > file: ~/boost-build.jam > containing: boost-build /local1/installs/boost_1_39_0/tools/build/v2 ; > > file: /local1/installs/boost_1_39_0/tools/build/v2/site_config.jam > containing: > using gcc : : /vol/apps/gcc-4.3.3_64/bin/g++ ; > using python : : /vol/apps/python-2.5.2_64/bin/python ; > > I'm trying to compile the python tutorial > 'boost_1_39_0/libs/python/example/tutorial' using bjam and get the > following errors when running it: > ------- > error: No best alternative for /python_for_extensions > next alternative: required properties: <python>2.5 <target-os>linux > matched > next alternative: required properties: <python>2.5 <target-os>linux > matched > /local1/installs/boost_1_39_0/tools/build/v2/build/property.jam:613: in > find-replace from module object(property-map)@1 > error: Ambiguous key <target-type>PYTHON_EXTENSION > .... > .. > . > ------ > > Here is the full output that is given with the '--debug-configuration' > option: > ---------------------------------------------- > ssnyders@machine:/local1/installs/boost_1_39_0/libs/python/example/tutorial > {650} > /local1/ssnyders/boost_1_39_0/bin/bjam --debug-configuration > notice: found boost-build.jam at > /local1/installs/boost_1_39_0/libs/python/example/boost-build.jam > notice: loading Boost.Build from > /local1/installs/boost_1_39_0/tools/build/v2 > notice: Searching /etc /usr/home/ssnyders > /local1/installs/boost_1_39_0/libs/python/example/../../../tools/build/v2 > /usr/share/boost-build /local1/installs/boost_1_39_0/tools/build/v2/kernel > /local1/installs/boost_1_39_0/tools/build/v2/util > /local1/installs/boost_1_39_0/tools/build/v2/build > /local1/installs/boost_1_39_0/tools/build/v2/tools > /local1/installs/boost_1_39_0/tools/build/v2/contrib > /local1/installs/boost_1_39_0/tools/build/v2/. for site-config > configuration file site-config.jam . > notice: Loading site-config configuration file site-config.jam from > /local1/installs/boost_1_39_0/tools/build/v2/site-config.jam . > notice: will use '/vol/apps/gcc-4.3.3_64/bin/g++' for gcc, condition > <toolset>gcc-4.3.3 > notice: using gcc libraries :: <toolset>gcc-4.3.3 :: > /vol/apps/gcc-4.3.3_64/bin /vol/apps/gcc-4.3.3_64/lib > /vol/apps/gcc-4.3.3_64/lib32 /vol/apps/gcc-4.3.3_64/lib64 > notice: using gcc archiver :: <toolset>gcc-4.3.3 :: ar > warning: toolset gcc initialization: can not find tool windres > warning: initialized from > /local1/installs/boost_1_39_0/tools/build/v2/build/project.jam:874 > notice: using rc compiler :: <toolset>gcc-4.3.3 :: as > notice: [python-cfg] Configuring python... > notice: [python-cfg] user-specified cmd-or-prefix: > "/vol/apps/python-2.5.2_64/bin/python" > notice: [python-cfg] Checking interpreter command > "/vol/apps/python-2.5.2_64/bin/python"... > notice: [python-cfg] running command > '"/vol/apps/python-2.5.2_64/bin/python" -c "from sys import *; print > 'version=%d.%d\nplatform=%s\nprefix=%s\nexec_prefix=%s\nexecutable=%s' % > (version_info[0],version_info[1],platform,prefix,exec_prefix,executable)" > 2>&1' > notice: [python-cfg] ...requested configuration matched! > notice: [python-cfg] Details of this Python configuration: > notice: [python-cfg] interpreter command: > "/vol/apps/python-2.5.2_64/bin/python" > notice: [python-cfg] include path: > "/vol/apps/python-2.5.2_64/include/python2.5" > notice: [python-cfg] library path: > "/vol/apps/python-2.5.2_64/lib/python2.5/config" > "/vol/apps/python-2.5.2_64/lib" > notice: Searching /usr/home/ssnyders > /local1/installs/boost_1_39_0/libs/python/example/../../../tools/build/v2 > /usr/share/boost-build /local1/installs/boost_1_39_0/tools/build/v2/kernel > /local1/installs/boost_1_39_0/tools/build/v2/util > /local1/installs/boost_1_39_0/tools/build/v2/build > /local1/installs/boost_1_39_0/tools/build/v2/tools > /local1/installs/boost_1_39_0/tools/build/v2/contrib > /local1/installs/boost_1_39_0/tools/build/v2/. for user-config > configuration file user-config.jam . > notice: Loading user-config configuration file user-config.jam from > /usr/home/ssnyders/user-config.jam . > notice: Searching ../../../.. for project-config configuration file > project-config.jam . > notice: Loading project-config configuration file project-config.jam from > ../../../../project-config.jam . > notice: [python-cfg] Configuring python... > notice: [python-cfg] user-specified version: "2.5" > notice: [python-cfg] user-specified cmd-or-prefix: > "/vol/apps/python-2.5.2_64" > notice: [python-cfg] Checking interpreter command > "/vol/apps/python-2.5.2_64/bin/python2.5"... > notice: [python-cfg] running command > '"/vol/apps/python-2.5.2_64/bin/python2.5" -c "from sys import *; print > 'version=%d.%d\nplatform=%s\nprefix=%s\nexec_prefix=%s\nexecutable=%s' % > (version_info[0],version_info[1],platform,prefix,exec_prefix,executable)" > 2>&1' > notice: [python-cfg] ...requested configuration matched! > notice: [python-cfg] Details of this Python configuration: > notice: [python-cfg] interpreter command: > "/vol/apps/python-2.5.2_64/bin/python2.5" > notice: [python-cfg] include path: > "/vol/apps/python-2.5.2_64/include/python2.5" > notice: [python-cfg] library path: > "/vol/apps/python-2.5.2_64/lib/python2.5/config" > "/vol/apps/python-2.5.2_64/lib" > notice: iostreams: using prebuilt zlib > notice: iostreams: using prebuilt bzip2 > error: No best alternative for /python_for_extensions > next alternative: required properties: <python>2.5 <target-os>linux > matched > next alternative: required properties: <python>2.5 <target-os>linux > matched > /local1/installs/boost_1_39_0/tools/build/v2/build/property.jam:613: in > find-replace from module object(property-map)@1 > error: Ambiguous key <target-type>PYTHON_EXTENSION <asynch-exceptions>off > <debug-symbols>on <exception-handling>on <extern-c-nothrow>off > <hardcode-dll-paths>true <host-os>linux <inlining>off > <install-dependencies>off <link>shared <main-target-type>PYTHON_EXTENSION > <optimization>off <os>LINUX <pch>on <preserve-test-targets>on > <profiling>off <python-debugging>off > <python.interpreter>/vol/apps/python-2.5.2_64/bin/python > <python.interpreter>/vol/apps/python-2.5.2_64/bin/python2.5 <python>2.5 > <rtti>on <runtime-debugging>on <runtime-link>shared <stdlib>native > <suppress-import-lib>true <symlink-location>project-relative > <target-os>linux <target>object(file-target)@408 <threadapi>pthread > <threading>single <toolset-gcc:version>4.3.3 <toolset>gcc > <user-interface>console <variant>debug <warnings-as-errors>off > <warnings>on > /local1/installs/boost_1_39_0/tools/build/v2/build/property.jam:590: in > object(property-map)(a)1.find from module object(property-map)@1 > /local1/installs/boost_1_39_0/tools/build/v2/build/type.jam:335: in > generated-target-ps-real from module type > /local1/installs/boost_1_39_0/tools/build/v2/build/type.jam:359: in > generated-target-ps from module type > /local1/installs/boost_1_39_0/tools/build/v2/build/type.jam:270: in > type.generated-target-suffix from module type > /local1/installs/boost_1_39_0/tools/build/v2/build/virtual-target.jam:501: > in virtual-target.add-prefix-and-suffix from module virtual-target > /local1/installs/boost_1_39_0/tools/build/v2/build/virtual-target.jam:460: > in _adjust-name from module object(file-target)@408 > /local1/installs/boost_1_39_0/tools/build/v2/build/virtual-target.jam:243: > in abstract-file-target.__init__ from module object(file-target)@408 > /local1/installs/boost_1_39_0/tools/build/v2/build/virtual-target.jam:553: > in object(file-target)(a)408.__init__ from module object(file-target)@408 > /local1/installs/boost_1_39_0/tools/build/v2/kernel/class.jam:88: in > class.new from module class > /local1/installs/boost_1_39_0/tools/build/v2/build/generators.jam:565: in > generator.generated-targets from module object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/tools/builtin.jam:893: in > linking-generator.generated-targets from module > object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/tools/unix.jam:67: in > generated-targets from module object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/build/generators.jam:462: in > construct-result from module object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/build/generators.jam:415: in > run-really from module object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/build/generators.jam:390: in > generator.run from module object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/tools/builtin.jam:808: in > linking-generator.run from module object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/tools/unix.jam:41: in > unix-linking-generator.run from module object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/tools/gcc.jam:506: in > object(gcc-linking-generator)(a)403.run from module > object(gcc-linking-generator)@403 > /local1/installs/boost_1_39_0/tools/build/v2/build/generators.jam:1000: in > try-one-generator-really from module generators > /local1/installs/boost_1_39_0/tools/build/v2/build/generators.jam:1062: in > try-one-generator from module generators > /local1/installs/boost_1_39_0/tools/build/v2/build/generators.jam:1274: in > construct-really from module generators > /local1/installs/boost_1_39_0/tools/build/v2/build/generators.jam:1346: in > generators.construct from module generators > /local1/installs/boost_1_39_0/tools/build/v2/build/targets.jam:1437: in > construct from module object(typed-target)@77 > /local1/installs/boost_1_39_0/tools/build/v2/build/targets.jam:1250: in > object(typed-target)(a)77.generate from module object(typed-target)@77 > /local1/installs/boost_1_39_0/tools/build/v2/build/targets.jam:710: in > generate-really from module object(main-target)@385 > /local1/installs/boost_1_39_0/tools/build/v2/build/targets.jam:682: in > object(main-target)(a)385.generate from module object(main-target)@385 > /local1/installs/boost_1_39_0/tools/build/v2/build/targets.jam:258: in > object(project-target)(a)74.generate from module object(project-target)@74 > /local1/installs/boost_1_39_0/tools/build/v2/build-system.jam:700: in load > from module build-system > /local1/installs/boost_1_39_0/libs/python/example/../../../tools/build/v2/kernel/modules.jam:283: > in import from module modules > /local1/installs/boost_1_39_0/libs/python/example/../../../tools/build/v2/kernel/bootstrap.jam:138: > in boost-build from module > /local1/installs/boost_1_39_0/libs/python/example/boost-build.jam:7: in > module scope from module > > ---------------------------------------------- > -- View this message in context: http://www.nabble.com/Compiling-python-extensions-examples-fail-tp24526932p… Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

Race condition with Boost::Thread
by Rob Yull 27 Aug '09

27 Aug '09

I've been writing a thread pool class as kind of an experiment with boost.threads, boost.bind, and boost.function. The code below is the worker thread, that will eventually pull jobs from a queue, execute them, then look and see if there is another job in the queue. I think I've got an interesting race condition. The problem I'm having is when the class goes out of scope. I'm getting an access violation from the run() member function (which is function being run by m_Thread). Right now, run() is just entering and leaving (no actual queue checking). By the time run() is getting executed, the class is starting to execute the destructor. I thought that having the join() call in the destructor should allow the thread to finish before the class data is destroyed. In Visual Studio 2005's debugger, when it is getting to run(), it looks like the object has already been destroyed (this = 0x7fffffff), so the access violation is coming from the assignment to the m_Running member variable. If I take out the assignment, and run the debug, when it is entering run(), the this pointer is still showing as invalid, but the function completes and ends up back at the join() in the destructor, where the this pointer in the destructor is valid again. So as far as the destructor seems to be conserned, the class hasn't been destroyed yet, but as far as the member class is, it has been destroyed already. If I do pad some instructions/time before the class goes out of scope, run() executes just fine (this pointer is valid). I have also tried moving the creation of m_Thread outside of the constructor to a separate member function call, but the result is the same. I know debugging multithreaded apps can be difficult. Is there right out something I'm doing wrong here or making some kind of bad assumpion about the execution of run()? #include <boost/config.hpp> #include <boost/thread.hpp> #include <boost/functional.hpp> class WorkerThread { public: typedef boost::function<void ()> JobType; WorkerThread() { m_Thread = boost::thread(boost::bind(&WorkerThread::run, boost::ref(this))); m_Running = true; } ~WorkerThread() { m_Thread.join(); } const bool isRunning() const { return m_Running; } private: volatile bool m_Running; boost::thread m_Thread; void run() { m_Running = false; // Access violation here } private: WorkerThread(const WorkerThread& rhs); WorkerThread& operator = (const WorkerThread& rhs); }; int main() { WorkerThread wt; // new thread should start executing here return 0; } -Rob Yull

4 6

generic programming : variadic virtual function
by Stephen Nuchia 27 Aug '09

27 Aug '09

I'm finally doing something nontrivial with templates and I've painted myself into an interesting corner. boost::enable_if looks like it might be a solution but I've not made it work yet. extracting the essence of the situation: template<typename S, typename W> class baseclass { virtual void operator()(S s, W w) = 0; } template<typename S> class baseclass<S,void> { virtual void operator()(S s) = 0; } template<typename S, typename W = void> class derived : public baseclass<S,W> { void operator()(S s); // if W is void void operator()(S s, W w); // otherwise } How can I make this work? MSVC9, if that makes a difference. I tried adapting some examples from the enable_if documentation but that led to wandering around aimlessly making random changes to the text. TIA, -swn

3 4

[serialization] skipping top node "<boost_serialization signature...>"
by Sebastian Mach 27 Aug '09

27 Aug '09

Good morning everyone I'd like to use boost::serialization for the generation of xml-files. The format of those files shall not be specific to or dependent of boost, i.e. I'd like boost to not write the root-node of "<boost_serialization signature="serialization::archive" version=...>". E.g., instead of <?xml version="1.0" encoding="UTF-8" standalone="yes" ?> <!DOCTYPE boost_serialization> <boost_serialization signature="serialization::archive" version="4"> <picogen-qt-scene class_id="0" tracking_level="0" version="0"> <comment>hello world!</comment> <fenum>two</fenum> </picogen-qt-scene> </boost_serialization> I would prefer <?xml version="1.0" encoding="UTF-8" standalone="yes" ?> <picogen-qt-scene class_id="0" tracking_level="0" version="0"> <comment>hello world!</comment> <fenum>two</fenum> </picogen-qt-scene> . I wonder if this is possible without tweakage of boost components (meaning that I would have to fork or write patches). I tried the serialization-documentation and google, but could not find appropriate information (or maybe I missed the right words). Best Regards Sebastian Mach http://picogen.org

3 5

[boost-users][python] Problem with lifetime of object?
by Germán Diago 27 Aug '09

27 Aug '09

Hello. I would like to know what's wrong with this: Object * create() { return new Object(); } void addToScene(const std::string & name, Object * o) { //Store object in a std::multimap<std::string, Object *> } BOOST_PYTHON_MODULE(mod) { def("createObject", &create, return_value_policy<reference_existing_object>()); .... } I start my application, it works ok and it uses addToScene and create: Now I try to add a new object to my application from python: >> obj = Object() >> addToScene("obj", obj) Should the problem have anything to do with object lifetime? Or the segmentation fault is because of another issue? Thanks in advance.

2 2

[asio] io_service::run() never returns and is blocking my thread
by Robert Dailey 27 Aug '09

27 Aug '09

Hello, I'm currently using a serial_port object to call async_read_some(). This results in io_service::run() blocking until at least 1 byte or more can be read from the serial port. If nothing is ever sent, then I can not safely exit the thread in which io_service::run() is executing in. In other words, a call to join() on that thread hangs indefinitely until data is received. My completion callback is not executed at all until data arrives for reading. How can I force run() to return when I interrupt my thread? --------- Robert Dailey

3 4

Problem when using Boost.Asio at Fedora 11
by Davi Vercillo C. Garcia 27 Aug '09

27 Aug '09

Hi all, I'm new at C++ and Boost and I'm with some problems when I try to use Boost.Asio library. I was trying to build the first example from Boost.Asio's tutorial. The code is like this: ===== CODE ===== #include <iostream> #include <boost/asio.hpp> #include <boost/date_time/posix_time/posix_time.hpp> int main( int argc, char **argv ) { boost::asio::io_service io; boost::asio::deadline_timer t(io, boost::posix_time::seconds(5)); t.wait(); std::cout << "Opa !" << std::endl; return 0; } ===== CODE ===== I'm trying to compile using this command: g++ -I/usr/include/boost/ example.cpp -o example", and I'm getting this error: ==== ERROR ==== /tmp/ccv68fST.o: In function `__static_initialization_and_destruction_0(int, int)': timer.cpp:(.text+0x51): undefined reference to `boost::system::get_system_category()' timer.cpp:(.text+0x5b): undefined reference to `boost::system::get_generic_category()' timer.cpp:(.text+0x65): undefined reference to `boost::system::get_generic_category()' timer.cpp:(.text+0x6f): undefined reference to `boost::system::get_generic_category()' timer.cpp:(.text+0x79): undefined reference to `boost::system::get_system_category()' /tmp/ccv68fST.o: In function `boost::asio::error::get_system_category()': timer.cpp:(.text._ZN5boost4asio5error19get_system_categoryEv[boost::asio::error::get_system_category()]+0x7): undefined reference to `boost::system::get_system_category()' /tmp/ccv68fST.o: In function `boost::asio::detail::posix_tss_ptr<boost::asio::detail::call_stack<boost::asio::detail::task_io_service<boost::asio::detail::epoll_reactor<false> > >::context>::~posix_tss_ptr()': timer.cpp:(.text._ZN5boost4asio6detail13posix_tss_ptrINS1_10call_stackINS1_15task_io_serviceINS1_13epoll_reactorILb0EEEEEE7contextEED2Ev[boost::asio::detail::posix_tss_ptr<boost::asio::detail::call_stack<boost::asio::detail::task_io_service<boost::asio::detail::epoll_reactor<false> > >::context>::~posix_tss_ptr()]+0xf): undefined reference to `pthread_key_delete' /tmp/ccv68fST.o: In function `boost::asio::detail::posix_tss_ptr<boost::asio::detail::call_stack<boost::asio::detail::task_io_service<boost::asio::detail::epoll_reactor<false> > >::context>::posix_tss_ptr()': timer.cpp:(.text._ZN5boost4asio6detail13posix_tss_ptrINS1_10call_stackINS1_15task_io_serviceINS1_13epoll_reactorILb0EEEEEE7contextEEC2Ev[boost::asio::detail::posix_tss_ptr<boost::asio::detail::call_stack<boost::asio::detail::task_io_service<boost::asio::detail::epoll_reactor<false> > >::context>::posix_tss_ptr()]+0x22): undefined reference to `pthread_key_create' collect2: ld returned 1 exit status ==== ERROR ==== I already installed Boost and Boost-devel packages. Does anyone know what is missing ? Thanks, -- Davi Vercillo C. Garcia B.Sc. Student - DCC-IM/UFRJ Fedora Project Contributor http://davivercillo.fedorapeople.org/ "If a million people say a foolish thing, it is still a foolish thing." - Anatole France

3 4

lock-free shared_ptr on solaris
by Frank Meerkötter 27 Aug '09

27 Aug '09

Hello, from looking at [1] i understood that there is no lock-free implementation of shared_ptrs on solaris. When snooping around in the source i noticed that there is a ./smart_ptr/detail/atomic_count_solaris.hpp It is using the solaris specific <atomic.h> to implement an atomic counter. frank@dexter:~/boost_1_40_0/boost$ find -type f | xargs grep atomic_count_solaris.hpp ./smart_ptr/detail/atomic_count_solaris.hpp:// boost/detail/atomic_count_solaris.hpp But apparently it is not in use. I am currently wondering why? KR, Frank [1] http://www.boost.org/doc/libs/1_40_0/libs/smart_ptr/shared_ptr.htm

1 0

Implementing a type-and-effect system
by david raila 27 Aug '09

27 Aug '09

I'm pondering the use of boost to implement a type-and-effect static checker which consists of: 1. a system to define and name annotations 2. a way to declare and associate class members with the annotations 3. a way to declare effects on methods that say how the method interacts with the annotation 4. a way to capture uses of the members in the methods and verify that the effects are satisfied A simple example might be logical heap partitioning and interference checking which in the *abstract* might look something like this: class foo { heapregion a; heapregion b; foo * left in a; foo *right in b; foo *link(); void update() reads a, writes b { right = a->link(); // ok left = /*anything*/ // fails, writing a } This is one very simple example, but hopefully illustrates the idea. Ideally this can be all done statically at compile time without any run-time overhead, even if that means a two-phase compilation process, a checking phase that instantiates the checking code, and an optimized compile phase that might not. If this can be done via metaprogramming then we can avoid writing a new refactoring tool and/or modifying an existing front end. I've been trying to determine whether to try and implement a custom system or use proto or is there something else that might help in this effort? Is anyone aware of similar uses of boost, other libraries, or have advice and/or feedback. Thanks. David

2 2