March 2010 - Boost-users - lists.preview.boost.org

Boost::spirit program crashes
by Lars Rohwedder 26 Mar '10

26 Mar '10

Here my stripped version still crashes with segfault: ----<snio>---- #include <boost/spirit.hpp> #include <iostream> using namespace boost::spirit; struct Configurable { const char* name; // name of the configurable module const rule<>& parser; // parser for config items, normally // a couple of OR-connected terms. }; rule<> __ = ch_p(' '); // separator char. was a bit more complex // in the original code std::string conn, port, ident_name, ident_pw; /* // the original parser crashes with segfault: const Configurable cfg = { "Hardtest", +( *__ | ("connect" >> __ >> str_p[assign_a(conn)] >> ':' >> int_p[assign_a(port)]) | ("identity" >> __ >> str_p[assign_a(ident_name)] >> __ >> str_p[assign_a(ident_pw)] ) | "hard" ) }; */ // a simplicistic parser that still crashes: const Configurable cfg = { "Hardtest", +( __ ) }; int main(int argc, char** argv) { // original config string: // "connect foo.bar:123 identity root fF2.oby!sy" or the like. parse("hard ", cfg.parser); // <--- CRASHES! // this works: // parse("hard ", +(__) ); }; ----<snap>---- The stack trace of GNU gdb is: Program received signal SIGSEGV, Segmentation fault. 0x000000000040184e in boost::spirit::impl::rule_base<boost::spirit::rule<boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t>, boost::spirit::rule<boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t> const&, boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t>::parse_main<boost::spirit::scanner<char const*, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> > > (this=0x7fff29fcea10, scan=@0x7fff29fcea00) at /usr/include/boost/spirit/core/non_terminal/impl/rule.ipp:189 189 hit = rule_base_access::get(*derived_this) (gdb) bt #0 0x000000000040184e in boost::spirit::impl::rule_base<boost::spirit::rule<boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t>, boost::spirit::rule<boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t> const&, boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t>::parse_main<boost::spirit::scanner<char const*, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> > > (this=0x7fff29fcea10, scan=@0x7fff29fcea00) at /usr/include/boost/spirit/core/non_terminal/impl/rule.ipp:189 #1 0x00000000004018f7 in boost::spirit::impl::rule_base<boost::spirit::rule<boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t>, boost::spirit::rule<boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t> const&, boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t>::parse<boost::spirit::scanner<char const*, boost::spirit::scanner_policies<boost::spirit::iteration_policy, boost::spirit::match_policy, boost::spirit::action_policy> > > (this=0x7fff29fcea10, scan=@0x7fff29fcea00) at /usr/include/boost/spirit/core/non_terminal/impl/rule.ipp:171 #2 0x000000000040197b in boost::spirit::parse<char const*, boost::spirit::rule<boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t> > (first_=@0x7fff29fcea58, last=@0x7fff29fcea68, p=@0x7fff29fcea10) at /usr/include/boost/spirit/core/impl/parser.ipp:28 #3 0x0000000000401a34 in boost::spirit::parse<char, boost::spirit::rule<boost::spirit::nil_t, boost::spirit::nil_t, boost::spirit::nil_t> > (str=0x402420 "hard ", p=@0x7fff29fcea10) at /usr/include/boost/spirit/core/impl/parser.ipp:45 #4 0x0000000000400973 in main (argc=1, argv=0x7fff29fceb88) at cfgTest.cc:38 I am using boost 1.35 (yes, it is quite old. Is this a bug in boost that is fixed in newer versions? I didn't found something related in the changelogs) on Linux x86_64, gcc 4.3.2 (Debian "Lenny"). Perhaps somebody knows an answer except "use spirit 2!" :-) Lars R.

2 3

error with "-std=c++0x" flag
by girish hilage 26 Mar '10

26 Mar '10

Hi, I am using "-std=c++0x" flag while compiling my program but it gives me errors as follows : /usr/lib/gcc/i586-redhat-linux/4.4.0/../../../../include/c++/4.4.0/tr1_impl/type_traits: In function ‘bool boost::filesystem::is_empty(const boost::filesystem::path&)’: /usr/lib/gcc/i586-redhat-linux/4.4.0/../../../../include/c++/4.4.0/tr1_impl/type_traits:326: error: ‘template<class _Tp> struct std::is_empty’ is not a function, /usr/include/boost/filesystem/operations.hpp:361: error: conflict with ‘template<class Path> typename boost::enable_if<boost::filesystem::is_basic_path<Path>, bool>::type boost::filesystem::is_empty(const Path&)’ /usr/include/boost/filesystem/operations.hpp:660: error: in call to ‘is_empty’ /usr/lib/gcc/i586-redhat-linux/4.4.0/../../../../include/c++/4.4.0/tr1_impl/type_traits: In function ‘bool boost::filesystem::is_empty(const boost::filesystem::wpath&)’: /usr/lib/gcc/i586-redhat-linux/4.4.0/../../../../include/c++/4.4.0/tr1_impl/type_traits:326: error: ‘template<class _Tp> struct std::is_empty’ is not a function, /usr/include/boost/filesystem/operations.hpp:361: error: conflict with ‘template<class Path> typename boost::enable_if<boost::filesystem::is_basic_path<Path>, bool>::type boost::filesystem::is_empty(const Path&)’ /usr/include/boost/filesystem/operations.hpp:662: error: in call to ‘is_empty’ Is there any workaround to this problem? Regards, Girish

1 0

[MPI] Code example leaks memory
by Yves Daemen 26 Mar '10

26 Mar '10

1 0

timed waits
by Just C 26 Mar '10

26 Mar '10

I was looking at this article about conditon_variables. http://www.justsoftwaresolutions.co.uk/threading/condition-variable-spuriou… For a timeout, the article has the following sample program with condition variables. template<typename Duration> bool timed_wait_and_pop(Data& popped_value, Duration const& wait_duration) { boost::system_time const timeout=boost::get_system_time()+wait_duration; boost::mutex::scoped_lock lock(the_mutex); while(the_queue.empty()) { if(!the_condition_variable.timed_wait(lock,timeout)) return false; } popped_value=the_queue.front(); the_queue.pop(); return true; } However, shouldn't this program instead start with boost::mutex::scoped_lock lock(the_mutex, timeout); if(!lock.owns()) return false; i.e. shouldn't it also take care of getting stuck while locking the mutex. i.e. if the timeout is 5 seconds & the getting the lock itself takes 10 seconds, then, the earlier the program will return is 10+ seconds.

2 1

template typedef?
by Roland Bock 26 Mar '10

26 Mar '10

Hi, I wonder if there is some easy way to express the following incorrect code in a correct way? typedef template<class T> boost::function<void(const T&)> foo_function; template<> foo_function<int> my_foo = bind(example_function); Regards, Roland

2 2

Re: [Boost-users] [program-options] static linked example segfaults for exception
by Neetu Garg 26 Mar '10

26 Mar '10

ThanksVladimir and David, I tried "-static" without "-static-libgcc" and "-Wl,-Bstatic" and it worked fine. My test sample that uses program option and single threaded works fine with the above change. But my other application that is multi threaded (using boost thread API) and uses program option still segfaults. When I try to add "libpthread" and "-lboost_thread-gcc41-mt" in linking I start getting segfaults. LDFLAGS= -static -L/u/neetug/Projects/performance/perfstat_cluster/lib -lpthread -lboost_program_options-gcc41-mt -lboost_filesystem-gcc41-mt -lboost_system-gcc41-mt -lboost_thread-gcc41-mt -lrt Appreciate you help Regards, NG ________________________________ From: David Kaplan <davkaplan(a)gmail.com> To: boost-users(a)lists.boost.org Sent: Mon, March 22, 2010 11:29:26 AM Subject: Re: [Boost-users] [program-options] static linked example segfaults for exception >>> g++ testprogramoptions.o -static-libgcc >>> -Wl,-Bstatic -L/u/neetug/Projects/performance/perfstat_cluster/lib >>> -lboost_program_options-gcc41-mt -lpthread -o testprogramoptions > >This is not the best way to link statically. Did you try just -static >>option, without -static-libgcc and -Wl,-Bstatic? > Or you could try: -l:libboost_program_options-gcc41- mt

2 1

[multiarray] Obtain MultiArray implementation instance from base, extents, strides
by Rhys Ulerich 25 Mar '10

25 Mar '10

Hi all, I'd like to obtain an N-dimensional MultiArray implementation given a base pointer, N extents, and N strides. My use case requires padding the stride in one dimension in a way seemingly not obtainable taking views of the usual boost::multi_array_ref. Code like the following would be ideal boost::array<std::size_t,3> extents = { 2, 3, 4 }; boost::array<std::size_t,3> strides = { 1, 2, 7 }; // Note 7 not 6 boost::scoped_array<int> raw(new int[extents[2]*strides[2]]); using boost::detail::multi_array::multi_array_view; multi_array_view<int,3> a(raw.get(), extents, strides); except that the appropriate constructor in multi_array_view is private (boost/multi_array/view.hpp:442). boost::detail::multi_array::sub_array would also be ideal if it's constructor was accessible(boost/multi_array/subarray.hpp:370). I can make either accessible by #defining BOOST_NO_MEMBER_TEMPLATE_FRIENDS, but that's evil. Am I missing something in Boost.MultiArray? Or is there no publicly accessible way to provide a custom stride list? - Rhys

1 1

asio: how to handle unreliable network/glitches/server down ?
by Avi Bahra 25 Mar '10

25 Mar '10

Ok. I am a newbie to asio so bear with me: I have the following piece of code for handling a unreliable network and/or server crashed/servicing where I do not want the client to fail. My solution basically involves looping round , sleeping and trying again. .... int no_of_tries = clientEnv.no_of_tries(); while ( no_of_tries > 0 ) { try { boost::asio::io_service io_service; Client theClient( io_service, clientEnv.get_cmd(), clientEnv.host(), clientEnv.port() ); io_service.run(); ... // Some kind of connection error fall through and try again. } catch (std::exception& e) { .... } sleep( RETRY_POLL_PERIOD ); no_of_tries--; } This does not seem right to me, but I am not sure how typical my case is. Are there any known design patterns/idioms that provide a more robust ways of handling unreliable networks/glitches and cases where the server needs to be taken down temporarily ? Best regards, Ta, Avi

1 0

[Bind][Lambda] Processing containers of pointers
by Robert Jones 25 Mar '10

25 Mar '10

Hi All I believe boost.bind has the capability to look through pointers and smart pointers to transparently process the pointed-to objects. Does boost.lambda.bind also have this capability? I.e., this bit of code (and I am specifically using boost 1.37 for this btw) #include <boost/lambda/lambda.hpp> #include <boost/lambda/bind.hpp> #include <boost/lambda/construct.hpp> #include <boost/shared_ptr.hpp> #include <algorithm> #include <vector> struct A { A( int ); void f( ); }; int main( ) { using namespace boost::lambda; std::vector< A * > pv; std::vector< boost::shared_ptr< A > > sv; std::for_each( pv.begin( ), pv.end( ), bind( &A::f, _1 ) ); std::for_each( sv.begin( ), sv.end( ), bind( &A::f, _1 ) ); // error here. } Thanks, - Rob.

4 7

Boost.Proto embedding a function calls into expressions
by Stefan Frehse 25 Mar '10

25 Mar '10

Hello Boost.Proto, firstly, thank you very much for your great work for providing such a powerful framework. My Setup: I defined a grammar, like grammar * grammar, a domain and so on. Based on the domain two evaluator for different contexts are provided. For standard operators that works great. For expression evaluation I have a function defined as follow: template<typename T, typename Expr, typename Arg0, typename Arg1, ... > generate (T& t, Expr const& expr, Arg0 const& arg0, ... ) { call the first or the second evaluator switched by T and passes the value arg0, ... } Therewith, I can write for example, generate (foo, _1 * _2, a, b); Coming to my problem. I would like to write, generate (foo, _1 * func (_2, _3), a ,b, c); Overall I would like to embedd function calls into the expression and the pass them to the generate function. I tried to implement a make_expr based on func, but that does not work for me, i.e., compiler error for fmi::ite (_0, _1, _2) (a,b,c); error: invalid use of incomplete type 'struct fmi::ite_::result<fmi::ite_(boost::fusion::vector1<const fmi::bv&>, boost::fusion::vector1<const fmi::bv&>, boost::fusion::vector1<mpl_::void_&>)>' error: declaration of 'struct fmi::ite_::result<fmi::ite_(boost::fusion::vector1<const fmi::bv&>, boost::fusion::vector1<const fmi::bv&>, boost::fusion::vector1<mpl_::void_&>)>' cpp:7: Can you imaging what the problem is? Otherwise I will attach further cleaned code. struct ite_ { template<typename Sig> struct result; template<typename This, typename If, typename Then, typename Else> struct result<This(If const&, Then const&, Else const&)> : fusion::result_of::make_vector<bv> {}; template<typename If, typename Then, typename Else> fusion::result_of::make_vector<bv>::type operator() (If const& _if, Then const& _then, Else const& _else) const { bv test; /// some more code return fusion::make_vector (test); } }; template<typename If, typename Then, typename Else> typename proto::result_of::make_expr<proto::tag::function , constraint_domain , ite_ const , If const& , Then const& , Else const&>::type ite (If const& _if, Then const& _then, Else const& _else) { return proto::make_expr<proto::tag::function, constraint_domain>( ite_() , boost::ref(_if) , boost::ref(_then) , boost::ref(_else)); } Examplarily, the placeholder definition: typedef constraint<proto::terminal<placeholder<mpl::int_<0> > >::type> var0_type Best wishes, sfrehse

2 1