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

custom allocators, pool_alloc, more questions
by B Hart 18 Mar '10

18 Mar '10

I asked for explanation about custom allocators, but unfortunately the answers I got led to more questions which didn't get answered. Supposing a code block: { std::set<unsigned int, std::less<unsigned int>, boost::pool_allocator<unsigned int>> v; const int K = 10485750; for(int i=0; i<K; i++) { v.insert(i); } } //Does memory get released automatically here? or after the line: boost::singleton_pool<boost::pool_allocator_tag, sizeof(unsigned int)>::release_memory(); //what is release_memory() doing besides just setting the node type to a default? Why does that matter? Last, is the life of the singleton instance for the duration of program execution, or just the codeblock?

1 0

Spirit question
by philip tucker 17 Mar '10

17 Mar '10

Please see attached code. Why does line A compile while line B doesn't? Thanks.

4 5

program option library segfaults for exception
by Neetu Garg 17 Mar '10

17 Mar '10

Sorry for the duplicate post, I think my message posted under different conversation. Hi, I have application that is using boost program option library. If I provide unknown option via CLI it should throw exception. The library is statically linked on windows and Linux environment. It works fine on windows platform I do get exception for unknown CLI option. On Linux platform I get "segfault" in place of exception for unknown CLI option. It is s a multi-threaded application and it also uses boost thread library. Application does throw exception on Linux also if I link dynamically. I am using following flags to link the library statically: LDFLAGS= -fexceptions -static-libgcc -Wl,-static -L/usr/lib -lboost_program_options-gcc41-mt -lboost_filesystem-gcc41-mt -lboost_thread-gcc41-mt -lboost_system-gcc41-mt -lnet7ssh -lbotan -ltar -lz -lrt -lpthread I am using following version of gcc: # g++ --version g++ (GCC) 4.1.1 Copyright (C) 2006 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. Here is the core dump for the segfault: (gdb) backtrace #0 __gnu_internal::get_global () at /usr/software/src/gcc-4.1.1/libstdc++-v3/libsupc++/eh_globals.cc:58 #1 0x082c0f67 in __cxa_get_globals () at /usr/software/src/gcc-4.1.1/libstdc++-v3/libsupc++/eh_globals.cc:71 #2 0x082c0b3f in __cxa_allocate_exception (thrown_size=32) at /usr/software/src/gcc-4.1.1/libstdc++-v3/libsupc++/eh_alloc.cc:154 #3 0x080a71f8 in boost::program_options::detail::cmdline::finish_option () #4 0x080a77eb in boost::program_options::detail::cmdline::run () #5 0x0806f3aa in boost::program_options::basic_command_line_parser<char>::run () #6 0x080660e9 in CLIHandler::collectCLIOptions () #7 0x0808a0d1 in perfstat::runPerfstat () #8 0x0808af3d in main () Current language: auto; currently c++ Appreciate your help Regards, NG

1 0

Signal code: Address not mapped (1) eroor
by alev mutlu 17 Mar '10

17 Mar '10

Hi, I am getting the following run time errors for the code below. Any suggestions? _alev [ceng34:00724] *** Process received signal *** [ceng34:00724] Signal: Segmentation fault (11) [ceng34:00724] Signal code: Address not mapped (1) [ceng34:00724] Failing at address: 0xfffffffffffffff8 [ceng34:00724] [ 0] /lib64/libpthread.so.0 [0x315880de80] [ceng34:00724] [ 1] /usr/lib64/libstdc++.so.6(_ZN9__gnu_cxx18__exchange_and_addEPVii+0x2) [0x385dab7672] [ceng34:00724] [ 2] /usr/lib64/libstdc++.so.6(_ZNSs6assignERKSs+0x9b) [0x385da9ca4b] [ceng34:00724] [ 3] test-boost(_ZN5fieldaSERKS_+0x47) [0x42197f] [ceng34:00724] [ 4] test-boost(_ZNSt6__copyILb0ESt26random_access_iterator_tagE4copyIPK5fieldPS3_EET0_T_S8_S7_+0x3c) [0x4221be] [ceng34:00724] [ 5] test-boost(_ZSt10__copy_auxIPK5fieldPS0_ET0_T_S5_S4_+0x29) [0x422203] [ceng34:00724] [ 6] test-boost(_ZNSt13__copy_normalILb0ELb0EE6copy_nIPK5fieldPS2_EET0_T_S7_S6_+0x25) [0x42222b] [ceng34:00724] [ 7] test-boost(_ZSt4copyIPK5fieldPS0_ET0_T_S5_S4_+0x2d) [0x42225b] [ceng34:00724] [ 8] test-boost(_ZN5boost3mpi6detail12scatter_implI5fieldEEvRKNS0_12communicatorEPKT_PS7_iiN4mpl_5bool_ILb0EEE+0x90) [0x42a630] [ceng34:00724] [ 9] test-boost(_ZN5boost3mpi7scatterI5fieldEEvRKNS0_12communicatorEPKT_PS6_ii+0x50) [0x42b162] [ceng34:00724] [10] test-boost(_ZN5boost3mpi7scatterI5fieldEEvRKNS0_12communicatorERKSt6vectorIT_SaIS7_EEPS7_ii+0x55) [0x42b1db] [ceng34:00724] [11] test-boost(main+0x20e) [0x420880] [ceng34:00724] [12] /lib64/libc.so.6(__libc_start_main+0xf4) [0x3157c1d8b4] [ceng34:00724] [13] test-boost(__gxx_personality_v0+0x169) [0x420489] [ceng34:00724] *** End of error message *** mpirun noticed that job rank 0 with PID 724 on node ceng34-ib exited on signal 11 (Segmentation fault). #include <boost/mpi/environment.hpp> #include <boost/mpi/communicator.hpp> #include <boost/mpi/collectives.hpp> #include <boost/thread/barrier.hpp> #include <boost/thread/mutex.hpp> #include <boost/config.hpp> #include <boost/archive/text_oarchive.hpp> #include <boost/archive/text_iarchive.hpp> #include <boost/serialization/access.hpp> #include <boost/serialization/string.hpp> #include <boost/mpi/skeleton_and_content.hpp> #include <boost/mpi/datatype.hpp> #include <boost/mpi/allocator.hpp> #include <boost/archive/tmpdir.hpp> #include <boost/serialization/utility.hpp> #include <boost/serialization/base_object.hpp> #include <boost/mpi.hpp> #include <boost/tokenizer.hpp> #include <boost/archive/tmpdir.hpp> #include <boost/archive/binary_oarchive.hpp> #include <boost/serialization/export.hpp> #include <boost/serialization/base_object.hpp> #include <boost/serialization/utility.hpp> #include <boost/serialization/hash_map.hpp> #include <boost/serialization/map.hpp> #include <string> #include <iostream> #include <vector> #include <fstream> #define MASTER 0 int arraySize = 7; using namespace std; namespace mpi = boost::mpi; struct field { string field_name; string field_type; string field_ref; string field_colType; }; namespace boost { namespace serialization{ template<class Archive> void serialize(Archive & ar, struct field & f, unsigned int version){ ar & f.field_name; ar & f.field_type; ar & f.field_ref; ar & f.field_colType; } } } int main(int argc, char* argv[]) { mpi::environment env(argc, argv); mpi::communicator world; int m_mySize, w_mySize, kk,i; struct field *relAllValues; struct field *relMyValues; vector<struct field> C; struct field c; w_mySize = arraySize / world.size(); cout << "w_mySize = " << w_mySize << endl; if (world.rank() == MASTER){ for(int i = 0; i < arraySize; i++){ c.field_name = "field_name=" ; c.field_type = "field_type="; c.field_ref = "field_ref="; c.field_colType = "field_colType="; C.push_back(c); } m_mySize = arraySize / world.size() + arraySize % world.size(); relMyValues = (struct field *) malloc(sizeof(struct field) * w_mySize); relAllValues = (struct field *) malloc(sizeof(struct field) * arraySize); } else{ relMyValues = (struct field *) malloc(sizeof(struct field) * w_mySize); } if (world.rank() == MASTER){ scatter(world, C, relMyValues, w_mySize, 0); } else{ scatter(world, C, relMyValues, w_mySize, 0); } if (world.rank() == 1){ cout << "MY RANK IS " << world.rank() << endl; } }

2 1

[Phoenix] : Using phoenix with zip_iterator
by Manjunath Kudlur 17 Mar '10

17 Mar '10

I have the following program where I make_zip_iterator on 2 vector<int>::iterator's and am trying to pass a Phoenix expression to std::for_each on the iterator. I wrote a lazy version of tuple.get<N> for that purpose. I get a compile error because the phoenix functor only accepts references, whereas the tuple element seems to get passed by value. What am I doing wrong? Manjunath #include <iostream> #include <vector> #include <algorithm> #include <boost/iterator/zip_iterator.hpp> #include <boost/tuple/tuple.hpp> #include <boost/spirit/include/phoenix_core.hpp> #include <boost/spirit/include/phoenix_operator.hpp> #include <boost/spirit/include/phoenix_statement.hpp> #include <boost/spirit/include/phoenix_scope.hpp> #include <boost/spirit/home/phoenix/function/function.hpp> using namespace boost::phoenix; using namespace boost::phoenix::arg_names; using namespace boost::phoenix::local_names; using namespace std; namespace fusion = boost::fusion; struct print_tuple { void operator()(const boost::tuple<int&, int&>& t) const { cout << t.get<0>() << " " << t.get<1>() << "\n"; } }; struct get0_ { template<typename Arg> struct result { typedef int type; }; template<typename Arg> int operator()(Arg a1) const { return a1.get<0>(); } }; function<get0_> get0; int main() { vector<int> a(10), b(10); for(unsigned i=0, e=a.size(); i!=e; ++i) { a[i] = i; b[i] = i; } std::for_each(a.begin(), a.end(), cout << arg1); std::for_each(boost::make_zip_iterator(boost::make_tuple(a.begin(), b.begin())), boost::make_zip_iterator(boost::make_tuple(a.end(), b.end())), cout << get0(arg1)); }

3 10

[iostreams][serialization] compressed UTF-8 XML archive
by Paul 17 Mar '10

17 Mar '10

(Was: Re: [Boost-users] Filesystem, serialization, character encoding and portable software) On 14 August 2008 00:44, Robert Ramey <ramey(a)rrsd.com> wrote: > Daniel Krügler wrote: > > Robert Ramey wrote: > >> the wide character xml archives use UTF8. > >> > >> the narrow character xml archives use the currently set locale. > >> > >> Robert Ramey > > > > Sorry for asking offhand: > > > > What is the reasoning behind this different behaviour? > > I assumed that most programs built with narrow characters used > the locale concept to deal with this. > > Wide character systems lend themselves to UTF coding so I > used that for wide char archives. In order to do this, I used > Ron Garcia's UTF code conversion facet for streams. > > It would be quite easy to generate UTF coding for narrow > character archives. Just do the following: > > a) Build the UTF code conversion facet for narrow character > input (its templated on character type). > > b) When the stream is opened, attach this facet to the stream. > > Note the the output char format is not really a property of the > serialization > library, but rather an artifact of the way it has been used. That is, the > serialization library depends on the standard stream library for this > property. > > I understand that to get a UTF-8 XML file, I need to use a wide-character stream, but I can't seem to understand how I can compress a serialized UTF-8 xml archive... it seems I can either do: namespace bio = boost::iostreams; namespace fs = boost::filesystem; fs::wofstream ofs(path,ios::binary); bio::filtering_stream<bio::output,wchar_t> out; out.push(bio::gzip_compressor()); // PROBLEM: gzip_compressor only accepts char out.push(ofs); boost::archive::xml_woarchive oa(out); --- OR --- fs::ofstream ofs(path,ios::binary); bio::filtering_stream<bio::output> out; out.push(bio::gzip_compressor()); out.push(ofs); boost::archive::xml_woarchive oa(out); // PROBLEM: only accepts wostream I've googled but haven't found anything relevant... either people use an ifstream with gzip_compressor, or they don't use gzip_compressor at all. thanks, Paul

3 2

[boost-users][test] Does memory leak detection work in linux?
by Andrey Torba 17 Mar '10

17 Mar '10

Hi all, Does memory leak detection work in linux? -- Regards, Andrey

3 4

How to write a config file parser with boost::spirit that supports line comments?
by Lars Rohwedder 17 Mar '10

17 Mar '10

Hi there, I want to write a parser for config files having many entries with this format: <name> "{" <token> [ <token> ... ] "}" The format is "free-form" so there can be line breaks and other whitespaces within each entry between the tokens etc. <name> is chset_p("a-zA-Z_"), <token> is chset_p("a-zA-Z0-9_.-"). This parser works fine so far. I extended the format to allow even whitespaces etc. in the tokens. So <token> is changed to (I omit the assign_a actions): ( +chset_p("a-zA-Z0-9_.-") | ('"' >> +(!ch_p('"')) >> '"') | ('\'' >> +(!ch_p('\'')) >> '\''; which also works fine as I tested. (I know, newlines would be allowed in these tokens but I hope noone use that.) But now I want to extend the format to also allow line comments, starting with '#', ending with '\r' or '\n', and I failed: Just scanning for '#' also recognized '#' within a quoted token. But that is not the start of a comment. :-( But first searching for quotes also catched unbalanced quote chars in comments and confused the parser. What would be a working solution for this - IMHO still simple - format? (I should mention that I am stuck with boost 1.35 so I cannot use spirit2 or all the other new features in the current release. :-(( ) Lars R.

1 0

custom allocators Re:pool_alloc
by B Hart 17 Mar '10

17 Mar '10

I am not understanding the use of pool with std::set. i.e. the release_memory() function referenced in last example shown on: http://www.boost.org/doc/libs/1_42_0/libs/pool/doc/interfaces.html seems to do nothing. I call GetProcessMemoryInfo() and don't see any change after the set has gone out of scope. maybe someone could relate what happens when set goes out of scope and release_memory() isn't called versus set goes out of scope and release_memory() is called. I want to do something like: 1.std::set<unsigned int, std::less<unsigned int>, boost::pool_allocator<unsigned int>> v; 2. for(int i=0; i<(K); i++) { v.insert(i); } 3. use v 4. release_memory() on v //dump all set elements in one action 5. repeat 1-4 with a new set

1 0

[serialization] xml files that can be hand-edited
by Paul 17 Mar '10

17 Mar '10

Hi, I'd like to build some insurance into my XML files. If there is some sort of problem in the future (either with boost, or my application), I'd like to be able to hand-edit the serialized XML archive files. Even better would be to allow XML-tools to manipulate a batch of the XML files, but one thing at a time. One big problem with the current XML format is that the class_id="3" tracking_level="0" version="0" preamble is only included the FIRST time a class is dumped to the file. Example: <brother class_id="3" tracking_level="0" version="2"> .. etc, and then later in the file ... <sister> ... etc ... If something happened (eg impossible-to-upgrade-format-change) and I needed to manually move the <sister> item above the <brother> item (maybe into a different branch of tags), then I would need to KNOW that the brother class is the same as the sister class (which is not obvious) AND move the preamble into the sister tag. I would also need to re-number all the class_ids, but those appear to be only used if I'm tracking a pointer AND I have multiple instances of that pointer in the file. If the item is not a pointer, it looks like the class_id is *always* thrown away. So, I'm thinking of: a) never allow tracking (its a compromise to give me the next item) b) never writing out class_id or tracking_level c) always writing out the version the result would look like this: <brother version="2"> .. etc, and then later in the file ... <sister version="2"> That way items in the XML file could be rearranged by hand and I would not need to do the difficult task of adjusting the tag attributes. Does this sound sensible? I had a look into inheriting and creating my own archive, but it appears all of the changes required happen in the basic_oarchive_impl, which seems pretty fundamental and hard to override. Will I need to create and maintain my own patch in order to "hack" this change. Thanks, Paul

2 1