September 2009 - Boost-users

Serialization demo program crashes trying to free an unallocated block
by Jean-Denis Muys 24 Sep '09

24 Sep '09

Hi, Warning: noob inside. I am new to boost, and quite a bit rusty with C+ +, having spent the past few years with other languages. Nonetheless. I downloaded, built and installed Boost 1.40 just fine on Mac OS X 10.6 (Snow Leopard). I set up a XCode 3.2 project to compile and run the demo.cpp program delivered with the Serialization library. It crashes deep inside this line: restore_schedule(new_schedule, filename.c_str()); GDB gives the following error message: malloc: *** error for object 0x1000581a0: pointer being freed was not allocated The stack crawl shows: #0 0x7fff81202ff6 in __kill #1 0x7fff812a4072 in abort #2 0x7fff811bb095 in free #3 0x7fff836031e8 in std::string::reserve #4 0x7fff836032fe in std::string::append #5 0x10028ccac in boost ::archive::text_iarchive_impl<boost::archive::text_iarchive>::load at text_iarchive_impl.ipp:55 #6 0x100024a5e in boost ::archive::load_access::load_primitive<boost::archive::text_iarchive, std::string> at iserializer.hpp:109 #7 0x100024a83 in boost ::archive ::detail::load_non_pointer_type<boost::archive::text_iarchive, std::string>::load_primitive::invoke at iserializer.hpp:332 #8 0x100024aa8 in boost ::archive ::detail::load_non_pointer_type<boost::archive::text_iarchive, std::string>::invoke at iserializer.hpp:417 #9 0x100024acd in boost::archive::load<boost::archive::text_iarchive, std::string> at iserializer.hpp:542 #10 0x100024afe in boost ::archive ::detail ::common_iarchive <boost::archive::text_iarchive>::load_override<std::string> at common_iarchive.hpp:61 #11 0x100024b2b in boost ::archive ::basic_text_iarchive <boost::archive::text_iarchive>::load_override<std::string> at basic_text_iarchive.hpp:62 #12 0x100024b59 in boost ::archive ::text_iarchive_impl <boost::archive::text_iarchive>::load_override<std::string> at text_iarchive.hpp:66 #13 0x100024b8c in boost ::archive ::detail ::interface_iarchive <boost::archive::text_iarchive>::operator>><std::string> at interface_iarchive.hpp:61 #14 0x100024bc3 in boost ::archive ::detail ::interface_iarchive <boost::archive::text_iarchive>::operator&<std::string> at interface_iarchive.hpp:68 #15 0x100026688 in bus_stop_corner::serialize<boost::archive::text_iarchive> at main.cpp: 123 #16 0x1000266c3 in boost::serialization::access::serialize<boost::archive::text_iarchive, bus_stop_corner> at access.hpp:109 and so on... The last frame for which I have source code is #5. There the crashes occurs in the following routine: template<class Archive> BOOST_ARCHIVE_DECL(void) text_iarchive_impl<Archive>::load(std::string &s) { std::size_t size; * this->This() >> size; // skip separating space is.get(); // borland de-allocator fixup #if BOOST_WORKAROUND(_RWSTD_VER, BOOST_TESTED_AT(20101)) if(NULL != s.data()) #endif s.resize(size); // <============== crash here if(0 < size) is.read(&(*s.begin()), size); } In that frame, s is the empty string (its internal buffer points to a 0 byte), and size is 11. Suspecting a bug in OS X std::string class, I checked that the following program works fine: int main() { std::string s; std::size_t size = 11; std::cout << s << std::endl; s.resize(size); std::cout << s << std::endl; } So I tried to follow the stack up tracking the string object to see where and how it was created. I lost track of it in frame #15 on the following line: ar & street1 & street2; Note that this is back to the sample code. I decided to split that line into two lines, and surely now the crashes occurs there: ar & street1; I'm going to go on tracing through that, but maybe this rings a bell with someone, that would save me time... I can post more info if needed. Best regards and thanks, Jean-Denis

2 4

[interprocess] Memory deallocation doesn't seem to work
by Stephan Menzel 24 Sep '09

24 Sep '09

Hi there, I seem to get a strange behavior with 1.40 on windows concerning deallocation. Basically, I have a managed shared memory segment and I allocate several blocks of mem in it typedef boost::interprocess::managed_windows_shared_memory shared_segment; typedef boost::shared_ptr<shared_segment> shared_segment_ptr; ... shared_segment_ptr p = new shared_segment(create_only, buffername.c_str(), begin_size); ... Now I have several allocate_aligned operations: ptr = m_segment->allocate_aligned(static_cast<size_t>(rawsize), 128); Now after a while the segment will have no more room for further blocks. So I delete some of them to gain more room: m_segment->deallocate(raw.get()); (raw is an interprocess::offset_ptr) I debugged into this and there's indeed deallocation functionality called but when I try to allocate afterwards... ptr = m_segment->allocate_aligned(static_cast<size_t>(rawsize), 128); ... I get an interprocess::bad_alloc This makes me believe that either deallocate didn't do a good job or I have to do something after the deallocate() to really make it happen. I looked into the docs for this but there was nothing mentioned so I suppose deallocate() should suffice. I think fragmentation is not really a problem here. We're talking about a few >1MB blocks in a 30MB segment. Plus some administrative overhead. Any ideas what's going on here? Cheers, Stephan

2 3

Boost-regex: Weird behaviour with non-greedy matching operator in regex_replace in boost 1.40?
by Florian Schwarz 24 Sep '09

24 Sep '09

Hi, the non-greedy matching seems to have a weird behaviour (maybe non-determeterministic?) to work if the pattern is preceeded with something. E.g. when I want to get all characters in a string except the last on, if its an 'o', I would use regex matchExpr("(.*?)o?"); So if I write string text("hallo"); regex matchExpr("(.*?)o?"); string valueExpr("$1"); string result; regex_replace(back_inserter(resul), text.begin(), text.end(), matchExpr, valueExpr); cout << "Match \"" << result << "\"" << endl; it will print the expected "hall". If I now use instead string text(" hallo"); regex matchExpr(" (.*?)o?"); it will print "hallo". And for string text("hhallo"); regex matchExpr("h(.*?)o?"); string valueExpr("$1"); regex_replace(back_inserter(resul), text.begin(), text.end(), matchExpr, valueExpr); cout << "Match \"" << result << "\"" << endl; it will print "allo" which seems even more strange to me. I'm using boost1.40 on ubuntu with g++ 4.2.4 To give you a complete example: #include <boost/regex.hpp> #include <iostream> using namespace std; using namespace boost; void test(string& text, regex& matchExpr, string valueExpr){ if (regex_match(text, matchExpr)){ string result; regex_replace(back_inserter(result), text.begin(), text.end(), matchExpr, valueExpr); cerr << "Match \"" << result << "\"" << endl; } } int main(){ { // Test 1 string text("hallo"); regex matchExpr("(.*?)o?"); string valueExpr("$1"); test(text, matchExpr, valueExpr); } { // Test 2 string text(" hallo"); regex matchExpr(" (.*?)o?"); string valueExpr("$1"); test(text, matchExpr, valueExpr); } {// Test 3 string text(" hallo"); regex matchExpr("^ (.*?)o?$"); string valueExpr("$1"); test(text, matchExpr, valueExpr); } { // Test 4 string text("hhallo"); regex matchExpr("h(.*?)o?"); string valueExpr("$1"); test(text, matchExpr, valueExpr); } { // Test 5 string text("hhallo"); regex matchExpr("^h(.*?)o?$"); string valueExpr("$1"); test(text, matchExpr, valueExpr); } } The program will print Match "hall" Match "hallo" Match "hall" Match "allo" Match "hall" I have the following questions: - why does test 1 match the expected "hall" while test 2 matches "hallo" - why does test 1 match the whole string while test 4 matches only a part of it. Many thanks for your help and best regards Florian Schwarz

3 5

Object Pool Usage
by david.weber＠l-3com.com 24 Sep '09

24 Sep '09

So, I'm investigating Object Pools, and I think I'm not getting something. The documentation isn't great, hence the question We have a lot of small objects that we will use throughout our application's lifespan, and I'm trying to see if ObjectPool is a good fit for us. Heck, I'm just trying to look at performance of it in general. So, I have the following code: -------------------------------------------------------------- void testNormal() { common::Timer t; t.start(); for(int i = 0; i < MAX_NUM_TESTS; i++) { testClass t; t.doWork(); } t.stop(); std::cout << "No Pool: Creating " << MAX_NUM_TESTS << " took " << str_cast(t.getTotalElapsedTime()) << std::endl; } void testPool() { common::Timer t; t.start(); boost::object_pool<testClass> pool; for(int i = 0; i < MAX_NUM_TESTS; i++) { testClass::ptr tc = pool.construct(); tc->doWork(); } t.stop(); std::cout << "Pool: Creating " << MAX_NUM_TESTS << " took " << str_cast(t.getTotalElapsedTime()) << std::endl; } int main(int argc, char* argv[]) { testNormal(); testPool(); return 0; } ------------------------------------------------ When run, I am getting: No Pool: Creating 500000 took 00:00:00.187785 Pool: Creating 500000 took 00:00:00.281679 So, I guess I'm not understanding the advantage of the pools. Could someone write/modify my example to make it make more sense? Thanks --dw

4 3

Re: [Boost-users] [Regex] Group of number at line end
by John Maddock 24 Sep '09

24 Sep '09

> i am looking for a regular expression which extracts a group of number > from > the end of a line like the following: > > "This line may contain digits (9) but ends with numbers like 1 2 3 4 5" > > When using a regex like "^(.+)(\\s+(\\d+))+$" I receive > "This line may contain digits (9) but ends with numbers like 1 2 3 4" and > "5". > > Can someone help me? What about: "^(.+)[^\\d\\s](\\s+\\d+)+$" HTH, John.

2 1

boost::filesystem::exists question
by Steven White 23 Sep '09

23 Sep '09

I am working on building a project that relies on OpenSG2 and VR Juggler 2.2 for various reasons between the two I am building Boost 1_37 for the boost co dependency. While working with VR Juggler I noticed the error that boost::filesystem::exists was not defined in libboost_filesystem-xgcc40-mt.dylib. Which I found a little odd. After looking in to it I noticed that it is indeed not defined in either on of my release libs, but was defined in all of the debug libraries. In case I had done something wrong when I compiled boost I started a recompile with the following options. bjam --tool-set=darwin --prefix=$HOME/usr -- build-type=complete -a stage It currently looks like the build will yield the same results. I have a few nm | grep | c++filt below for the new libs. So I have two questions. Is the behavior below normal. If not what might be the cause? nm bin.v2/libs/filesystem/build/darwin-4.0.1/release/link-static/runtime-link-static/threading-multi/libboost_filesystem-xgcc40-mt-s-1_37.a | grep exists | c++filt 000003b0 T boost::filesystem::detail::symbolic_link_exists_api(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) 0002cc00 S __ZN5boost10filesystem6detail24symbolic_link_exists_apiERKSs.eh nm bin.v2/libs/filesystem/build/darwin-4.0.1/debug/link-static/runtime-link-static/threading-multi/libboost_filesystem-xgcc40-mt-sd-1_37.a | grep exists | c++filt 000005a4 T boost::filesystem::detail::symbolic_link_exists_api(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) 00025cf4 S __ZN5boost10filesystem6detail24symbolic_link_exists_apiERKSs.eh 00021fe8 S boost::files nm stage/lib/libboost_filesystem-xgcc40-mt-d-1_37.dylib |grep exists | c++filt 00002822 T boost::filesystem::detail::symbolic_link_exists_api(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) 00004872 T boost::filesystem::exists(boost::filesystem::file_status) volcano boost_1_37_0]$ nm stage/lib/libboost_filesystem-xgcc40-mt-1_37.dylib | grep exists | c++filt 00002010 T boost::filesystem::detail::symbolic_link_exists_api(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)

1 0

[enable_if] Doesn't work for static member functions ?
by Jean-Louis Leroy 23 Sep '09

23 Sep '09

Consider : template<typename U> struct Test { template<typename T> typename enable_if<is_same<T, int>, void>::type test() { cout << "int" << endl; } template<typename T> typename enable_if<is_same<T, char>, void>::type test() { cout << "char" << endl; } }; template<typename T> typename enable_if<is_same<T, int>, void>::type test() { cout << "int" << endl; } template<typename T> typename enable_if<is_same<T, char>, void>::type test() { cout << "char" << endl; } int _tmain(int argc, _TCHAR* argv[]) { test<int>(); test<char>(); Test<int>::test<double>(); // error return 0; } MSVC9 will give a compilation error (see below) where marked. Am I missing something obvious ? Thanks, J-L 1>c:\users\jll\documents\visual studio 2008\projects\boost_1_40_0\libs\rdb\test\test.cpp(325) : error C2143: syntax error : missing ';' before '<' 1> c:\users\jll\documents\visual studio 2008\projects\boost_1_40_0\libs\rdb\test\test.cpp(333) : see reference to class template instantiation 'Test<U>' being compiled 1>c:\users\jll\documents\visual studio 2008\projects\boost_1_40_0\libs\rdb\test\test.cpp(325) : error C4430: missing type specifier - int assumed. Note: C++ does not support default-int etc etc etc

2 1

[[value_initialized] when T is const
by Edward Diener 23 Sep '09

23 Sep '09

I like the idea of using value_initialized for a variable to initialize a value to its value intialized value when no initial value is given. But if an initial value is given it does not work if the T is a const type. Was this intended ? Essentially if a variable is a non-const object, one can say, for example: boost::value_initialized<int> data; // data initialized to its value initialized value boost::value_initialized<int> data; boost::get(data) = arg; // data is initialized to value arg For a const value we have: boost::value_initialized<int const> data; // data initialized to its value initialized value but we can not do: boost::value_initialized<int> data; boost::get(data) = arg; // ERROR // try to initialize data to value arg So there is no way to initialize a const T to a value when using boost::value_initialized<T>. This presents a problem in a template class where T can be a const type: template <class T> struct SomeTemplate<T> { boost::value_initialized<T> data; SomeTemplate() { } // always good SomeTemplate(T arg) { boost::get(data) = arg; // FAILS with a const T, // such as if SomeTemplate<int const> is instantiated with any int value // as in 'SomeTemplate<int const> st(4);' } }; In this case boost::value_initialized can not be used. This makes it much less usable in my estimation since the creator of a template class wanting to use boost::value_initialized can only do so if he restricts that template type to a non-const value. If he does not make that restriction then he must go back to initialization by hand and forgo the advantages of value_initialized entirely. It would seem that this limitation could be lifted if the boost::value_initialized template had a constructor which enabled one to initialize the value in cases where initialization to a value initialized state was not desired. Then, despite T being a const type, one could initialize the variable to a specific value.

2 11

[intrusive] list traversing by list_base_hooks
by Tim Blechmann 23 Sep '09

23 Sep '09

hi all, is there a way to traverse a boost::intrusive::list directly from the list_base_hook? i'd like to traverse the list directly from the node something like: my_node: public list_base_hook<> { my_node * previous_node(); my_node * next_node(); }; is it possible to implement something like that? thnx, tim -- tim(a)klingt.org http://tim.klingt.org /"\ ASCII Ribbon Campaign \ / no HTML in email & vCards X no proprietary attachments / \ use open standards

2 1

Calling boost::_mfi::dm on temporaries (e.g. in transform_iterator)
by Sebastian Theophil 23 Sep '09

23 Sep '09

I've encountered a problem when using boost::mem_fn in transform_iterators. In the attached example, I've chained two transform_iterators. The first iterator returns an Obj by value and the second extracts a member of the temporary object. The example shows that the return value is not quite what is expected. boost::mem_fn(&Obj::m_member) returns an int const&. The transform_iterator uses the return value of the UnaryFunction template parameter to initialize his value_type. In the given example, this logic seems too simple. At least if the UnaryFunction is of type boost::_mfi::dm the transform_iterator could do better: If Reference is boost::use_default If UnaryFunction is boost::_mfi::dm && Iterator::value_type is not a reference itself reference = boost::remove_reference( result_of(UnaryFunction) ) Else reference = result_of(UnaryFunction) Else reference = Reference Is this too much of a special case to fix it in the transform_iterator? The same logic would apply to all (unknown) functors that returned references to member variables. In general, the transform_iterator can probably not decide if a reference returned from a UnaryFunction will cease to exist when the temporary is destroyed. It could have been a global object. On the other hand, transform_iterators should work correctly if only boost objects are used, shouldn't they? I'd like to hear what you think, Sebastian #include "boost/iterator/transform_iterator.hpp" #include <vector> struct Obj { Obj(int n) : m_member(n) {} Obj(Obj const& obj) : m_member(obj.m_member) {} ~Obj() { m_member=INT_MIN; } int m_member; }; static Obj GetObjByValue( int n ) { return Obj(n); } void main() { std::vector<int> v(1, 5); int n=*boost::make_transform_iterator( boost::make_transform_iterator( v.begin(), GetObjByValue ), boost::mem_fn(&Obj::m_member) ); _ASSERT(n==INT_MIN); } -- Sebastian Theophil . stheophil(a)think-cell.com Software Engineer think-cell Software GmbH . Invalidenstr. 34 . 10115 Berlin, Germany http://www.think-cell.com . phone +49-30-666473-10 . toll-free (US) +1-800-891-8091 Directors: Dr. Markus Hannebauer, Dr. Arno Schoedl . Amtsgericht Berlin-Charlottenburg, HRB 85229

1 0