- Boost-users - lists.boost.org

interprocess::list and destructors of stored objects
by alexmark 02 Dec '06

02 Dec '06

Hi to all! For the container type boost::interprocess::list<T, A> I need the destructor of stored object T be called when erase() is called for a corresponding list item. By default the T destructor is not called, memory only is deallocated. What is the correct way to force boost::interprocess::list to invoke destructors for contained objects while list item or the container itself is deleted? TYA.

3 2

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 01 Dec '06

01 Dec '06

________________________________________ From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Gottlob Frege Sent: Friday, December 01, 2006 3:25 PM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > On 12/1/06, Nat Goodspeed <ngoodspeed at solidworks dot com> wrote: > > It seems possible to implement weak_ptr notification by building a list > > of weak_ptr instances referencing a given object. If we don't want that > > list to consume additional heap memory, the list could itself be > > intrusive in the weak_ptr objects. If we want it to be efficient, we > > build a doubly-linked list. (This may call for a policy-based > > implementation so the consumer can decide which overhead is least > > noxious.) > > Interestingly, since you define the intrusive_ptr_add_ref and > release, you can probably implement the list management without > changing intrusive_ptr. [Nat] That's true, you could. But that would impose the requirement on everyone who wanted to use intrusive_ptr with this new variety of weak_ptr. Perhaps intrusive_ptr could do something implicit before calling intrusive_ptr_release. Maybe that could be implemented as an intrusive_ptr policy. > Would you prefer to have the list of weak_ptrs kept on the > intrusive_ptr (on the actual 'intruded' object, actually), or > have a global map of intrusive_ptr object to weak_ptr list? [Nat] I was thinking of storing the head pointer in the object itself. But it might be more consistent with the present intrusive_ptr implementation to call an intrusive_ptr_get_first_weak_ptr function, requiring the consumer to decide where to put it.

1 0

Re: [Boost-users] Boost::wave, sample driver and how to use it
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Sebastian Redl > Sent: Friday, December 01, 2006 3:37 PM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] Boost::wave, sample driver and how to use it > > mosfet wrote: > > VOID foo(BYTE bTest, CHAR szTest); > > ULONG foo(LPBYTE bTest, CHAR szTest); > > > If this is Windows, those are typedefs, not macros. The preprocessor > does not affect them. [Nat] <off-topic> Might gccxml be a better answer for the OP? Caveat: AFAIK, gccxml doesn't deal well with MFC headers and doesn't yet have its own equivalents. </off-topic>

1 0

Re: [Boost-users] Boost::wave, sample driver and how to use it
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of mosfet > Sent: Thursday, November 30, 2006 3:07 PM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Boost::wave, sample driver and how to use it > > foo.h > //-----// > > VOID foo(BYTE bTest, CHAR szTest); > ULONG foo(LPBYTE bTest, CHAR szTest); > > I would like to get > > void foo(unsigned char bTest, char szTest); > unsigned long foo(BYTE * bTest, char szTest); > > but maybe I haven't understood what a preprocessor was. > If someone could help, it's really urgent. [Nat] You mentioned Windows. Does that mean you're using Microsoft C++? Wouldn't it work to use something like cl /E ... or cl /P ... ?

1 0

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Ovanes Markarian > Sent: Friday, December 01, 2006 2:23 PM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > > Just a small question: > > Isn't it so, that linked lists (double or single), can be spreaded in the > memory, so that cache > misses and page faults are subject to happen? I think deque is a better > choice (which can be > better optimized in terms of locality)... [Nat] My suggestion was based on the (implied) intrusive_ptr constraint of no additional heap resources. But if we were to use a policy-based approach, a deque implementation could be one of the policies.

1 0

Problem building Boost on Windows
by Ronald Levine 01 Dec '06

01 Dec '06

C:\kits\boost_1_33_1>bjam "-sTOOLS=vc-8_0" install ****************************************************** Building Boost.Iostreams with bzip2 support disabled. To enable bzip2, consult the Boost.Iostreams documentation ****************************************************** ****************************************************** Building Boost.Iostreams with zlib and gzip support disabled. To enable zlib and gzip, consult the Boost.Iostreams documentation ****************************************************** Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (and if you don't know what ICU is then you probably don't need it). ...patience... ...patience... ...found 13371 targets... ...updating 4941 targets... MkDir1 C:\Boost spawn: Invalid argument

1 0

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Steven T. Hatton > Sent: Friday, December 01, 2006 7:03 AM > To: Boost-users(a)lists.boost.org > Subject: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > > Intrusive pointers are, IMO, the correct mechanism to hold a > tree together. They are efficient, effective, and easy to > understand. They work better than shared pointers for > recursive data structures. Unfortunately the lack of a > corresponding weak_ptr to go with boost::intrusive_ptr means I > either use a raw pointer to the parent, devise a strategy for > decrementing the pointers myself, or create my own weak > pointer. [Nat] I've thought about this problem to some extent. This is the way I see it so far. The one big advantage of a weak_ptr over a raw pointer is that the weak_ptr can tell you when its referent becomes invalid. If your data structure requires storing a pointer in a place which itself becomes invalid as soon as the referent becomes invalid (e.g. recursive deletion of a subtree), then I'm not convinced that weak_ptr is significantly better than a raw pointer. (Even "snipping loose" a subtree, but preserving it, doesn't seem to argue for weak_ptr as a parent pointer. In such a case, you'd want to clear the parent pointer anyway, wouldn't you?) Does this rationale make sense, or am I overlooking something? I'm not rejecting the idea of a weak_ptr that works with intrusive_ptr. I'm just saying that for certain use cases, raw pointers seem to address the requirements equally well -- with less runtime overhead. This case looks like one of them.

2 1

serialization problem
by nico35＠noos.fr 01 Dec '06

01 Dec '06

Hello, I am using boost1.33.1 and spirit1.6 and borland c++ builder. I have a compilation problem with the following source code. The error message is [C++ Error] void_cast.hpp(154): E2451 Undefined symbol 'get_instance' [C++ Error] void_cast.hpp(155): E2451 Undefined symbol 'get_instance' [C++ Error] void_cast.hpp(154): E2285 Could not find a match for 'void_caster::void_caster(undefined,undefined)' Can anybody help me fixing this compilation problem ? Than you in advance. Marc. //--------------------------------------------------------------------------- #include <vcl.h> #pragma hdrstop #include <boost/serialization/utility.hpp> // includes for xml serialization #include <boost/archive/xml_oarchive.hpp> #include <boost/archive/xml_iarchive.hpp> #include <boost/archive/basic_xml_oarchive.hpp> #include <boost/archive/basic_xml_iarchive.hpp> #include <boost/archive/impl/xml_iarchive_impl.ipp> #include <boost/archive/impl/xml_oarchive_impl.ipp> #include <boost/archive/impl/basic_xml_oarchive.ipp> #include <boost/archive/impl/basic_xml_iarchive.ipp> #include <boost/archive/basic_text_oprimitive.hpp> #include <boost/archive/basic_text_iprimitive.hpp> #include <boost/archive/impl/basic_text_oprimitive.ipp> #include <boost/archive/impl/basic_text_iprimitive.ipp> #include <boost/archive/impl/basic_xml_grammar.hpp> #include <libs/serialization/src/basic_xml_grammar.ipp> #include <boost/serialization/nvp.hpp> #include <boost/serialization/base_object.hpp> #include <boost/serialization/export.hpp> #include <boost/serialization/type_info_implementation.hpp> #include <boost/serialization/extended_type_info_no_rtti.hpp> #include <boost/serialization/extended_type_info_typeid.hpp> #include<iostream> #include<fstream> #include <vector> #include <boost/serialization/vector.hpp> class BaseClass { protected: friend class boost::serialization::access; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & BOOST_SERIALIZATION_NVP(m_degrees); ar & BOOST_SERIALIZATION_NVP(m_minutes); ar & BOOST_SERIALIZATION_NVP(m_seconds); } int m_degrees; int m_minutes; float m_seconds; public: BaseClass() { m_degrees = 0; m_minutes = 0; m_seconds = 0.0; } BaseClass(int degrees, int minutes, float seconds) { m_degrees = degrees; m_minutes = minutes; m_seconds = seconds; } virtual void display() const { std::cout << "------------------------------------" << std::endl; std::cout << "m_degrees= " << m_degrees << std::endl; std::cout << "m_minutes= " << m_minutes << std::endl; std::cout << "m_seconds= " << m_seconds << std::endl; std::cout << "------------------------------------" << std::endl; } }; BOOST_CLASS_TYPE_INFO( BaseClass, extended_type_info_no_rtti<BaseClass> ) BOOST_CLASS_EXPORT(BaseClass) class DerivedClass1: public BaseClass { protected: friend class boost::serialization::access; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & BOOST_SERIALIZATION_BASE_OBJECT_NVP(BaseClass); // ar & BOOST_SERIALIZATION_NVP(m_vector); } mutable std::vector<int> m_vector; public: DerivedClass1(): BaseClass() { } DerivedClass1(int degrees, int minutes, float seconds): BaseClass(degrees, minutes, seconds) { } void push_back(int arg) const { m_vector.push_back(arg); } virtual void display() const { std::cout << "***********************************************" << std::endl; BaseClass::display(); for (std::vector<int>::const_iterator iter=m_vector.begin(); iter!=m_vector.end(); iter++) { std::cout << " value= " << *iter << std::endl; } std::cout << "***********************************************" << std::endl; } }; BOOST_CLASS_TYPE_INFO( DerivedClass1, extended_type_info_no_rtti<DerivedClass1> ) BOOST_CLASS_EXPORT(DerivedClass1) BOOST_CLASS_TYPE_INFO( std::vector<double>, extended_type_info_no_rtti<std::vector<double> > ) BOOST_CLASS_EXPORT(std::vector<double>) int main() { const DerivedClass1 g(35, 59, 24.567f); g.push_back(10); g.push_back(20); g.push_back(30); g.display(); { std::ofstream ofs("filename"); boost::archive::xml_oarchive oa(ofs); oa << boost::serialization::make_nvp("g1", g); } DerivedClass1 newg; { std::ifstream ifs("filename"); boost::archive::xml_iarchive ia(ifs); ia >> boost::serialization::make_nvp("g1", newg); } newg.display(); return 0; } //--------------------------------------------------------------------------- /* ------------------------------------------------------------- Noos, votre bureau virtuel sur Internet : Mail... Web : webmail.noos.fr

2 1

Re: [Boost-users] [bind, function, mem_fn] Connection to a base class
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Johannes Brunen > Sent: Friday, December 01, 2006 7:10 AM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] [bind,function,mem_fn] Connection to a base class [Nat] This is top-of-the-head, no tests. > I have a base class X with two functions f and g > > class X { > virtual bool f() const; > virtual bool g(int) const; > }; > > and another class Y with functions f1, f2 and F > > class Y { > void f1(); > void f2(); > void F(???); [Nat] Seems like you'd want to make F accept boost::function<bool()>, and call it without arguments. > }; > > Y::f1() { > // call F so that F works on X::f > F(??? X::f); [Nat] F(boost::bind(&X::f, this)) > } > > Y::f2() { > // call F so that F works on Y::g but with g's int parameter known > and determined in Y::f2 > int v = 5; > F(??? X::g, v); [Nat] F(boost::bind(&X::g, this, v)) > }

1 0

how to clear a Graph
by Ufuk Okuyucu 01 Dec '06

01 Dec '06

Hi, I've just been handed a code using boost's cuthill-mckee algorithm. I found a memory leak within the code block where a number of edges being added to a graph but never cleaned at the end. Now, graph has a "clear" method where it's supposed to delete vertices and edges but it does not effect the memory leak, when I check the used memory size its still there. I need a way to delete a graph object and free all its internal mallocations if any. ...and I haven't used boost before... Here is the code block: //create incidence graph G2 from adj_list //apply cuthill-mckee on G2 //write out resulting permutation to perm_result Graph G2(matrix_size); //Graph data structure used in Cuthill-Mckee //add all the edges in adj_list to G2 for(int row=0;row<matrix_size;row++) { for(ROW_ADJ::iterator col=adj_list[row].begin(); col != adj_list[row].end(); col++) { add_edge (row, *col, G2); //<<-----this is where it gets all that memory } } cout <<"edges added\n"; print_system_stats(); graph_traits<Graph>::vertex_iterator ui, ui_end; property_map<Graph,vertex_degree_t>::type deg = get(vertex_degree, G2); for (boost::tie(ui, ui_end) = vertices(G2); ui != ui_end; ++ui) deg[*ui] = degree(*ui, G2); property_map<Graph, vertex_index_t>::type index_map = get(vertex_index, G2); //printf("original bandwidth: %d\n",bandwidth(G2)); printf("# vertices: %d\n",num_vertices(G2)); printf("# edges: %d\n",num_edges(G2)); vector<Vertex> inv_perm(num_vertices(G2)); vector<size_type> perm(num_vertices(G2)); Vertex start_vertex = vertex(0,G2); cuthill_mckee_ordering(G2, start_vertex, inv_perm.rbegin(), get(vertex_color, G2), get(vertex_degree, G2)); for (size_type c = 0; c != inv_perm.size(); ++c) perm[index_map[inv_perm[c]]] = c; //cout << " bandwidth: " << bandwidth(G2, make_iterator_property_map(&perm[0], index_map, perm[0])) << endl; //perm has the new ordering, copy it to perm_result for (int i=0;i<matrix_size;i++) perm_result[i] = perm[i]; G2.clear(); //<<-------this statement does not release all the memory inv_perm.clear(); perm.clear(); -- * Ufuk Okuyucu** *

1 0