- Boost-users - lists.boost.org

Re: [Boost-users] Thread-safe smart pointers, hashtables without locking
by Ray Whitmer 08 May '06

08 May '06

> I'd also say the idea of moving a pointer out of the way, and then > using the > flip-gate-count-thing to wait for any 'potentially-using' readers > to finish, > definitely has 'potential'. You would need to show detailed code to > really > be sure. And not to us - to the guys on comp.programming.threads. Thanks. I will plan on pursuing it, then. > Anyhow, if you haven't already, take a look at memory barriers, > and, as a > canonical example, the broken 'double checked locking pattern'. You > can't > do lock free without understanding this stuff. I was somewhat aware of memory ordering and especially compiler reordering, but I need to take another look at it and find the proper set of macros for use in non-kernel code since volatile does not (yet) guarantee hardware behavior between processors. > And you are correct, you can't make { reading the > location of a shared_ptr + incrementing its ref-count } atomic, it > least not > on most processors. Actually, I was not concerned with incrementing the ref count. The flip-gate-count-thing solves, I believe, the problem of reference counts not being up to date all the time. I realize that I still have to prove this, but I do not think it is hard. What worries me more is whether the smart pointer has one or two fields in it, because this affects whether readers find an incomplete/ broken smart pointer as they whiz by. With a single pointer field in it, the load and store are atomic. > I don't think the boost implementation of intrusive_ptr's operator= > () is > atomic, but it could be made to be. shared_ptr has two pointers in it, (one to the shared count, one for the pointer) whereas intrusive_ptr has a single pointer in it (since the count is internal to the referenced class). This means that the shallow bits of shared_ptr are not atomically initialized, whereas those of intrusive_ptr are. An obvious way to solve this would be to make shared_ptr contain a single pointer to a wrapper class that contains both the shared count and a shared pointer, but this adds an unacceptable level of indirection to basic pointer operations. > Your wording of 'much less likely' is scary, and I think similar > wording > made James apprehensive as well. There is no grey - it is black or > white, > thread safe or not. I will have to adjust my language. I tend to use uncertain language both because I consider my knowledge incomplete and because of the lack of a generic standard that guarantees behavior, i.e. the fact that different Linux kernel architectures need different memory barriers, etc. is demonstration of the uncertainty of a general C solution. If I can rely on something like what the Linux kernel relies on, it should not be scary. Behavior needs to be guaranteed on supported platforms, and other platforms need to be detected and fail to compile lock-free or revert to a locking fallback implementation, etc. Ray Whitmer

2 1

Re: [Boost-users] [operators] loss of performance
by Victor A. Wagner Jr. 08 May '06

08 May '06

At 23:05 2006-05-07, you wrote: >Hi all - > >I just tried out the operators library for the first time today. I >was quite happy with how easily it helped me implement my test class. >Unfortunately, I'm running into some performance issues, in particular >with operator +. > >The interesting thing is that the new operator += is actually faster >than my old += by about a factor of 2, but my operator + is slower >than my old + by a factor of nearly 10! > >I'm new here, and not really sure how much code is appropriate to post >to the list, so here's a minimal rundown: > >-------------------------------------------------------------------------- >Old code: > >template<typename real> >Vector3<real> operator + (const Vector3<real> &a, const Vector3<real> &b); > >template<typename real> >class Vector3 { >protected: > real m_v[3]; >.. >.. > friend Vector3 operator +<> (const Vector3 &a, const Vector3 &b); >}; > >template<typename real> >inline Vector3<real> operator + (const Vector3<real> &a, const >Vector3<real> &b){ > Vector3<real> t; > t[0] = a[0] + b[0]; > t[1] = a[1] + b[1]; > t[2] = a[2] + b[2]; > return t; >} > >------------------------------------------------------------------------------------------------- >New code (note that the new code is templatized on size, and that my >instantiations >were of size 3 to match the above code): > >template<typename T, unsigned int S> >class vec > : boost::additive< vec<T, S> > , boost::multiplicative< vec<T, S>, T > > > >{ >private: > typedef unsigned int uint; > typedef const T &const_reference; > typedef T &reference; > >private: > boost::array<T,S> m_v; > > template<typename U> > vec &operator += (const vec<U,S> &t){ > for(uint i = 0; i < S; ++i){ > m_v[i] += t.m_v[i]; > } > return *this; > } >}; > >Any ideas how to increase the performance of the new code here? A >factor of 10 makes it seem like I am just missing something important. you showed us the code that you claim is faster, what do you want us to do? >Thanks, > Brian >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users Victor A. Wagner Jr. http://rudbek.com The five most dangerous words in the English language: "There oughta be a law"

2 1

[date_time] strange warning
by george 08 May '06

08 May '06

hello! try { boost::gregorian::date date_start(boost::gregorian::from_string(start)); boost::gregorian::date date_end(boost::gregorian::from_string(end)); boost::gregorian::date today(boost::gregorian::day_clock::local_day()); if(date_start > date_end || date_start < today || date_end.year() > today.year()) return false; period_=new boost::gregorian::date_period(date_start,date_end); return true; } catch(...) { return false; } .../boost/date_time/constrained_value.hpp: In member function ‘virtual bool Reservation::validation(std::string&)’: .../boost/date_time/constrained_value.hpp:64: warning: ‘year$D34041$value_’ is used uninitialized in this function .../date_time/constrained_value.hpp:64: warning: ‘year$D34041$value_’ is used uninitialized in this function why I get this warning? any idea? thanks

2 2

Re: [Boost-users] [multi_index] const mem_fun serialization bug.
by JOAQUIN LOPEZ MU?Z 08 May '06

08 May '06

----- Mensaje original ----- De: Julien Pervillé <julien.perville(a)free.fr> Fecha: Sábado, Mayo 6, 2006 9:15 pm Asunto: [Boost-users] [multi_index] const mem_fun serialization bug. > Hello boost-users. > > I am writing to report an annoying serialization bug encountered > while developing a project with Boost.multi_index, which is a > wonderfully useful data structure. Thank you Joaquín! > > First a fast overview of my project: [...] > After hunting through the headers, I located the triggered > exception to line 1080 of boost/multi_index/ordered_index.hpp > (line number from boost 1.33) > > Here is the context from the "rearranger" method where the > exception is thrown: > > > 1080 if(position!=x){ In Boost 1.33, tis is actually line 1088. Maybe a typo in your post? Anyway, I don't think this is important. [...] > I am looking forward to test patches and offer my help testing, > so that this little quirk gets solved. Hello Julien, thank you very much for using Boost.MultiIndex and for providing such a detailed bug report, with reproducible test case and all --actually, having the test case has made it easy to spot the problem. This is a bug in B.MI, but it has nothing to do with const_mem_fun. Rather it lies in the serialization loading algorithm. I'd like you to make the following changes in your local copy of B.MI (I'm assuming you've got 1.33 as per what you say in the post): 1. Replace line 80 in boost/multi_index/detail/duplicates_iterator.hpp: if(pred(begin_chunk->value,node->value))advance(); with the following: if(node!=end&&pred(begin_chunk->value,node->value))advance(); This bug was previously spotted by Toby Smith, and it's currently fixed in the CVS version of B.MI, soon to be shipped with Boost 1.34. More details at: http://lists.boost.org/boost-users/2006/01/16710.php 2. Replace the beginning of rearranger() at line 1076 of boost/multi_index/ordered_index.hpp: void rearranger(node_type* position,node_type *x) { node_type* before; if(!position){ ... with the following: void rearranger(node_type* position,node_type *x) { node_type* before; if(!position||comp(key(position->value),key(x->value))){ ... 3. Undo your previous hack which bypassed the archive_exception throw. That's it. I can confirm this makes seri-bug.cpp work OK. Could you please verify whether it also solves the problems in your real project? I'll devote some days to carefully study the issue and, if nothing weird shows, will commit the fix in the CVS HEAD and 1_33_4 branches next week. Please comment your results back. Thanks for reporting, best regards, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

3 2

a problem on boost::asio
by 李慧霸 08 May '06

08 May '06

I'm trying asio on WindowsXPSP2 + VS2003, and I found that there are always some connections established from 127.0.0.1:xxxx to 127.0.0.1:yyyy, is it a bug?

1 0

How to test whether a vertex or edge exists?
by sean yang 08 May '06

08 May '06

I want to add a vertex, which a string as its ID to a graph if a vertex representing such a string does not exist in the graph. Is there a function call to test if a vertex exists in a graph? ----------------------------------------------------------------------- //Here is the simpilfied situation that I want to handle, "IF THERE IS NOT a vertex representing the //string of the current line" is the predicate I want to know typedef adjacency_list<vecS, vecS, directedS, property<vertex_color_t, default_color_type>, property<edge_weight_t, int> > Graph; Graph g(); ifstream infile; infile.open(argv[1]); string line; while( getline(infile, line) ){ if(line.find("Node")!= npos && /*IF THERE IS NOT a vertex representing the string of the current line*/) add_vertex( line, g); } -------------------------------------------------------------------- A related question I want to ask is whether there exist a function call that test whether an edge representing [vertext string("Node1") --> vertext string("Node2") ] exists in a graph. Thanks a lot, Sean _________________________________________________________________ Express yourself instantly with MSN Messenger! Download today - it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

2 3

A Problem with boost::bind
by Hossein Haeri 07 May '06

07 May '06

Dear all, A fragment of code of mine is like this: template <class Container, size_t i> struct print { ostream& operator () (const Container& c, ostream& out) const { if(out) { typedef typename Container::value_type value_type; typedef print<value_type, i - 1> NewPrint; NewPrint newPrint; for(typename Container::const_iterator j = c.begin(); j != c.end(); ++j) newPrint(*j, out); /*for_each(c.begin(), c.end(), bind(newPrint, _2, out)); //Error!*/ } return out; } }; AFAICS, there should be no harm in replacing that for with the for_each in the comments -- the one is marked with "//Error" I mean, except that "out" which is of type ostream& is not Copy-Constructible, and therefore _2 cannot be used against it. Anybody having a workarounds for that? Doesn't Boost have anything analog to boost::ref which can give me a Copy-Constructible beast? For your reference, you can find the whole program and the compiler error message in the P.S. TIA, --Hossein P.S. #include <iostream> #include <algorithm> #include <iterator> #include <vector> #include <boost/bind.hpp> #include <boost/mem_fn.hpp> using std::ostream; using std::cin; using std::cout; using std::copy; using std::for_each; using std::ostream_iterator; using std::vector; using boost::bind; using boost::mem_fn; template <class Container, size_t i> struct print { ostream& operator () (const Container& c, ostream& out) const { if(out) { typedef typename Container::value_type value_type; typedef print<value_type, i - 1> NewPrint; NewPrint newPrint; /* for(typename Container::const_iterator j = c.begin(); j != c.end(); ++j) newPrint(*j, out); */ for_each(c.begin(), c.end(), bind(newPrint, _2, out)); //Error! } return out; } }; template <class Container> struct print<Container, 1> { ostream& operator () (const Container& c, ostream& out) const { if(out) copy(c.begin(), c.end(), ostream_iterator<typename Container::value_type>(out, " ")); return out; } ostream& DoIt (const Container& c, ostream& out) const { if(out) copy(c.begin(), c.end(), ostream_iterator<typename Container::value_type>(out, " ")); return out; } }; int main() { int a1[] = {1, 2, 3, 4}; const size_t n(sizeof(a1)/sizeof(*a1)); vector<int> v(a1, a1 + n); print<vector<int>, 1> () (v, cout); int a2[n] = {5, 6, 7, 8}; vector<vector<int> > m; m.push_back(v); m.push_back(vector<int>(a2, a2 + n)); print<vector<vector<int> >, 2> () (m, cout); cin.get(); return 0; } Compiler: Default compiler Executing g++.exe... g++.exe "E:\Print.cpp" -o "E:\Print.exe" -I"C:\Dev-Cpp\include\c++" -I"C:\Dev-Cpp\include\c++\mingw32" -I"C:\Dev-Cpp\include\c++\backward" -I"C:\Dev-Cpp\include" -L"C:\Dev-Cpp\lib" C:/Dev-Cpp/include/c++/3.3.1/bits/ios_base.h: In copy constructor ` std::basic_ios<char, std::char_traits<char> >::basic_ios(const std::basic_ios<char, std::char_traits<char> >&)': E:/Print.cpp:33: instantiated from `std::ostream& print<Container, i>::operator()(const Container&, std::ostream&) const [with Container = std::vector<std::vector<int, std::allocator<int> >, std::allocator<std::vector<int, std::allocator<int> > > >, unsigned int i = 2]' E:/Print.cpp:72: instantiated from here C:/Dev-Cpp/include/c++/3.3.1/bits/ios_base.h:668: error: ` std::ios_base::ios_base(const std::ios_base&)' is private E:/Print.cpp:33: error: within this context E:/Print.cpp: In member function `std::ostream& print<Container, i>::operator()(const Container&, std::ostream&) const [with Container = std::vector<std::vector<int, std::allocator<int> >, std::allocator<std::vector<int, std::allocator<int> > > >, unsigned int i = 2]': E:/Print.cpp:72: instantiated from here E:/Print.cpp:33: error: initializing argument 3 of ` boost::_bi::bind_t<boost::_bi::unspecified, F, typename boost::_bi::list_av_2<A1, A2>::type> boost::bind(F, A1, A2) [with F = print<std::vector<int, std::allocator<int> >, 1>, A1 = boost::arg<2>, A2 = std::basic_ostream<char, std::char_traits<char> >]' C:/Dev-Cpp/include/boost/bind.hpp: At global scope: C:/Dev-Cpp/include/boost/bind.hpp: In instantiation of `boost::_bi::result_traits<boost::_bi::unspecified, print<std::vector<int, std::allocator<int> >, 1> >': C:/Dev-Cpp/include/boost/bind/bind_template.hpp:16: instantiated from `boost::_bi::bind_t<boost::_bi::unspecified, print<std::vector<int, std::allocator<int> >, 1>, boost::_bi::list2<boost::arg<2>, boost::_bi::value<std::basic_ostream<char, std::char_traits<char> > > > >' E:/Print.cpp:33: instantiated from `std::ostream& print<Container, i>::operator()(const Container&, std::ostream&) const [with Container = std::vector<std::vector<int, std::allocator<int> >, std::allocator<std::vector<int, std::allocator<int> > > >, unsigned int i = 2]' E:/Print.cpp:72: instantiated from here C:/Dev-Cpp/include/boost/bind.hpp:62: error: no type named `result_type' in ` struct print<std::vector<int, std::allocator<int> >, 1>' C:/Dev-Cpp/include/boost/bind/bind_template.hpp: In instantiation of `boost::_bi::bind_t<boost::_bi::unspecified, print<std::vector<int, std::allocator<int> >, 1>, boost::_bi::list2<boost::arg<2>, boost::_bi::value<std::basic_ostream<char, std::char_traits<char> > > > >': E:/Print.cpp:33: instantiated from `std::ostream& print<Container, i>::operator()(const Container&, std::ostream&) const [with Container = std::vector<std::vector<int, std::allocator<int> >, std::allocator<std::vector<int, std::allocator<int> > > >, unsigned int i = 2]' E:/Print.cpp:72: instantiated from here C:/Dev-Cpp/include/boost/bind/bind_template.hpp:16: error: no type named `type' in `struct boost::_bi::result_traits<boost::_bi::unspecified, print<std::vector<int, std::allocator<int> >, 1> >' C:/Dev-Cpp/include/boost/bind/bind_template.hpp:19: error: no type named `type' in `struct boost::_bi::result_traits<boost::_bi::unspecified, print<std::vector<int, std::allocator<int> >, 1> >' C:/Dev-Cpp/include/boost/bind/bind_template.hpp:25: error: no type named `type' in `struct boost::_bi::result_traits<boost::_bi::unspecified, print<std::vector<int, std::allocator<int> >, 1> >' C:/Dev-Cpp/include/boost/bind/bind_template.hpp:31: error: no type named `type' in `struct boost::_bi::result_traits<boost::_bi::unspecified, print<std::vector<int, std::allocator<int> >, 1> >' C:/Dev-Cpp/include/boost/bind/bind_template.hpp:31: confused by earlier errors, bailing out Execution terminated ___________________________________________________________ Switch an email account to Yahoo! Mail, you could win FIFA World Cup tickets. http://uk.mail.yahoo.com

4 10

Re: [Boost-users] [MULTI_INDEX] intersection algorithm?
by JOAQUIN LOPEZ MU?Z 07 May '06

07 May '06

Guillaume Lazzara ha escrito: > Hello, > > I am currently using multi_index as "set" container > and I wanted to know how to use intersection algorithm. > I tried with the STL interction algorithm but it > crashed. How can I do? Hello Guillaume. In principle, there's nothing preventing multi_index_container's from working with set_intersection. Could you provide a case where the referred to crash shows? Failing that, a little more detailed explanation or some code? Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

[SOC] Applicants -- Please Monitor Google for Comments
by Jeff Garland 07 May '06

07 May '06

Hi All - First let me say that we've had a tremendous response thus far to SOC. Some really excellent applications and ideas. I have high hopes that some very exciting Boost projects will come out of SOC. The reason I'm sending this is that I'm trying to reach out to applicants that may not be aware that proposal feedback is being provided on the SOC website. Please check your proposal for comments. Many of the proposals need additional details. There is no automated email when we add comments to a proposal, so you won't be notified. Goggle is looking into adding this, but it isn't there now. There are enough comments and proposals that it would be impractical for us to ensure that everyone got notified when a comment is added. Thanks, Jeff

1 0

Re: [Boost-users] [shmem] New bugfix version in Boost Vault andsandbox cvs
by JOAQUIN LOPEZ MU?Z 06 May '06

06 May '06

----- Mensaje original ----- De: Ion Gaztañaga <igaztanaga(a)gmail.com> Fecha: Sábado, Mayo 6, 2006 9:51 pm Asunto: Re: [Boost-users] [shmem] New bugfix version in Boost Vault andsandbox cvs > > Or, depending on > > how shmem fits into the picture, should B.MI take the lead in > > supporting shmem functionality internally? Y'all have more > > flexibility in the B.MI case as it isn't part of the standard. > > To be able to support Multi-index in shared memory we would need > to > rewrite it to replace all raw pointers with allocator::pointer > typedefs. > I haven't looked through multiindex code but this can be a big > task. > There is nothing that shmem can do to support multiindex if it > uses raw > pointers. But I think that multiindex and shmem can be a good base > for > in-memory data-bases. > I wholeheartedly agree with you shmem and multi_index should be able to interoperate nicely. My other activities in B.MI, and my rather shallow knowleddge of shmem, prevent me from attacking this task right away, but I'd be happy if some bold Booster could do some analysis/prototype of the changes involved in B.MI for it to work with shmem, so that eventually I can integrate those changes. If someone is up to the task, do count on my support/assitance for matters concerning the inner structure of B.MI. Best regards, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

1 0