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

"smart list" container?
by David M. Jones 12 Oct '04

12 Oct '04

Hello, I have run into a situation where the following std::list-like container would be useful. - The container itself would act more or less identically to std::list; - The iterators on the container would act like boost::weak_ptr in that they would know that they are invalid when the list element that they point to is removed. Does anyone know of such a container in any of the boost libraries, in another library or a way to accomplish the task using other means (such as iterator adaptors)? I looked at the newly proposed "smart containers" library, but it seems to be solving a different problem despite the similarity with the name I am using to describe what I want. Thanks in advance, David M. Jones

2 1

two boost.threads taking much longer than one
by Andrew Chapman 11 Oct '04

11 Oct '04

Hi everyone, I've got an operation which takes 14 seconds on a single thread, but is taking up to 45 seconds when running via two boost.threads on a dual proc machine (2x 3ghz Xeons, Fedora Core 1). The thread functor objects are being passed a lot of references (about 20) to some very large std::vectors. One thread always reads/writes from the start of the vectors to their midpoint, and the second thread from (midpoint+1) to the end of the vectors. The only thing I can think of is that the threads are taking a long time to fork and get running, perhaps moving a lot of stack data around or something? When I look at the output of 'top' with threads displayed, I see the two threads appear, but they only use about 50-60% CPU utilization each. The system monitor shows mostly red (Kernel) as opposed to blue (User). Any ideas? -- Andrew Chapman Senior Technical Director - Framestore CFC

3 3

Re: Re: [Boost-users] function_traits inheritance
by Victor Morilla 11 Oct '04

11 Oct '04

Peter Dimov wrote: >Victor Morilla wrote: >> >> template <typename FnType, FnType* fn> >> struct Translator<FnType, fn, 0> : >> function_traits<FnType> >> { >> typedef result_type type; > typedef typename function_traits<FnType>::result_type type; >> }; The question is, why result_type type definition is not inherited from function_traits ? ______________________________________________ Renovamos el Correo Yahoo!: ¡100 MB GRATIS! Nuevos servicios, más seguridad http://correo.yahoo.es

2 1

function_traits inheritance
by Victor Morilla 11 Oct '04

11 Oct '04

Guess someone why this code fails ? (g++3.3) ---- begin ----------------------- #include <boost/type_traits.hpp> #include <iostream> using namespace std; using namespace boost; int foo() { return 10; } template <typename FnType, FnType* fn, int Arity = function_traits<FnType>::arity> struct Translator; template <typename FnType, FnType* fn> struct Translator<FnType, fn, 0> : function_traits<FnType> { typedef result_type type; }; int main(int argc, char** argv) { Translator<int (), &foo>::type x; } ------------------- end code ------------------ error: ISO C++ forbids declaration of `result_type' with no type --------------------- end error ------------------- And another question: is there a portable equivalent to: Translator<typeof(foo), &foo> Thanks ______________________________________________ Renovamos el Correo Yahoo!: ¡100 MB GRATIS! Nuevos servicios, más seguridad http://correo.yahoo.es

3 2

Problems in compilation with Boost
by Debayan Chaudhuri 11 Oct '04

11 Oct '04

Hi, I am evaluating Boost for use in our project. The platform OS is FreeBSD 4.9; which comes with gcc-2.95.3. I downloaded boost_1.30.0 and compiled the thread library on the FreeBSD box. When I try to compile a small test program using Boost threads; I encounter hundreds of compiler errors; mostly in the Boost header files. Is there a workaround to this ? Which version of Boost works with gcc-2.95.3 and what are the compiler options that will make it compile properly? regards, Debayan.

2 2

RE: [Boost-users] Problems with compiling Filesystem library on Solaris
by Barak Simon 11 Oct '04

11 Oct '04

By now I had time to check your suggestion, but I'm afraid I still get the same error when building boost libs: "/db/pub/eq/build/src/boost_1_31_0/boost/iterator/iterator_facade.hpp", line 622: Error: Could not find a match for boost::iterator_core_access::equal<boost::iterator_core_access::Facade1, boost::iterator_core_access::Facade2>(const boost::filesystem::directory_iterator, const boost::filesystem::directory_iterator, boost::is_convertible<boost::filesystem::directory_iterator, boost::filesystem::directory_iterator>). "/db/pub/eq/build/src/boost_1_31_0/libs/filesystem/build/../src/operations_posix_windows.cpp", line 176: Where: While instantiating "boost::operator==<boost::filesystem::directory_iterator, boost::filesystem::path, boost::single_pass_traversal_tag, boost::filesystem::path&, int, boost::filesystem::directory_iterator, boost::filesystem::path, boost::single_pass_traversal_tag, boost::filesystem::path&, int>(const boost::iterator_facade<boost::filesystem::directory_iterator, boost::filesystem::path, boost::single_pass_traversal_tag, boost::filesystem::path&, int>&, const boost::iterator_facade<boost::filesystem::directory_iterator, boost::filesystem::path, boost::single_pass_traversal_tag, boost::filesystem::path&, int>&)". "/db/pub/eq/build/src/boost_1_31_0/libs/filesystem/build/../src/operations_posix_windows.cpp", line 176: Where: Instantiated from non-template code. Are you sure that is the only line I need to comment out? Thanks, Barak ________________________ Barak Simon GED IT Core Platform 190 George St. Sydney, 2000 Phone +61 2 925 85070 Fax +61 2 925 95050 barak.simon+external@db. com To: boost-users(a)lists.boost.org Sent by: cc: boost-users-bounces@list Subject: RE: [Boost-users] Problems with compiling Filesystem library on Solaris s.boost.org 11/10/2004 11:45 Please respond to boost-users Thanks for the swift reply! In the interim I commented out the 2-3 problematic methods (which I am not using anyway) in the filesystem lib, but I like your solution better. A question on that though: On your solution, what exactly am I commenting out, that is, can the workaround break some other boost library or a significant assumption? Thanks, Barak Tomas.Puverle@morganstan ley.com To: boost-users(a)lists.boost.org Sent by: cc: boost-users-bounces@list Subject: RE: [Boost-users] Problems with compiling Filesystem library on Solaris s.boost.org 08/10/2004 18:47 Please respond to boost-users > Hi, > I get the following error from the Filesystem library when > building boost 1_31 on Solaris with sunpro compiler: > > "/db/pub/eq/build/src/boost_1_31_0/boost/iterator/iterator_fac > ade.hpp", line 622: Error: Could not find a match for > boost::iterator_core_access::equal<boost::iterator_core_access > ::Facade1, boost::iterator_core_access::Facade2>(const > boost::filesystem::directory_iterator, const > boost::filesystem::directory_iterator, > boost::is_convertible<boost::filesystem::directory_iterator, > boost::filesystem::directory_iterator>). > > "/db/pub/eq/build/src/boost_1_31_0/libs/filesystem/build/../sr c/operations_posix_windows.cpp", line 176: Where: While instantiating > "boost::operator==<boost::filesystem::directory_iterator, > boost::filesystem::path, boost::single_pass_traversal_tag, > boost::filesystem::path&, int, > boost::filesystem::directory_iterator, > boost::filesystem::path, boost::single_pass_traversal_tag, > boost::filesystem::path&, int>(const > boost::iterator_facade<boost::filesystem::directory_iterator, > boost::filesystem::path, boost::single_pass_traversal_tag, > boost::filesystem::path&, int>&, const > boost::iterator_facade<boost::filesystem::directory_iterator, > boost::filesystem::path, boost::single_pass_traversal_tag, > boost::filesystem::path&, int>&)". > > "/db/pub/eq/build/src/boost_1_31_0/libs/filesystem/build/../sr c/operations_posix_windows.cpp", line 176: Where: Instantiated from non-template code. > > > > It looks suspiciously similar to bug "[ 976241 ] Compiler > error for tokenizer on Solaris" from 2004-06-20, with the > difference that I am using version 5.3 of sunpro rather than 5.5. I have submitted a similar problem and the proposed fix recently. I am just waiting on our developers to confirm that it works before submitting it to boost, but it seems to fix the tokenizer problem, too. For a quick workaround (before the patch is applied) do this: At the top of boost/type_traits/detail/bool_trait_def.hpp you will see something like this: #if defined(__SUNPRO_CC) # define BOOST_TT_AUX_BOOL_TRAIT_VALUE_DECL(C) \ typedef mpl::bool_< C > type; \ enum { value = type::value }; \ /**/ # define BOOST_TT_AUX_BOOL_C_BASE(C) //1 Comment out the line marked with //1. That seems to have fixed my problems with both 5.3 and 5.5. HTH, Tom -------------------------------------------------------- NOTICE: If received in error, please destroy and notify sender. Sender does not waive confidentiality or privilege, and use is prohibited. _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users -- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden. _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users -- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

1 0

Function type transformation and manipulation traits?
by Aaron W. LaFramboise 11 Oct '04

11 Oct '04

Are there any Boost support for manipulating function types? Operations I have in mind include: setting return type, setting argument type, changing number of arguments, et cetera. If not, is there some generally accepted way to accomplish this, and is any work being done to add this feature to Boost? Aaron W. LaFramboise

2 3

RE: [Boost-users] Problems with compiling Filesystem library on Solaris
by Barak Simon 11 Oct '04

11 Oct '04

Thanks for the swift reply! In the interim I commented out the 2-3 problematic methods (which I am not using anyway) in the filesystem lib, but I like your solution better. A question on that though: On your solution, what exactly am I commenting out, that is, can the workaround break some other boost library or a significant assumption? Thanks, Barak Tomas.Puverle@morganstan ley.com To: boost-users(a)lists.boost.org Sent by: cc: boost-users-bounces@list Subject: RE: [Boost-users] Problems with compiling Filesystem library on Solaris s.boost.org 08/10/2004 18:47 Please respond to boost-users > Hi, > I get the following error from the Filesystem library when > building boost 1_31 on Solaris with sunpro compiler: > > "/db/pub/eq/build/src/boost_1_31_0/boost/iterator/iterator_fac > ade.hpp", line 622: Error: Could not find a match for > boost::iterator_core_access::equal<boost::iterator_core_access > ::Facade1, boost::iterator_core_access::Facade2>(const > boost::filesystem::directory_iterator, const > boost::filesystem::directory_iterator, > boost::is_convertible<boost::filesystem::directory_iterator, > boost::filesystem::directory_iterator>). > > "/db/pub/eq/build/src/boost_1_31_0/libs/filesystem/build/../sr c/operations_posix_windows.cpp", line 176: Where: While instantiating > "boost::operator==<boost::filesystem::directory_iterator, > boost::filesystem::path, boost::single_pass_traversal_tag, > boost::filesystem::path&, int, > boost::filesystem::directory_iterator, > boost::filesystem::path, boost::single_pass_traversal_tag, > boost::filesystem::path&, int>(const > boost::iterator_facade<boost::filesystem::directory_iterator, > boost::filesystem::path, boost::single_pass_traversal_tag, > boost::filesystem::path&, int>&, const > boost::iterator_facade<boost::filesystem::directory_iterator, > boost::filesystem::path, boost::single_pass_traversal_tag, > boost::filesystem::path&, int>&)". > > "/db/pub/eq/build/src/boost_1_31_0/libs/filesystem/build/../sr c/operations_posix_windows.cpp", line 176: Where: Instantiated from non-template code. > > > > It looks suspiciously similar to bug "[ 976241 ] Compiler > error for tokenizer on Solaris" from 2004-06-20, with the > difference that I am using version 5.3 of sunpro rather than 5.5. I have submitted a similar problem and the proposed fix recently. I am just waiting on our developers to confirm that it works before submitting it to boost, but it seems to fix the tokenizer problem, too. For a quick workaround (before the patch is applied) do this: At the top of boost/type_traits/detail/bool_trait_def.hpp you will see something like this: #if defined(__SUNPRO_CC) # define BOOST_TT_AUX_BOOL_TRAIT_VALUE_DECL(C) \ typedef mpl::bool_< C > type; \ enum { value = type::value }; \ /**/ # define BOOST_TT_AUX_BOOL_C_BASE(C) //1 Comment out the line marked with //1. That seems to have fixed my problems with both 5.3 and 5.5. HTH, Tom -------------------------------------------------------- NOTICE: If received in error, please destroy and notify sender. Sender does not waive confidentiality or privilege, and use is prohibited. _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users -- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

1 0

Help ... BOOST::thread 's certain constuctor methods fails with STLPort
by Raj 09 Oct '04

09 Oct '04

Hi I have build boost libraries for a) regular gcc b) gcc-STLPort the following code works perfectly for "gcc" and throws up the following errors for "gcc-STLPort". // this is my function object that // takes in input parameters as part of its constructor. func_obj instance(data,m_regMap); boost::shared_ptr<thread> pThread(new boost::thread(instance)); ERROR: mystuff.cpp:101: undefined reference to `boost::thread::thread[in-charge] (boost::function0<void, _STL::allocator<boost::funct ion_base> > const&)' If I assume that the boost runtime library has not been linked, that is not the case.(ldd tells me that i have) whereever I use 'xtime','lock' or 'condition' it doesn't complain. WHAT IS STRANGE IS THE FOLLOWING COMPILES boost::thread thrd; boost::thread thrd(func_obj); // it is the actual fo not an instance. Any help would be greatly appreciated. This happens on Linux/with STLPort 4.6.2/gcc 3.2.3

1 0

How to compile BOOST on Linux with NPTL threads and STLPort
by Raj 09 Oct '04

09 Oct '04

Hi, I am trying to use BOOST thread and regex libraries on Linux. I am able to compile them with sTOOLSS=gcc-stlport option. But, I would also like to specify that I would like to use the NPTL thread lib instead of pthread. How can I do it ? Any help would be greatly appreciated. Thanks in Advance Raja

2 3