- Boost-users - lists.boost.org

Re: [Boost-users] Boost-users Digest, Vol 1984, Issue 1
by John Dlugosz 04 May '09

04 May '09

> Date: Sat, 02 May 2009 13:01:07 -0600 > From: tom fogal <tfogal(a)alumni.unh.edu> > Subject: [Boost-users] Recent compilation troubles with gcc atomics > To: boost-users(a)lists.boost.org > Message-ID: <auto-000019291789(a)sci.utah.edu> > > Hello all, > http://gcc.gnu.org/ml/gcc/2009-05/msg00039.html > > In it, he notes that the availability of these builtins is not solely > a function of the compiler version. In particular, the platform (CPU, > OS, runtime support) seems to be a factor, as well as the backend code > generation options (i.e. -march gcc compiler option). > > He also suggests that the only reliable methodology for discovering the > availability of these functions/builtins is by doing link-time tests. > I mention as much because (unverified) I think boost's current solution > is #ifdefs based on architecture and compiler version, which I think is > fighting a losing battle. > > I hope this information is useful for those stuck on the problem, > and/or boost developers. > I'm in the middle of refactoring my "atomic_counter" templates, originally in the August 1998 issues of Visual C++ Developer (<http://www.dlugosz.com/Repertoire/refman/Classics/atomic_counter_white paper.html#toptop>). Besides design changes, it targets AMD-64. My experiments showed that the compiler's intrinsic support was spotty at best, and the documentation wasn't clear as to their availability. In particular, the VC++ compiler only generates intrinsic opcodes in 64-bit targets, and links to a library function for 32. The original issue is that the supplied API is too limited. If I want a post-increment (x++), the library function does the CPU opcode which is a post increment, then incs the result again to give the new value as the result, so then my code has to subtract it back out. Why dance around it? Furthermore, it didn't handle all different word sizes. More have been added to the win32 API, but it's still missing char! My latest solution is to expose them via functions written in assembly language. They cannot be inlined. But, the synchronized opcode is so slow anyway, and the call is now so cheap, being eaten by the prefetch stages, that I concluded that it didn't matter. On AMD-64, the calling convention was defined in the CPU's documentation so all platforms should follow it. The same functions should work for all operating systems, but would need macros for the actual (mangled C++) function names. Way back when, when doing it originally, I found that inline asm in an inline function in Microsoft's compiler didn't work out very well, since inlining would destroy the calling convention and values would be in the wrong registers before the body of the code! On the other hand, the WatCom compiler had a very good inline/intrinsic generator that would let you specify asm code to emit and which registers to set up for the inlined "call". I haven't seen that compiler in years, though. I'm tempted to make the new 32-bit version as a separate asm file, too, rather than inline asm. This will be more portable across different compilers and OS's, with the potential of using macros to give the correct C++ names, and if all compilers of interest can't generate the same "fastcall" calling, have prolog/epilog macros that takes the passed parameters and puts them in registers and cleans up the call, specified separately. And, use NASM to assemble it, which allows very powerful macros and can generate many target object file formats for the different operating systems. It's still x64 and x86 specific asm code. But the very concepts at this low level is CPU specific anyway. I notice the documentation for Windows on IA64 platform has "acquire/release" stuff. That is, different ways of doing it. Trying to sweet-talk each different compiler is a losing battle, especially since its supplied set of primitives is very incomplete. I conclude that just doing it in assembly language is easier. --John TradeStation Group, Inc. is a publicly-traded holding company (NASDAQ GS: TRAD) of three operating subsidiaries, TradeStation Securities, Inc. (Member NYSE, FINRA, SIPC and NFA), TradeStation Technologies, Inc., a trading software and subscription company, and TradeStation Europe Limited, a United Kingdom, FSA-authorized introducing brokerage firm. None of these companies provides trading or investment advice, recommendations or endorsements of any kind. The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you received this in error, please contact the sender and delete the material from any computer.

1 0

Using boost::assign::map_list_of with multimap
by Sean Farrow 04 May '09

04 May '09

Hi: Is there a way to use boost::assign::map_list_of() with std::multimap? If not could a list_multi_map_of function? Any help appreciated. Sean. __________ Information from ESET NOD32 Antivirus, version of virus signature database 4052 (20090504) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com

2 1

Re: [Boost-users] Boost-users Digest, Vol 1983, Issue 1
by John Dlugosz 04 May '09

04 May '09

> Date: Fri, 01 May 2009 20:15:50 -0600 > From: tom fogal <tfogal(a)alumni.unh.edu> > Subject: Re: [Boost-users] General C++: Class Method Signatures > Miss-Match > To: boost-users(a)lists.boost.org > Message-ID: <auto-000019286349(a)sci.utah.edu> > > Dominique Devienne <ddevienne(a)gmail.com> writes: > > On Fri, May 1, 2009 at 10:32 AM, Dan Day <coolmandan(a)gmail.com> > wrote: > > > I thought that int represented the natural word size of the > > > architecture. If so, then wouldn't unsigned int be 64-bit on a > > > 64-bit platform as well? > > I had heard this before as well. I took a quick look through c1x-1336 > (current (?) draft of C) though, and I could not find anything to this > effect. > > Of course, that doesn't say it's not there, and I did not check the C++ > standard. > > > The few 64-bit platforms I've seen keep [unsigned] int at > > 32-bits. --DD > What is "natural"? The AMD-x64 is built on top of the 32-bit architecture. The instruction set (in 64-bit mode) uses 32 bits as the nominal and default data size for most operations. Even for memory addresses, it's a 32-bit relative-to-IP value. It takes an opcode prefix to indicate the use of 64-bit values. The memory architecture can handle 32-bit values without penalty, and you fit more of them in the cache. So 32-bit ints are not at all "unnatural". In principle, using 32-bit ints for casual use instead of gratuitously using 64-bits where you don't need the extra size will run faster. It reduces the memory bandwidth for both data and code size, even though the internal execution is the same speed in either case. --John (sorry about the footer; it's not my idea) TradeStation Group, Inc. is a publicly-traded holding company (NASDAQ GS: TRAD) of three operating subsidiaries, TradeStation Securities, Inc. (Member NYSE, FINRA, SIPC and NFA), TradeStation Technologies, Inc., a trading software and subscription company, and TradeStation Europe Limited, a United Kingdom, FSA-authorized introducing brokerage firm. None of these companies provides trading or investment advice, recommendations or endorsements of any kind. The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you received this in error, please contact the sender and delete the material from any computer.

1 0

Character Encoding (was: General C++: Class Method Signatures)
by John Dlugosz 04 May '09

04 May '09

> Date: Fri, 1 May 2009 17:51:03 -0500 > From: Dominique Devienne <ddevienne(a)gmail.com> > Subject: Re: [Boost-users] General C++: Class Method Signatures > Miss-Match > To: boost-users(a)lists.boost.org > Message-ID: > <255d8d690905011551v108b95c4sbfeaa21c443ac23e(a)mail.gmail.com> > Content-Type: text/plain; charset=ISO-8859-1 > > On Fri, May 1, 2009 at 10:38 AM, Etienne Philip Pretorius > <icewolfhunter(a)gmail.com> wrote: > > Wide chars as far as I know are different from platform to platform. > > Microsoft uses 16 bit wide chars while *nix uses 32 bit wide chars. > And for > > unicode I need at least 21 bits. > > It's not only a question of char size, the encoding matters. I don't > know for sure, but I bet Windows' can represent all code points by > using pairs of 16-bit wide chars (surrogate pairs). As long as you > have a way to convert the current wstring instances to a know encoding > like UTF-8, UTF-16 (bom, le, be), or UTF-32 for wire transport or > persistence, the actual representation of wstring doesn't matter. The Windows API uses 16-bit characters and calls it "Unicode". Originally it was UCS2, meaning each code point was simply a 16-bit unsigned integer. Later, they added some support for surrogate pairs, which allows codes >64K to be represented as a pair of values. This is now known as UTF-16. This only works for some of the text functions, though. Much of the API and any casual handling functions will simply count each 16-bit cell as 1 character unit. So yes, if you are using the font-related functions to draw text, you can represent any code point using UTF-16. But for opening a file, it treats the name as a sequence of 16-bit values without any understanding of surrogate pairs or duplicate meanings. For manipulating wstrings in C++, neither the standard library wcs* functions nor the wstring class deals with surrogate pairs. So if you call wsclen, for example, you get the number of elements in the wchar_t array, not the number of Unicode code points, if some of them are encoded pairs. > Of course I don't know what the encoding conversion methods would be > for Windows and Linux and *nix in general. Does Boost.IOStream provide > those in a portable fashion? I'd be interested on pointers on this > topic if you know about those methods. Thanks, --DD My limited knowledge is that Linux can use UTF-8 as the code page. Some distros do that by default. General pointer on this topic: When specifying a file or interchange format, (1) use Unicode, and (2) specify the encoding or options thereof. IMHO, manipulating strings as UTF-16 is neither here nor there. You bloat the size if you are dealing with mostly Western language text, but *still* have to deal with multi-unit sequences. So if you care about characters, really, as opposed to just how much room do you need for the representation, then use 32-bit characters (I call it "xstring") or stick with UTF-8 if that's what's loaded/saved. --John (mind the footer...) TradeStation Group, Inc. is a publicly-traded holding company (NASDAQ GS: TRAD) of three operating subsidiaries, TradeStation Securities, Inc. (Member NYSE, FINRA, SIPC and NFA), TradeStation Technologies, Inc., a trading software and subscription company, and TradeStation Europe Limited, a United Kingdom, FSA-authorized introducing brokerage firm. None of these companies provides trading or investment advice, recommendations or endorsements of any kind. The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you received this in error, please contact the sender and delete the material from any computer.

1 0

[statechart] status page
by Bill Clark 04 May '09

04 May '09

The status page for boost::statechart, http://www.boost.org/development/tests/release-1_38_0/user/statechart_relea…, appears not to be working. It gives an error message, "Error extracting file: The specified zipfiles were not found." Is the status information available elsewhere?

3 4

Using a shared_ptr in a boost::function definition?
by Ashley McConnell 04 May '09

04 May '09

Hi Folks, I can't quite figure out how to do this correctly (*note* the types ending in Ptr are typedef'd boost::shared_ptr types) I have a function like so: - void setCompleteFunction(boost::function< void (int errorCode, const string &data, HeaderMapPtr headers) > f); I try to bind to it: - wc->setCompleteFunction(boost::bind(&NetworkEventHandler::receiveSkinDownloadResult, this, _1, _2, _3, wc,myFile, filename.str(), carId, crc, networkPlayerId)); I get quite a few errors indicating that it can't match to this pattern, this is the most sensible message: - error C2784: 'boost::_bi::bind_t<R,F,_bi::list_av_9<A1,A2,A3,A4,A5,A6,A7,A8,A9>::type> boost::bind(boost::type<T>,F,A1,A2,A3,A4,A5,A6,A7,A8,A9)' : could not deduce template argument for 'boost::type<T>' from 'void (__thiscall Sirocco::NetworkEventHandler::* )(int,const std::string &,Sirocco::HeaderMapPtr,Sirocco::IWebConnectionPtr,Sirocco::FilePtr,const std::string &,const std::string &,const std::string &,const std::string &)' If i don't put the shared_ptr in there (and remove the _3) it works as expected, any idea what I could be doing wrong? Thanks for your help All the best, Ash -- View this message in context: http://www.nabble.com/Using-a-shared_ptr-in-a-boost%3A%3Afunction-definitio… Sent from the Boost - Users mailing list archive at Nabble.com.

3 5

Re ducing adjacency_list's memory usage
by behzad 04 May '09

04 May '09

Hi, I wanted to ask if anybody knows how I can reduce memory usage of adj_list graph, even by eliminating some of it's features. Any help is greatly appreciated. -- View this message in context: http://www.nabble.com/Reducing-adjacency_list%27s-memory-usage-tp23363864p2… Sent from the Boost - Users mailing list archive at Nabble.com.

2 1

[Signals2] Error on building project RELEASE
by Daniele Barzotti 04 May '09

04 May '09

Hi, I'm writing a library that use the following: #include <boost/asio.hpp> #include <boost/asio/serial_port.hpp> #include <boost/thread.hpp> #include <boost/signals2/signal.hpp> #include <boost/scoped_ptr.hpp> #include <boost/enable_shared_from_this.hpp> I retrieved Signals2 from svn on 15/04/09. When I build it like STATIC DEBUG lib all works fine, but when I switch to STATIC RELEASE I got two errors... The preprocessor statements are: FOR RELEASE: WIN32;_LIB;_WINDOWS;EUROATLIB_EXPORTS;_WIN32_WINNT=0x0501 FOR DEBUG : WIN32;_DEBUG;_LIB;_WINDOWS;EUROATLIB_EXPORTS;_WIN32_WINNT=0x0501 What could be the problem? Tnx, Daniele. -------------------------------------------------------- ATDispatcher.cpp C:\Programmi\Microsoft Visual Studio 9.0\VC\include\xutility(419) : error C2664: 'boost::signals2::detail::stack_allocator<T,n_stack_elements>::stack_allocator(boost::signals2::detail::stack_storage<T,10> *)' : cannot convert parameter 1 from 'boost::signals2::detail::stack_allocator<T,n_stack_elements>' to 'boost::signals2::detail::stack_storage<T,n_stack_elements> *' with [ T=std::_Aux_cont, n_stack_elements=10 ] and [ T=boost::shared_ptr<void>, n_stack_elements=10 ] and [ T=std::_Aux_cont, n_stack_elements=10 ] No user-defined-conversion operator available that can perform this conversion, or the operator cannot be called C:\Programmi\Microsoft Visual Studio 9.0\VC\include\xutility(417) : while compiling class template member function 'std::_Container_base_aux_alloc_real<_Alloc>::_Container_base_aux_alloc_real(_Alloc)' with [ _Alloc=boost::signals2::detail::stack_allocator<boost::shared_ptr<void>,10> ] C:\Programmi\Microsoft Visual Studio 9.0\VC\include\vector(421) : see reference to class template instantiation 'std::_Container_base_aux_alloc_real<_Alloc>' being compiled with [ _Alloc=boost::signals2::detail::stack_allocator<boost::shared_ptr<void>,10> ] C:\Programmi\Microsoft Visual Studio 9.0\VC\include\vector(439) : see reference to class template instantiation 'std::_Vector_val<_Ty,_Alloc>' being compiled with [ _Ty=boost::shared_ptr<void>, _Alloc=boost::signals2::detail::stack_allocator<boost::shared_ptr<void>,10> ] G:\lib\boost_signals2\boost/signals2/detail/stack_vector.hpp(28) : see reference to class template instantiation 'std::vector<_Ty,_Ax>' being compiled with [ _Ty=boost::shared_ptr<void>, _Ax=boost::signals2::detail::stack_allocator<boost::shared_ptr<void>,10> ] G:\lib\boost_signals2\boost/signals2/detail/slot_call_iterator.hpp(39) : see reference to class template instantiation 'boost::signals2::detail::stack_vector<T,NumStackElements>' being compiled with [ T=boost::shared_ptr<void>, NumStackElements=10 ] G:\lib\boost_signals2\boost/signals2/detail/slot_call_iterator.hpp(41) : see reference to class template instantiation 'boost::signals2::detail::slot_call_iterator_cache<ResultType,Function>' being compiled Serial.cpp C:\Programmi\Microsoft Visual Studio 9.0\VC\include\xutility(419) : error C2664: 'boost::signals2::detail::stack_allocator<T,n_stack_elements>::stack_allocator(boost::signals2::detail::stack_storage<T,10> *)' : cannot convert parameter 1 from 'boost::signals2::detail::stack_allocator<T,n_stack_elements>' to 'boost::signals2::detail::stack_storage<T,n_stack_elements> *' with [ T=std::_Aux_cont, n_stack_elements=10 ] and [ T=boost::shared_ptr<void>, n_stack_elements=10 ] and [ T=std::_Aux_cont, n_stack_elements=10 ] No user-defined-conversion operator available that can perform this conversion, or the operator cannot be called C:\Programmi\Microsoft Visual Studio 9.0\VC\include\xutility(417) : while compiling class template member function 'std::_Container_base_aux_alloc_real<_Alloc>::_Container_base_aux_alloc_real(_Alloc)' with [ _Alloc=boost::signals2::detail::stack_allocator<boost::shared_ptr<void>,10> ] C:\Programmi\Microsoft Visual Studio 9.0\VC\include\vector(421) : see reference to class template instantiation 'std::_Container_base_aux_alloc_real<_Alloc>' being compiled with [ _Alloc=boost::signals2::detail::stack_allocator<boost::shared_ptr<void>,10> ] C:\Programmi\Microsoft Visual Studio 9.0\VC\include\vector(439) : see reference to class template instantiation 'std::_Vector_val<_Ty,_Alloc>' being compiled with [ _Ty=boost::shared_ptr<void>, _Alloc=boost::signals2::detail::stack_allocator<boost::shared_ptr<void>,10> ] G:\lib\boost_signals2\boost/signals2/detail/stack_vector.hpp(28) : see reference to class template instantiation 'std::vector<_Ty,_Ax>' being compiled with [ _Ty=boost::shared_ptr<void>, _Ax=boost::signals2::detail::stack_allocator<boost::shared_ptr<void>,10> ] G:\lib\boost_signals2\boost/signals2/detail/slot_call_iterator.hpp(39) : see reference to class template instantiation 'boost::signals2::detail::stack_vector<T,NumStackElements>' being compiled with [ T=boost::shared_ptr<void>, NumStackElements=10 ] G:\lib\boost_signals2\boost/signals2/detail/slot_call_iterator.hpp(41) : see reference to class template instantiation 'boost::signals2::detail::slot_call_iterator_cache<ResultType,Function>' being compiled LRRPTokenizer.cpp --------------------------------------------------------

3 4

Program options negative arguments
by David Doria 04 May '09

04 May '09

If I setup a multitoken option like this: desc.add_options() ("NumberList", po::value<vector<int> >()->multitoken(), "List of numbers.") and then call the program with: ./Program --NumberList 1 2 -3 4 it treats the number "-3" as a new flag so it produces an "unknown option" error. Is there a way to handle this negative argument? Thanks, David

2 3

Specifying daylight savings start/end in zone spec CSV file.
by Jason Cipriani 04 May '09

04 May '09

The timezone data file used by boost::local_time::tz_database is described here: http://www.boost.org/doc/libs/1_39_0/doc/html/date_time/local_time.html#tz_… My question is: Should the DST start time be specified in local time with or without DST applied? What about the DST end time? For example, America/New_York's start time is 2:00 AM. Is that 2:00 AM EST, or EDT? The end time is similar. The documentation says that the start time and end time is the "number of hours past midnight"; so I'm assuming that the start time is in non-DST, and the end time is in DST, but I just want to make sure. Thanks, Jason

1 0