November 2012 - Boost-users - lists.preview.boost.org

Serialization newbie, memory being overwritten?
by Lasse Laursen 03 Dec '12

03 Dec '12

Hello to you, I wonder if you might be able to solve this conundrum I've been pondering. The boost (1.51) serialization library is giving me some trouble. I've been battling back and forth with the library, and I'm beginning to think I'm missing some key element. At first, I had issues to just getting things written out to a file without crashes. As far as I could tell, this particular problem was compounded by the fact, that I tried to simplify my code by just writing a single primitive referenced by a pointer to an object containing said primitive. As I've since come to learn, serializing pointed-to primitives is apparently not trivial. I have since, gotten the serialization class to properly write out a few simple member primitives from my class, but when I read them back using the exact same function, it seems as though a lot - if not all - of the pre-existing data in the object, which I am not serializing, gets overwritten. As far as I have been able to tell from the tutorials, I am doing everything exactly by the book. I was also concerned that maybe boost doesn't allow partial serialization, but this shouldn't be an issue either as far as I can tell. I must be making some sort of colossal mistake that I just cannot see. My in and out code is fairly straight forward: std::string streamPath = PTMFilePath.string(); std::ifstream ifs(streamPath); boost::archive::xml_iarchive ia(ifs); ia >> BOOST_SERIALIZATION_NVP(aStream); ----- std::string streamPath = PTMFilePath.string(); std::ofstream ofs(streamPath); boost::archive::xml_oarchive oa(ofs); oa << BOOST_SERIALIZATION_NVP(aStream); ------ My serialization function: template<class Archive> void serialize(Archive &ar, const unsigned int file_version) { ar & BOOST_SERIALIZATION_NVP(nrOfSamples); } ------ aStream is a pointer to a custom class, where I have allowed boost access. I am really confused as to what is causing this. Both pointed to data and regular primitives in the aStream class are wiped/altered as soon as the program enteres the serialize function, and even before the actual data seems to be read. I must be overlooking something really basic here. I hope someone can whack me with the solution stick... Thanks, Lasse

3 6

Incorrect assertion in Boost.Exception?
by Thorsten Dahlheimer 02 Dec '12

02 Dec '12

Hello, we want to use Boost.Exception (1.47.0) with MS Visual Studio 2010 in a project that consists of several DLLs and a main executable. We have situations where we store an error_info into an exception from inside a DLL and access it again from another DLL/the EXE, e.g.: dll.cpp: void func() { BOOST_THROW_EXCEPTION(my_exception() << my_info(42)); } exe.cpp: try { func(); } catch (boost::exception& e) { std::cout << *boost::get_error_info<my_info>(e) << std::endl; } Now at runtime, when the error info is accessed, the following assertion in boost::exception_detail::error_info_container_impl::get() fails: BOOST_ASSERT( BOOST_EXCEPTION_DYNAMIC_TYPEID(*p).type_==ti.type_ ); As far as we can see, this happens because MSVC++ uses seperate type_info objects in each dll/exe, and the above assertion compares pointers to type_infos instead of the objects themselves. We believe this is a bug on the Boost side, since type_info objects for the same type are generally not guaranteed to be unique. Are we correct with that? Thanks, Thorsten

2 2

Re: [Boost-users] [Range & c++0x Lambdas] Can this be done?
by John M. Dlugosz 01 Dec '12

01 Dec '12

I had the same issue and posted it a few months ago. I wrote to the MS VS forum that their std::result_of should work correctly with both old and new formats; in particular use the declared result_type if it exists (which in general lets the user override the deduced type, great for compatibility issues). Actually doing that robustly is a bit of a maze of metaprogramming. I ended up making a local my_namespace::result_of that I just handles what I needed, and only when I asked for it. That doesn't help the existing range templates, but such a change could be put back into boost::result_of. Meanwhile, for different reasons, I made a macro that abstracts the different syntax between C++11 lambdas and Apple Block Closures. The latter generates a suitable lambda class complete with result_of member. That makes me think that you could have a macro that wraps a compiler-generated lambda expression in another class that inherits the operator() etc. and adds in the result_of typedef to simply use decltype. —John Date: Wed, 21 Nov 2012 10:15:18 -0700 From: Nathan Crookston <nathan.crookston(a)gmail.com> To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] [Range & c++0x Lambdas] Can this be done? Message-ID: <CAMb0s=1qtLcOiXTd5h7b2qkOSiYqjEVeN-m=rafJRHKDU6y6Cg(a)mail.gmail.com> Content-Type: text/plain; charset="iso-8859-1" Rob, On Wed, Nov 21, 2012 at 6:39 AM, Robert Jones <robertgbjones(a)gmail.com>wrote: >> Can the last line, labelled NOT Ok, be made to work? I think the lambda >> does >> not publish its result as a bind does, so I suspect it's hopeless. Any >> thoughts? >> // NOT Ok >> boost::range::push_back( out, in | transformed( []( S & s ) { return >> s.i; } ) ); >> } >> >> It seems trivial to have a nested result_type in all cases with a lambda. > However, in C++11 there's no need for it due to decltype. > > Sadly, some compilers fall into a gap where result_of (boost or std) > doesn't use decltype and lambdas don't publish tr1-style result_type -- > leaving code like the above in the lurch. > > Michel Morin wrote some code for adapting such a case which may be of > interest to you: http://lists.boost.org/boost-users/2012/01/72879.php > > HTH, > Nate >

5 21

Boost.Wave getting raw input tokens for code transformation
by Frank Birbacher 30 Nov '12

30 Nov '12

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi! I'm trying to utilize Wave for simple code transformations. My starting point is to do an identity mapping through Wave: read in a source file, tokenize it, and output it again. The output shall be the same as the input on a byte by byte basis. I modified the cpp_lexer example to just output the token.value() texts. From that I discovered whitespace in "# define" will be dropped. Ok, fixed this by modifying Wave source code. Now my main problem are line continuations using backslash like in multiline macro definitions. Those get processed at a very low level, I guess. The backslash will never be reported as a token, but the adjacent characters will be fused into one token. Is there a way to define a "raw" token mode that would report the backslash and keep the adjacent characters as distinct tokens? I don't understand the tokenizer level at all. Frank -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.17 (Darwin) Comment: GPGTools - http://gpgtools.org Comment: keyserver x-hkp://pool.sks-keyservers.net iEYEARECAAYFAlC1w/8ACgkQhAOUmAZhnmrUsQCeIVBxkYTUg1qIU2DTsSZuh5u1 2v8AniMF44HV4agBashrP81N6XN3bTW1 =A8Sp -----END PGP SIGNATURE-----

2 2

serialization of recursive objects
by kean 29 Nov '12

29 Nov '12

Hello everybody, I defined the following objects to be serialized: class MyElementObject { friend class boost::serialization::access; public: template<class Archive> void serialize(Archive & ar, const unsigned int version) { } }; template<class T> class MyRecursiveObject { friend class boost::serialization::access; public: T element; std::vector<MyRecursiveObject<T> > children; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & element; ar & children; } }; I then run the following code: int main() { //MyRecursiveObject initialization MyRecursiveObject<MyElementObject> rec_object; rec_object.children.push_back(MyRecursiveObject<MyElementObject>()); rec_object.children[0].children.push_back(MyRecursiveObject<MyElementObject>()); //create vector of pointers to MyRecursiveObject's elements vector<MyElementObject *> elt_ptrs; elt_ptrs.push_back(&rec_object.element); elt_ptrs.push_back(&rec_object.children[0].element); elt_ptrs.push_back(&rec_object.children[0].children[0].element); //serialize MyRecursiveObject and the vector of pointers { ofstream ofs("filename"); boost::archive::text_oarchive oa(ofs); oa << rec_object; oa << elt_ptrs; } //create new MyRecursiveObject and vector of pointers for deserialization MyRecursiveObject<MyElementObject> rec_object_deserialized; rec_object_deserialized.children.push_back(MyRecursiveObject<MyElementObject>()); rec_object_deserialized.children[0].children.push_back(MyRecursiveObject<MyElementObject>()); vector<MyElementObject *> elt_ptrs_deserialized; //deserialize { ifstream ifs("filename"); boost::archive::text_iarchive ia(ifs); ia >> rec_object_deserialized; ia >> elt_ptrs_deserialized; } //compare deserialized pointers cout<<"elt_ptrs first level="<<elt_ptrs_deserialized[0] <<" expected="<<&rec_object_deserialized.element<<endl; cout<<"elt_ptrs second level="<<elt_ptrs_deserialized[1] <<" expected="<<&rec_object_deserialized.children[0].element<<endl; cout<<"elt_ptrs third level="<<elt_ptrs_deserialized[2] <<" expected="<<&rec_object_deserialized.children[0].children[0].element<<endl; return 0; } And I always get an output similar to the following one: elt_ptrs first level=<b>0x7fff57c787c0* expected=*0x7fff57c787c0* elt_ptrs second level=*0x18e7020* expected=*0x18e7020* elt_ptrs third level=*0xffff8000ab5564f0* expected=*0x18e7450* As can be seen from the output I manage to deserialize pointers that point to elements down to the second recursion level of MyRecursiveObject. As soon as I try to do it with pointers to the third level or even deeper the deserialization fails. Am I using boost::serialization wrongly? Note that the MyRecursiveObject is always successfully deserialized, no matter how many levels it has. I encounter the problem only deserializing pointers to its elements. Thank you in advance Kean -- View this message in context: http://boost.2283326.n4.nabble.com/serialization-of-recursive-objects-tp463… Sent from the Boost - Users mailing list archive at Nabble.com.

2 1

Is it ok to pass &my_boost_array_variable[2] into a C function expecting a Type*
by Anton Daneyko 29 Nov '12

29 Nov '12

Hi, I have a two variables: boost::array my_boost_array_variable_1<float, 3>; boost::array my_boost_array_variable_2<float, 3>; now I want to call a C function with the following signature: int MPI_Allreduce ( void *sendbuf, void *recvbuf, ...); I want to pass the address of a particular element of my my_boost_array_variable_1 and my_boost_array_variable_2 as a first and second parameter of MPI_Allreduce: MPI_Allreduce(&my_boost_array_variable_1[2], &my_boost_array_variable_2[2], ... or should I rather do: MPI_Allreduce(my_boost_array_variable_1.c_array() + 2, my_boost_array_variable_2.c_array() + 2, ... Thanks, Anton

2 1

strange linkage problem with boost_date_time 1.52 (x86_64-linux-gnu/gcc version 4.7.2)
by mypop＠freenet.de 29 Nov '12

29 Nov '12

Hi erveryone, i have a little problem linking against my newly installed: boost_date_time library. I even don't know if this is the right place for this question because everything works fine so far in the past. 1. I never had problems compiling and linking this sources before (on 3 other systems). 2. The project also contains some other directories with some kind of commandline tools. These tools also link against the regex and date_time libraries without any problems. Here is the part where the libs are defined in my Makefile.am libmodwscmi_la_CXXFLAGS = $(INCLUDES) $(PROJECT_INCLUDES) $(APR_CFLAGS) libmodwscmi_la_LDFLAGS = $(LDADD) $(BOOST_LDFLAGS) $(LIB_BOOST_REGEX) $(LIB_BOOST_DATETIME) `$(APR1CONFIG) --link-ld --libs` libmodwscmi_la_LIBADD = $(top_builddir)/utils/mysqlpp/libmysqlpp.la $(top_builddir)/utils/jwsmtp/libjwsmtp.la After compiling is complete the linker gets called by this command: g++ -fPIC -DPIC -shared -nostdlib /usr/lib/gcc/x86_64-linux-gnu/4.7/../../../x86_64-linux-gnu/crti.o /usr/lib/gcc/x86_64-linux-gnu/4.7/crtbeginS.o .libs/libmodwscmi_la-mod_wscmi.o .libs/libmodwscmi_la-mod_gsoap.o .libs/libmodwscmi_la-mod_utils.o .libs/libmodwscmi_la-mod_mysql.o .libs/libmodwscmi_la-md5.o .libs/libmodwscmi_la-regEx.o .libs/libmodwscmi_la-utils.o .libs/libmodwscmi_la-exception.o .libs/libmodwscmi_la-dateTime.o .libs/libmodwscmi_la-validator.o .libs/libmodwscmi_la-mysqlPool.o .libs/libmodwscmi_la-mysqlCons.o .libs/libmodwscmi_la-mysqlQueries.o .libs/libmodwscmi_la-sqlite3.o -Wl,-rpath -Wl,/var/www/projects/mod_wscmi-1.0/utils/mysqlpp/.libs -Wl,-rpath -Wl,/var/www/projects/mod_wscmi-1.0/utils/jwsmtp/.libs -L/usr/local/lib -lboost_date_time -lboost_regex -L/usr/lib /usr/lib/libapr-1.so -L/usr/local/mysql/lib ../utils/mysqlpp/.libs/libmysqlpp.so ../utils/jwsmtp/.libs/libjwsmtp.so -L/usr/lib/gcc/x86_64-linux-gnu/4.7 -L/usr/lib/gcc/x86_64-linux-gnu/4.7/../../../x86_64-linux-gnu -L/usr/lib/gcc/x86_64-linux-gnu/4.7/../../../../lib -L/lib/x86_64-linux-gnu -L/lib/../lib -L/usr/lib/x86_64-linux-gnu -L/usr/lib/../lib -L/usr/lib/gcc/x86_64-linux-gnu/4.7/../../.. -lstdc++ -lm -lc -lgcc_s /usr/lib/gcc/x86_64-linux-gnu/4.7/crtendS.o /usr/lib/gcc/x86_64-linux-gnu/4.7/../../../x86_64-linux-gnu/crtn.o -pthread -O3 -Wl,-R/usr/local/lib -pthread -Wl,-soname -Wl,libmodwscmi.so.0 -o .libs/libmodwscmi.so.0.0.0 As you can see above, both libraries are present: -lboost_date_time -lboost_regex. They also can be found on the harddrive (in /use/local/lib): ... -rw-r--r-- 1 root 136876 Nov 20 09:19 libboost_date_time.a lrwxrwxrwx 1 root 28 Nov 20 09:18 libboost_date_time.so -> libboost_date_time.so.1.52.0 -rwxr-xr-x 1 root 94337 Nov 20 09:18 libboost_date_time.so.1.52.0 ... -rw-r--r-- 1 root 2619610 Nov 20 09:19 libboost_regex.a lrwxrwxrwx 1 root 24 Nov 20 09:18 libboost_regex.so -> libboost_regex.so.1.52.0 -rwxr-xr-x 1 root 1251576 Nov 20 09:18 libboost_regex.so.1.52.0 ... The complete build runs fine! There are no errors or warnings but the fianlly library is unusable because: boost_date_time is not present. Ldd shows up something like this: ldd .libs/libmodwscmi.so.0.0.0 linux-vdso.so.1 => (0x00007fff17da4000) libboost_regex.so.1.52.0 => /usr/local/lib/libboost_regex.so.1.52.0 (0x00007ff289006000) libapr-1.so.0 => /usr/lib/libapr-1.so.0 (0x00007ff288db5000) libmysqlpp.so.0 => /var/www/projects/mod_wscmi-1.0/utils/mysqlpp/.libs/libmysqlpp.so.0 (0x00007ff288b35000) libstdc++.so.6 => /usr/lib/x86_64-linux-gnu/libstdc++.so.6 (0x00007ff288832000) libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007ff288472000) libgcc_s.so.1 => /lib/x86_64-linux-gnu/libgcc_s.so.1 (0x00007ff28825c000) libpthread.so.0 => /lib/x86_64-linux-gnu/libpthread.so.0 (0x00007ff28803f000) libuuid.so.1 => /lib/x86_64-linux-gnu/libuuid.so.1 (0x00007ff287e39000) libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00007ff287c35000) libmysqlclient.so.18 => /usr/lib/x86_64-linux-gnu/libmysqlclient.so.18 (0x00007ff2876f1000) libm.so.6 => /lib/x86_64-linux-gnu/libm.so.6 (0x00007ff2873f4000) /lib64/ld-linux-x86-64.so.2 (0x00007ff2895e9000) libz.so.1 => /lib/x86_64-linux-gnu/libz.so.1 (0x00007ff2871dd000) Only libboost_regex.so.1.52.0 was linked in and i have absolutly no idea why: libboost_date_time does not show up. So maybe someone of the list can give me an adivse of what's is going on here. Thanks and greetings Michael --- E-Mail ist da wo du bist! Jetzt mit freenetMail ganz bequem auch unterwegs E-Mails verschicken. Am besten gleich informieren unter http://mail.freenet.de/mobile-email/index.html

3 3

Fwd: [boost] [thread] Crash when calling pthread_cond_wait from thread.cpp in iOS simulator
by Brent Lewis 29 Nov '12

29 Nov '12

I've determined that pthread_cond_wait works fine on an iPhone (hardware) and works fine when called from my own module on simulator (I use boost::condition_variable, and since it's header only and calling pthread_cond_wait, it's my module) But when I create a new thread, which goes through thread.cpp in the thread library, call_once in once.hpp has to (sometimes, based on synchronization randomness) call pthread_cond_wait which results in the message: Detected an attempt to call a symbol in system libraries that is not present on the iPhone: (null) called from function _ZN5boost9call_onceIPFvvEEEvRNS_9once_flagET_ in image Blackjack If I make it continue, I get a problem report from OSX which tells me that the stack trace is: 0 libsystem_kernel.dylib 0x903a5a6a __pthread_kill + 10 1 libsystem_c.dylib 0x96afbacf pthread_kill + 101 2 libsystem_sim_c.dylib 0x03a0457b abort + 140 3 libSystem.dylib 0x032ad4ad _interposition_vtable_unimplemented + 200 4 libSystem.dylib 0x032c0938 pthread_cond_wait$UNIX2003 + 60 5 Blackjack 0x00277c8c void boost::call_once<void (*)()>(boost::once_flag&, void (*)()) + 156 (once.hpp:112) 6 Blackjack 0x0027506b thread_proxy + 139 (thread.cpp:135) 7 libsystem_c.dylib 0x96afa557 _pthread_start + 344 8 libsystem_c.dylib 0x96ae4cee thread_start + 34 Is it maybe some discrepancy in the settings I'm using to build boost that's causing to link something wrong... something else? I've attached the complete crash log. In case it is build related, this is how I've built boost using a shell script: LIBS=" --with-thread --with-serialization --with-date_time" CPLUS_INCLUDE_PATH=/Applications/Xcode.app/Contents/Developer/Platforms/iPhoneSimulator.platform/Developer/SDKs/iPhoneOS6.0.sdk/usr/include/ PATH=/Applications/Xcode.app/Contents/Developer/Platforms/iPhoneSimulator.platform/Developer/usr/bin:$PATH ./b2 toolset=darwin runtime-link=static link=static variant=release include=$CPLUS_INCLUDE_PATH cxxflags=--visibility=default cxxflags=-Wno-long-long --build-dir=$HOME/boost/lib/build --includedir=$HOME/boost/include --libdir=$HOME/boost/lib/build/arm-release architecture=arm target-os=iphone $LIBS -q install

1 0

[counting_range] why?
by Robert Jones 28 Nov '12

28 Nov '12

Ok, maybe I'm having a senior moment here, but why is the second line wrong? Thx, Rob. #include <boost/range/counting_range.hpp> int main( ) { // Ok auto r = boost::counting_range( 0, 3 ); // Not Ok. auto s = boost::counting_range<int>( 0, 3 ); }

3 6

[graph] File Dependency Example with updates
by Brian Budge 28 Nov '12

28 Nov '12

Hi all - I am toying around with the file dependency example that is part of the boost graph documentation as a way of adapting the concepts to my own use cases. What I want to do is update the graph over time (add more files with more dependencies), and incrementally decide what needs to be built, and in what order. It makes sense that doing a breadth-first search from a changed node will give me everything that might be affected; however, if I change many nodes, is there something better? I could imagine that I might mark nodes as they are encountered so that other BFS from other changed nodes won't traverse those as well. With boost graph, how can I control the BFS to stop recursion if a vertex has already been visited? I guess then once I have seen all the stale vertices, I can make a smaller graph that only has the stale vertices and their connections, and I can topologically sort the sub-graph to find the parallel build order? Sorry, I'm a bit out of practice with my graph algorithms in general, and I'm very new to the boost graph library, so I hope I haven't said anything dumb :) Any design suggestions welcome. Thanks, Brian

2 6