May 2015 - Boost-users - lists.preview.boost.org

boost::intrusive::slist push_back() => error: static assertion failed: (cache_last)
by Jonas Printzén 09 May '15

09 May '15

Hi All! First post... Attempt switching to intrusive slist for efficiency, the following example is simply cut-paste from docs. Won't compile! g++ 4.9.1 complains as inlined in code below. #include <boost/intrusive/slist.hpp> using namespace boost::intrusive; struct point : slist_base_hook<> { int64_t x,y; }; using points = slist<point>; int main() { points plist; point p1,p2,p3; ACCEPT( 0U == plist.size() ); plist.push_back( p1 ); // => error: static assertion failed: (cache_last) ACCEPT( 1U == plist.size() ); } The code is from the docs of my version (1.55) and im pretty sure this or similar worked before. (1.54 i think) Please help, what am I missing?

2 1

Re: [Boost-users] lexical_cast between double and string slow in Visual Studio 2013
by David Roberts 07 May '15

07 May '15

> That issue is unknown. I'd really appreciate the investigation. I have done some more investigation, and there are two factors that only cause the slowness when they both occur together. > Try excluding the lexical_cast from test, I have a feeling that this is only MSVC related issue: > > #include <sstream> > #include <string> > > int main (int, char **) > { > for (double count = 0.0; count < 1000000.0; count += 1.41) > { > std::stringstream ss; > ss << count; > std::string result = std::move(ss.str()); > ss.str(std::string()); > > ss << result; > ss >> count; > } > > return 0; > } > Running your test program does not exhibit the problem. It runs in around 3 seconds on my machine when built with either Visual Studio 2010 or Visual Studio 2013. However, changing it very slightly to match more closely what lexical_cast does internally does recreate the problem: #include <sstream> #include <string> int main (int, char **) { for (double count = 0.0; count < 1000000.0; count += 1.41) { std::stringstream ss; ss.unsetf(std::ios::skipws); ss.precision(17); ss << count; std::string result = std::move(ss.str()); ss.str(std::string()); ss << result; ss >> count; } return 0; } The effect of setting the precision to 17 is that lots of 9s appear in the string representations. (The number 17 is what boost::detail::lcast_get_precision(double*) chooses.) Without the precision call the contents of the string called result start off like this: 0 1.41 2.82 4.23 5.64 7.05 8.46 9.87 11.28 12.69 With precision set to 17 they start off like this: 0 1.4099999999999999 2.8199999999999998 4.2299999999999995 5.6399999999999997 7.0499999999999998 8.4599999999999991 9.8699999999999992 11.279999999999999 12.69 This happens for both Visual Studio 2010 and Visual Studio 2013. Then the next difference is that Visual Studio 2013 spends a lot longer handling all the extra 9s. Changing the program so that the double is converted to a string using std::stringstream without a precision call and then back to double using lexical_cast takes about 3 seconds for both Visual Studio 2010 and Visual Studio 2013. It is the combination of having all the extra 9s to parse and using Visual Studio 2013 that makes the test using lexical_cast to go both ways slow. Both Visual Studio 2010 and Visual Studio 2013 do the conversion by calling std::num_get<char,std::istreambuf_iterator<char,std::char_traits<char> > >::do_get() which then calls a function called _Stodx() which is implemented in xstod.c. This function is very different for the two versions. In Visual Studio 2010 it's a relatively thin wrapper around the C function strtod(). In Visual Studio 2013 _Stodx() has got a completely new implementation that's generated by #including xxstod.h with some macros defined. The original C function strtod() is much faster than the new _Stodx() when there are lots of 9s at the end of the strings being parsed. This modification to the program: #include <sstream> #include <string> #include <stdlib.h> int main (int, char **) { for (double count = 0.0; count < 1000000.0; count += 1.41) { std::stringstream ss; ss.unsetf(std::ios::skipws); ss.precision(17); ss << count; std::string result = std::move(ss.str()); ss.str(std::string()); ss << result; char *endptr; count = strtod(ss.str().c_str(), &endptr); } return 0; } has a runtime of about 3 seconds even though it's got to cope with all the 9s. I guess only someone from Microsoft or Dinkumware could comment on why _Stodx() was reimplemented. But the other thing is that by setting precision to 17 lexical_cast is bloating the string representations of the doubles with lots of 9s in both Visual Studio 2010 and Visual Studio 2013. Setting precision to 15 instead prevents this, and makes the original test run faster even with Visual Studio 2013 (about 4 seconds rather than 10). Regards, David

5 4

[MultiIndex] getting the index from set::lower_bound() iterator (or a replacement/extension)
by U.Mutlu 07 May '15

07 May '15

Hi, I need a container like std::set which delivers also the index when using the member function lower_bound(). IMO std::set lacks this important functionality. I cannot use std::distance(myset.begin(), it), because that is very slow as it walks again all the items from the beginning. Can boost MultiIndex be used for this, or any other alternatives/workarounds? Here's a demonstration of what I need: struct TS { /*...*/ }; set<TS> myset; TS S(/*...*/); auto it = myset.lower_bound(S); if (it != myset.end()) { size_t ix = ...howto get the index, ie. the rank?... //... } -- Thx Uenal

3 2

boost container help
by Uthpal Urubail 07 May '15

07 May '15

Experts, I am looking for help on a container where I could group the objects having similar properties Ex: struct phyProp { unsigned int color; double height; double weight; std::string location; }; Can I use boost for grouping the objects having same property? [With tolerance?] Any help is highly appreciated. Regards, UJ

3 2

error: type/value mismatch at argument 1 in template parameter
by Lane 06 May '15

06 May '15

bounded_buffer is taken from the boost circular_buffer examples which wraps operations with lock acquisition and release (more type-safe). I've changed up the calls slightly for my app and the problem is, I am having difficulty compiling. I was hoping to make the calls like below w/o the need for a template function (as in the example) since I'm in the function where I'd like to make the calls. --- bounded_buffer<MyData> bb(200); Consumer<bounded_buffer> consumer(bb); Producer<bounded_buffer> producer(bb); boost::thread* c_thread = new boost::thread(consumer); boost::thread* p_thread = new boost::thread(producer); threads.add_thread(c_thread); threads.add_thread(p_thread); --- Here are the compile errors: error: type/value mismatch at argument 1 in template parameter list for 'template<class Buffer> class Consumer' error: expected a type, got 'bounded_buffer' The example code I'm using is, http://www.boost.org/doc/libs/1_55_0/libs/circular_buffer/example/circular_…

2 1

Can a mapped_file larger than 2GB be random seeked?
by Narcélio Filho 05 May '15

05 May '15

(I'm resending this because it seems that my other message got garbled.) I'm writting on random positions on a mapped file, but it seems that if I seek past 2GB, write something, then seek to a lower position, I got a segfault. Here is the code: #include <cassert> #include <iostream> #include <fstream> #include <boost/iostreams/stream.hpp> #include <boost/iostreams/device/mapped_file.hpp> using namespace std; using boost::iostreams::stream; using boost::iostreams::mapped_file; int main(int argc, char** argv) { const size_t GIGA = 1024l*1024l*1024l; const size_t FILESIZE = 2l * GIGA + 10000; ofstream block("file", ios::binary | ios::out); assert(block.good()); block.seekp(FILESIZE); block.write("", 1); block.close(); mapped_file mappedFile("file"); stream<mapped_file> file(mappedFile, ios::binary|ios::out); assert(mappedFile.is_open()); char buffer[10] = {0}; cout << 0 << endl; file.seekp(0); file.write(buffer, 1); cout << FILESIZE - 1 << endl; file.seekp(FILESIZE - 1); file.write(buffer, 1); cout << 1 * GIGA << endl; file.seekp(1 * GIGA); file.write(buffer, 1); return 0; } I'm compiling under Ubuntu 14.10 (amd64): $ g++ boost-mmap.cpp -lboost_iostreams $ ./a.out 0 2147493647 Segmentation fault (core dumped) Everything is 64-bits: $ file a.out a.out: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.32, BuildID[sha1]=d39e3dc0c99e666631747e496bf798b92e4d4c71, not stripped I have the same issue compiling with Visual Studio and running under Windows 8 (amd64). What I'm doing wrong? -- []s, Narcélio.

3 4

boost thread + container.
by Andreas Hehn 05 May '15

05 May '15

Hello boost-users readers, I stumbled across some funny behavior of the headers of boost::thread and boost::container/boost::move for C++O3. I am unsure if this is a known shortcoming of those libraries or if it is a bug. I didn't find anything on the topic in the documentation so far. The problem is dependent on the include order of the two files: * boost/container/vector.hpp * boost/thread/thread.hpp The following code causes several compilation errors (boost-1.58.0/git, g++-4.9.2) related to the move emulation of boost::thread: #include <boost/container/vector.hpp> #include <boost/thread/thread.hpp> int main() { boost::container::vector<boost::thread> v; v.resize(10); return 0; } It prints several error messages starting with boost/container/detail/copy_move_algo.hpp:285:112: error: no matching function for call to ‘move(boost::thread&)’. The following error messages are also boost::move releated "failed" enable_ifs and so on. If I switch the order of the includes everything compiles just fine. Is there a requirement on the order of the includes that I overlooked or is this a bug? Best, Andreas

1 0

How to stage one sub libary in C++ boost
by J Ronald 05 May '15

05 May '15

1. I have been built one sub library as below successfully: b2 locale toolset=msvc address-model=64 link=static threading=multi runtime-link=shared variant=release 2. Then, I tried to run: b2 stage Unexpectedly, it'll build other sub libraries, instead of copy the built sub sub library to the stage directory as the document says. Is there any automatic way to get the final built result in one directory?

1 0

Efficiently pass data from thread to main
by Lane 05 May '15

05 May '15

Using boost, I start a thread from the main process as such. MyWorkThread newThread; newThread = boost::thread(newThread); And in the worker thread class (MyWorkThread), I have a read function which reads from a USB device as much data as it can deliver. This function can be called several hundred times a second. MyThread::read() } //... while () { MyData data; if (hasMore(dev, &data)) // lots of data here // need to get it back to the main thread } } } I'm trying to figure out how to efficiently get this data back to the main process to be placed in a buffer since it can be called so often. - Would making a callback to the main process be the way to go? If so, any pointers how using boost? - Or should I just have a global data buffer and store it directly?

2 1

[serialization] std::shared_ptr with Boost >= 1.53
by Rüdiger Berlich 04 May '15

04 May '15

Hi there, I need to support various versions of Boost, but am in the process of porting our entire code base to C++11. This involves replacing those constructs of Boost that are now in the standard with their C++11-counterparts. A prime example is std::shared_ptr. However, most of our code needs to be serializable. I know that serialization of std::shared_ptr is possible with Boost 1.58. But 1.53 (the oldest version I need to support) cannot serialize std::shared_ptr . Is there any reason not to include the 1.58-Version of #include <boost/serialization/shared_ptr.hpp> instead of the older version, if I detect a Boost-version < 1.58 ? Thanks and Best Regards, Beet

4 12