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

Program Options and Filesystem ifstream
by Scott Morgan 19 Aug '15

19 Aug '15

Having a problem getting the following to compile under a Linux system (Mint 17.2, boost 1.54). It works okay with MSVC2010 and boost 1.57 though. **begin: main.cpp** #include <boost/filesystem/fstream.hpp> namespace fs = boost::filesystem; #include <boost/program_options.hpp> namespace po = boost::program_options; void test(fs::path& cfg_path, po::options_description& cfg_desc, po::variables_map& options) { po::store(po::parse_config_file(fs::ifstream(cfg_path), cfg_desc, true), options); } **end: main.cpp** Is this just a problem with 1.54's version of filesystem and/or program options? Scott **begin: GCC output** main.cpp: In function ‘void test(boost::filesystem::path&, boost::program_options::options_description&, boost::program_options::variables_map&)’: main.cpp:8:72: error: no matching function for call to ‘parse_config_file(boost::filesystem::ifstream, boost::program_options::options_description&, bool)’ po::store(po::parse_config_file(fs::ifstream(cfg_path), cfg_desc, true), options); ^ main.cpp:8:72: note: candidates are: In file included from /usr/include/boost/program_options.hpp:17:0, from main.cpp:3: /usr/include/boost/program_options/parsers.hpp:185:5: note: boost::program_options::basic_parsed_options<charT> boost::program_options::parse_config_file(std::basic_istream<charT>&, const boost::program_options::options_description&, bool) [with charT = char] parse_config_file(std::basic_istream<charT>&, const options_description&, ^ /usr/include/boost/program_options/parsers.hpp:185:5: note: no known conversion for argument 1 from ‘boost::filesystem::ifstream {aka boost::filesystem::basic_ifstream<char>}’ to ‘std::basic_istream<char>&’ /usr/include/boost/program_options/parsers.hpp:198:5: note: template<class charT> boost::program_options::basic_parsed_options<charT> boost::program_options::parse_config_file(const char*, const boost::program_options::options_description&, bool) parse_config_file(const char* filename, const options_description&, ^ /usr/include/boost/program_options/parsers.hpp:198:5: note: template argument deduction/substitution failed: main.cpp:8:72: note: cannot convert ‘boost::filesystem::basic_ifstream<char>((*(const boost::filesystem::path*)cfg_path))’ (type ‘boost::filesystem::ifstream {aka boost::filesystem::basic_ifstream<char>}’) to type ‘const char*’ po::store(po::parse_config_file(fs::ifstream(cfg_path), cfg_desc, true), options); ^ **end: GCC output**

2 3

multi_index member key extractors crash MSVC 2015
by Filip Konvička 19 Aug '15

19 Aug '15

Hi, it looks like using boost::multi_index::member<> crashes MSVC 2015 RTM when a const-qualified type is passed to it as the 2nd argument, but the member is actually non-const. #include <boost\multi_index_container.hpp> #include <boost\multi_index\member.hpp> #include <boost\multi_index\indexed_by.hpp> #include <boost\multi_index\ordered_index.hpp> struct X { int number; // non-const }; using namespace boost::multi_index; int main() { typedef multi_index_container<X, indexed_by<ordered_unique<member<X, const int, &X::number> > > > C; C c; c.find(1); // C1001 return 0; } ...\include\boost\multi_index\member.hpp(65): fatal error C1001: An internal error has occurred in the compiler. (compiler file 'f:\dd\vctools\compiler\utc\src\p2\main.c', line 246) I found the following on a related subject, but I'm not sure whether this is a separate issue or not. http://www.boost.org/development/tests/master/developer/output/teeks99-08l-… https://connect.microsoft.com/VisualStudio/feedback/details/1577162/vc-inte… https://groups.google.com/forum/#!topic/boost-devel-archive/gi-xGnCkvPI Thanks, Filip

2 1

[Asio] Measuring and improving SSL servier scalability
by Konrad Zemek 18 Aug '15

18 Aug '15

Hi all, I've built a few different Asio-based SSL servers in my career, but in my recent project the performance and resource-efficiency was paramount (as it's a library project), so I benchmarked it to make sure. Long story short, when benchmarking I found out that OpenSSL's error handling is a *very* serious bottleneck when scaling Asio+SSL over multiple threads. This can be "fixed", e.g. by using BoringSSL instead of OpenSSL, as BoringSSL uses thread-local storage for error data instead of locking a global mutex. As I didn't find any resources discussing bad Asio+SSL multithreaded performance (aside of one or two messages somewhere on this mailing list that complained about the problem; unfortunately I can't seem to find them right now), I've written up my knowledge into a blog post [1]. I believe it might be very useful to anyone who wants to build a multi-threaded SSL server/client on top of Asio. Konrad [1] https://konradzemek.com/2015/08/16/asio-ssl-and-scalability/

2 2

[units] implicit type conversions to specific unit quantity
by Michael Powell 17 Aug '15

17 Aug '15

Hello, I'd like to convert from an SI length quantity to a US inches quantity. I believe the following snippet will do what I want it to, but wanted to clarify if that was the case: http://www.boost.org/doc/libs/1_59_0/doc/html/boost_units/Quantities.html I am referencing the online docs, but needed some clarity in actually how to go about using it in practice: <code> using boost::units::si::meter_base_unit; using boost::units::us::inch_base_unit; // i.e. 0.3048 * 12 (inches/feet, feet/meter) BOOST_UNITS_DEFINE_CONVERSION_FACTOR(inch_base_unit, meter_base_unit, double, 3.6576); //... quantity<inch_unit> to_inches_length_quantity(quantity<length> const & L) { const quantity<inch_unit::unit_type> result_(L); return result_; } </code> Just besides that, there isn't a cleaner way of doing it? Some have mentioned going as far as creating a whole other system for my calculator's purposes, for example. Thank you... Best regards, Michael Powell

2 2

Re: [Boost-users] Building Boost 1.59.0 on Solaris 11 with Solaris Studio 12.4
by Leo Carreon 17 Aug '15

17 Aug '15

From: Leo Carreon Sent: Sunday, August 16, 2015 8:39 PM To: Boost Users Mailing List Subject: Building Boost 1.59.0 on Solaris 11 with Solaris Studio 12.4 >>Hi Boosters, >>I have been attempting to build Boost 1.59.0 on Solaris 11 using Solaris Studio 12.4 in C++11 mode. >>So far, I have managed to locate on the web solutions to 3 issues. >>Now I’m down to 22 targets failing to build. >>Has anyone else attempted to do the same? I would appreciate it if you could share some of your fixes if you have any. >>Has anyone reported that they have been able to cleanly build in this environment? >>Kind regards, >>Leo This is an update of my previous mail. I now have fixed 4 issues instead of 3. I also had a look at the build log generated from the above. It appears that the following issues are the cause of the remaining build problems which I haven’t figured out how to resolve yet: 1. ) "libs/thread/src/pthread/thread.cpp", line 103: Error: The operation "boost::shared_ptr<boost::detail::tss_cleanup_function> && bool" is illegal. "libs/thread/src/pthread/thread.cpp", line 144: Error: Overloading ambiguity between "boost::call_once<void(*)()>(boost::once_flag&, void(*)())" and "boost::call_once<void(&)()>(boost::once_flag&, void(&)())". "libs/thread/src/pthread/thread.cpp", line 150: Error: Overloading ambiguity between "boost::call_once<void(*)()>(boost::once_flag&, void(*)())" and "boost::call_once<void(&)()>(boost::once_flag&, void(&)())". "libs/thread/src/pthread/thread.cpp", line 803: Error: The operation "bool && boost::shared_ptr<boost::detail::tss_cleanup_function>" is illegal. "libs/thread/src/pthread/thread.cpp", line 807: Error: The operation "boost::shared_ptr<boost::detail::tss_cleanup_function> || bool" is illegal. "libs/thread/src/pthread/thread.cpp", line 817: Error: The operation "boost::shared_ptr<boost::detail::tss_cleanup_function> || bool" is illegal. 2.) "./boost/thread/detail/nullary_function.hpp", line 125: Error: Cannot use std::shared_ptr<boost::detail::nullary_function<void()>::impl_base> to initialize bool. "./boost/thread/detail/nullary_function.hpp", line 125: Error: An expression of scalar type was expected. 3.) "libs/coroutine/src/posix/stack_traits.cpp", line 56: Error: Overloading ambiguity between "boost::call_once<void(&)(unsigned long*), unsigned long*>(boost::once_flag&, void(&)(unsigned long*), unsigned long*&&)" and "boost::call_once<void(*)(unsigned long*), unsigned long*>(boost::once_flag&, void(*)(unsigned long*), unsigned long*)". "libs/coroutine/src/posix/stack_traits.cpp", line 64: Error: Overloading ambiguity between "boost::call_once<void(&)(rlimit*), rlimit*>(boost::once_flag&, void(&)(rlimit*), rlimit*&&)" and "boost::call_once<void(*)(rlimit*), rlimit*>(boost::once_flag&, void(*)(rlimit*), rlimit*)". 4.) "./boost/proto/detail/preprocessed/make_expr_.hpp", line 73: Error: type is not a member of boost::proto::transform<boost::proto::switch_<boost::phoenix::phoenix_generator, boost::proto::tag_of<boost::proto::_>()>, void>::result<boost::phoenix::phoenix_generator(boost::proto::exprns_::basic_expr<boost::proto::tagns_::tag::assign, boost::proto::argsns_::list2<boost::proto::exprns_::basic_expr<boost::proto::tagns_::tag::terminal, boost::proto::argsns_::term<boost::phoenix::argument<1>>, 0>, boost::phoenix::actor<boost::proto::exprns_::basic_expr<boost::proto::tagns_::tag::terminal, boost::proto::argsns_::term<boost::phoenix::argument<1>>, 0>>>, 2>)>. 5.) "libs/python/src/converter/builtin_converters.cpp", line 474: Error: Overloading ambiguity between "std::complex<double>::complex(double, double)" and "std::complex<double>::complex(complex double[2])". "libs/python/src/converter/builtin_converters.cpp", line 474: Error: Cannot return long from a function that should return std::complex<double>. 2 Error(s) detected. 6.) "./boost/thread/once.hpp", line 38: Error: Overloading ambiguity between "boost::call_once<void(*)()>(boost::once_flag&, void(*)())" and "boost::call_once<void(*)()&>(boost::once_flag&, void(*)()&)". 7.) "./boost/spirit/home/classic/core/non_terminal/impl/object_with_id.ipp", line 145: Error: Overloading ambiguity between "boost::call_once<void(*)()>(boost::once_flag&, void(*)())" and "boost::call_once<void(&)()>(boost::once_flag&, void(&)())". 8.) "./boost/spirit/home/classic/phoenix/closures.hpp", line 427: Error: Overloading ambiguity between "boost::call_once<void(*)()>(boost::once_flag&, void(*)())" and "boost::call_once<void(&)()>(boost::once_flag&, void(&)())". These remaining issues are preventing me from building my own library because it depends on some of the Boost libraries that are still failing to build.

1 0

Re: [Boost-users] [boost] Making the Windows binaries 'official'
by Tom Kent 15 Aug '15

15 Aug '15

On Fri, Aug 14, 2015 at 8:34 PM, Glen Fernandes <glen.fernandes(a)gmail.com> wrote: > On Fri, Aug 14, 2015 at 9:03 PM, Tom Kent wrote: > > Is this something the community would be in favor of? I know that many of > > the people on these lists don't use windows and if they do are generally > > capable of building the windows binaries themselves, however there is a > > large group out there that prefers to just get the installer for them > (like > > they do for other projects like Qt, wxWidgets, etc). > > > > Thoughts? > > While I have not been a MSVC user for a while now: I think it's a good > thing for our VC users that binaries are available to them. > > Is (1) necessary? I was under the impression binaries becoming > available a few days after a source release is not uncommon. > This is basically how it is now....if people don't think it is necessary we can stick with this. However, I think it would be more professional if we change, hence my e-mail to the list. > SourceForge raises the usual concern: Can you trust a Windows > executable or DLL that you get from sourceforge.net anymore? > The .7z file and installers are SHA-256 hashed, and the list of hashes is signed with my GPG key. That said, I think we need to start planning an alternative to sourceforge for releases. It is not clear to me that it will still be around in a few years. That's a whole different thread though. > > Also, how do you cater to users who want variants that you're not > providing? i.e. Already you probably have Debug/Release * > 32-bit/64-bit * MT/MD/MTd/MDd * LIB/DLL (per 6 compiler versions). > What happens if someone wants /Gz or /Gv instead of /Gd etc. > I provide whatever bjam specifies with --build-type=complete. The bigger question is external libraries. I link with zlib, and bzip2 but not MPI for instance. I have had users ask for different version numbers of those libraries, but my answer is then that you're on your own for building anything else. These kind of questions are only relevant to far less than 1% of the people who download the binaries. Tom

2 1

[type_erasure] BOOST_TYPE_ERASURE_MEMBER with template?
by Nat Goodspeed 15 Aug '15

15 Aug '15

The code with which I'm working has two similar template classes: they have a common subset of 8 methods that could be used on an instance of either. For performance reasons, these methods are not virtual. As far as the compiler knows, the two template classes are unrelated. Naturally we can write a template function to accept an instance of either template class, as long as it uses only methods from the common subset. I also want to support runtime polymorphism, permitting a non-template function to accept an instance of either template class. Obviously that will incur some performance hit, but the coder can decide whether or not his/her function is performance-critical. If so, s/he can write that function to accept one or the other specific template class -- or write a template function. We want to incur the performance hit only when the coder deems that an acceptable tradeoff. The classic solution would be to derive both template classes from a common base class and make the common methods virtual -- but then that penalizes every consumer. I haven't yet used Boost.TypeErasure, but it seems to me potentially applicable to this problem. I see the example[0] illustrating use of BOOST_TYPE_ERASURE_MEMBER(). I was going to define requirements for each of the 8 common methods ... but then I became confused. Our common methods use a template parameter from the parent template class as a method parameter type. Can I invoke BOOST_TYPE_ERASURE_MEMBER in a generic way that would support all valid template parameters? Two of the 8 common methods are template methods, accepting a parameter whose type is a template parameter for the method rather than the containing class. I don't know whether that introduces further complexity. Thank you for any help you can provide! [0] http://www.boost.org/doc/libs/1_59_0/doc/html/boost_typeerasure/basic.html

2 1

Making the Windows binaries 'official'
by Tom Kent 15 Aug '15

15 Aug '15

<history> After the boostpro.com organization dissolved, there was never an official place to go to get pre-built binaries for the releases. Some random components were made available on the sourceforge download page and I had been hosting them on my own website (http://boost.teeks99.com) since I started building them around the release of 1.37. A couple years ago, before the release of 1.54, I asked the community if they would like my builds to be hosted on the sourceforge site, and the answer was positive: http://lists.boost.org/Archives/boost/2013/04/202460.php Since the 1.54 release, I've generally been uploading the windows binaries based on the RCs and enabling the downloads just a few minutes after the release notice goes out. </history> I would now like to propose that we make these libraries an official part of each Boost release. I don't think this entails much, just the following: 1) Have the release team coordinate with me to ensure that the binaries are ready before announcing the release. 2) Include a link to them in the release announcement. 3) Link to them from the website. It would also be good to get others who can run the scripts to perform the build, if I am someday unable to do this. The biggest hurdle here is having a (virtual) machine with all six versions of visual studio on it. Is this something the community would be in favor of? I know that many of the people on these lists don't use windows and if they do are generally capable of building the windows binaries themselves, however there is a large group out there that prefers to just get the installer for them (like they do for other projects like Qt, wxWidgets, etc). Thoughts? Tom Kent

1 0

[Graph] Descriptor invalidation on remove_vertex() with adjacency_list VL=vecS
by Chris Dellin 14 Aug '15

14 Aug '15

Hello, The documentation for the adjacency_list [0] is clear that if the vertex list container is vecS, then calling remove_vertex() will invalidate any vertex/edge descriptors into the graph. However, I was wondering if this is actually the case if the removed vertex was the one most recently added, that is, remove_vertex(vertex(num_vertices(g)-1)). In this case, I would think that no shifting of vertex data or adjustment of edge source/target indices would be necessary, and existing descriptors are guaranteed to be still valid. Is this safe to do with the current implementation? If not, why not? Thanks, - Chris Dellin [0] http://www.boost.org/doc/libs/1_59_0/libs/graph/doc/adjacency_list.html

2 1

[shared_ptr] boost::shared_ptr deleter question
by Tamas.Ruszkai＠leica-geosystems.com 14 Aug '15

14 Aug '15

Hi All, I wonder when the shared_ptr deleter is destroyed. I have a shared_ptr with a custom deleter. I would expect when the last shared pointer to the object is gone then not only the deleter method will be called, but the deleter function object is also released. As I see the deleter is taken by value and stored by value in the reference counter, so I do not understand what is going on. How is the deleter mean to be freed? Thanks, Tamas struct MyClassDeleter { ~MyClassDeleter() { // Break point is set here } void operator()(MyClass* item) { // Break point is set here } } { auto sharedPtr = boost::shared_ptr<MyClass>(new MyClass, MyClassDeleter) } // shared_ptr leaves the scope here, the custom deleter is called (parenthesis operator) // the deleter instance is however not destroyed, the breakpoint in the destructor does not hit

2 2