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

Re: [Boost-users] Fwd: Boost coroutine/context libraries
by Nat Goodspeed 19 Jan '17

19 Jan '17

On Jan 18, 2017 2:39 PM, "Gregory Laird" <glaird(a)northgatebowl.com> wrote: Question: If I allocate too small of a stack, and something gets written outside of the allocation, is there any way to know that this has occurred? http://www.boost.org/doc/libs/1_63_0/libs/context/doc/html/context/stack/pr…

1 0

1.63.0: diffenre package with the same version
by Ivan Kabaivanov 15 Jan '17

15 Jan '17

Hi, first time poster here. I've been compiling boost from source for about 8 years now. I download the packages from sourceforge.net (https://sourceforge.net/projects/boost/files/boost/1.63.0/) Lately (last few releases) I notice a troubling trend -- the same package, say boost_1_63_0.tar.bz2 will have a different md5 hash if downloaded again a few weeks after being downloaded for the first time. Case in point: boost_1_63_0.tar.bz2, downloaded around Fri Dec 30 17:11:50 2016 +0200 (GMT+2) had md5 hash d17537e28aa2131fa192ce2870ce72a3. The very same file, dowloaded today (January 15, 2017) has a new md5 hash: 1c837ecd990bb022d07e7aab32b09847. What's more troubling is that a cursory comparison of the two files' contents reveals differences. To begin with, the old package had 53599 files, while the new one has 53655. Further, files that are present in both packages have small differences. $ diff -Naur ../../OLD/boost_1_63_0/./libs/numeric/odeint/doc/html/boost/numeric/odeint/base_tag.html ../../NEW/boost_1_63_0/./libs/numeric/odeint/doc/html/boost/numeric/odeint/base_tag.html--- ../../OLD/boost_1_63_0/./libs/numeric/odeint/doc/html/boost/numeric/odeint/base_tag.html 2016-12-29 17:12:21.000000000 +0200+++ ../../NEW/boost_1_63_0/./libs/numeric/odeint/doc/html/boost/numeric/odeint/base_tag.html 2016-12-22 14:47:46.000000000 +0200@@ -7,7 +7,7 @@ <link rel="home" href="../../../index.html" title="Chapter 1. Boost.Numeric.Odeint"> <link rel="up" href="../../../header/boost/numeric/odeint/stepper/stepper_categories_hpp.html" title="Header <boost/numeric/odeint/stepper/stepper_categories.hpp>"> <link rel="prev" href="dense_output_stepper_tag.html" title="Struct dense_output_stepper_tag">-<link rel="next" href="base_tag_stepp_idp64335824.html" title="Struct base_tag<stepper_tag>">+<link rel="next" href="base_tag_stepp_idp34699008.html" title="Struct base_tag<stepper_tag>"> </head> <body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"> <table cellpadding="2" width="100%"><tr>@@ -20,7 +20,7 @@ </tr></table> <hr> <div class="spirit-nav">-<a accesskey="p" href="dense_output_stepper_tag.html"><img src="../../../../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../../../header/boost/numeric/odeint/stepper/stepper_categories_hpp.html"><img src="../../../../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../../../index.html"><img src="../../../../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="base_tag_stepp_idp64335824.html"><img src="../../../../../../../../doc/src/images/next.png" alt="Next"></a>+<a accesskey="p" href="dense_output_stepper_tag.html"><img src="../../../../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../../../header/boost/numeric/odeint/stepper/stepper_categories_hpp.html"><img src="../../../../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../../../index.html"><img src="../../../../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="base_tag_stepp_idp34699008.html"><img src="../../../../../../../../doc/src/images/next.png" alt="Next"></a> </div> <div class="refentry"> <a name="boost.numeric.odeint.base_tag"></a><div class="titlepage"></div>@@ -45,7 +45,7 @@ </tr></table> <hr> <div class="spirit-nav">-<a accesskey="p" href="dense_output_stepper_tag.html"><img src="../../../../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../../../header/boost/numeric/odeint/stepper/stepper_categories_hpp.html"><img src="../../../../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../../../index.html"><img src="../../../../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="base_tag_stepp_idp64335824.html"><img src="../../../../../../../../doc/src/images/next.png" alt="Next"></a>+<a accesskey="p" href="dense_output_stepper_tag.html"><img src="../../../../../../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../../../header/boost/numeric/odeint/stepper/stepper_categories_hpp.html"><img src="../../../../../../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../../../index.html"><img src="../../../../../../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="base_tag_stepp_idp34699008.html"><img src="../../../../../../../../doc/src/images/next.png" alt="Next"></a> </div> </body> </html> I know, these are trivial differences. But they are differences nevertheless and this is not acceptable for package maintainers. I've heard stories of how SourceForge is (or used to be ?) modifying packages to include crapware and sponsored content. So my first reaction was to suspect SF.net of pulling some stupid/shady stuff. But I'm beginning to wonder it it was SF.net really. Please, investigate what's happening and please fix it. As a package maintainer I can no longer vouchsafe for the integrity of the upstream package. Thanks,IvanK.

2 1

Problems Building Boost 1.63 on Solaris 11.3 x86 with Oracle Developer Studio12.5
by Shawn Firth 14 Jan '17

14 Jan '17

Ladies and Gentlemen, We have been using Boost on Windows and HP-UX for over five years. Recently I have been attempting to build Boost 1.63 on Solaris 11.3 x86 with Oracle Developer Studio12.5. I started with: ./bootstrap.sh --prefix=/export/home/me/boost_1_63_0 -libdir=/export/home/me/boost_1_63_0/lib --with-toolset=sun ...and that is where I still am. I have been trying every b2 option/configuration that I can file via Google to get Boost 1.63, both with and without C++11, and I am striking out. The Boost documentation is very terse on Sun builds, however the Boost 1.63 release notes say it has been tested for SunOS: Oracle Solaris Studio 12.5 and Oracle Solaris Studio, C++11 12.5. Here's all that the Boost 1.64 doco has to say about Sun builds: sun Oracle Only very recent versions are known to work well with Boost. Note that the Oracle/Sun compiler has a large number of options which effect binary compatibility: it is vital that the libraries are built with the same options that your appliction will use. In particular be aware that the default standard library may not work well with Boost, unless you are building for C++11. The particular compiler options you need can be injected with the b2 command line options cxxflags=``and ``linkflags=. For example to build with the Apache standard library in C++03 mode use b2 cxxflags=-library=stdcxx4 linkflags=-library=stdcxx4. http://www.boost.org/doc/libs/1_63_0/more/getting_started/unix-variants.html Sun Studio The sun module supports the Sun Studio<http://developers.sun.com/sunstudio/index.jsp> C++ compilers for the Solaris OS. The module is initialized using the following syntax: using sun : [version] : [c++-compile-command] : [compiler options] ; This statement may be repeated several times, if you want to configure several versions of the compiler. If the command is not specified, Boost.Build will search for a binary named CC in /opt/SUNWspro/bin and in PATH. When using this compiler on complex C++ code, such as the Boost C++ library<http://boost.org/>, it is recommended to specify the following options when intializing the sun module: -library=stlport4 -features=tmplife -features=tmplrefstatic See the Sun C++ Frontend Tales<http://blogs.sun.com/sga/entry/command_line_options> for details. The following options can be provided, using <option-name>option-value syntax: cflags Specifies additional compiler flags that will be used when compiling C sources. cxxflags Specifies additional compiler flags that will be used when compiling C++ sources. compileflags Specifies additional compiler flags that will be used when compiling both C and C++ sources. linkflags Specifies additional command line options that will be passed to the linker. Starting with Sun Studio 12, you can create 64-bit applications by using the address-model=64 property. http://www.boost.org/build/doc/html/bbv2/reference/tools.html#bbv2.referenc… Can someone provide me with the user-config.jam and b2 options so that I can successfully build Boost 1.63 on Solaris 11.3 x86 with Oracle Developer Studio12. 3, 4, or 5? Failing user-config.jam and b2 options for Solaris 11 on x86, can you provide me with user-config.jam and b2 options for Solaris 11 on SPARC with Oracle Developer Studio12.3, 4, or 5? Thank you for your time... Shawn Firth RG Technology Senior Systems Architect shawn.firth(a)robbinsgioia.com<mailto:shawn.firth@robbinsgioia.com> [Description: Description: RGG] www.robbinsgioia.com<http://www.robbinsgioia.com/> ________________________________ CONFIDENTIALITY NOTICE: The information contained in this electronic mail and its attachments are privileged and confidential information and are intended for the use of the individual(s) or entity named above and those who have been specifically authorized to receive it. If you are not the intended recipient, you are hereby notified that any use, copying or distribution of this communication is strictly prohibited.

2 4

boost 1.63 fails to build with zlib and bzip2 in different directories
by Andrea Bocci 14 Jan '17

14 Jan '17

Hi, I'm trying to build custom binaries for boost 1.63 using zlib and bzip2 from a local installation rather than the system ones, following the instructions at http://www.boost.org/doc/libs/1_63_0/more/getting_started/unix-variants.htm… and http://www.boost.org/doc/libs/1_63_0/libs/iostreams/doc/installation.html but it looks like the build system does not like having the zlib and bzip2 libraries in different folders. Here's what I'm doing to set up the build environment: # create a common directory for zlib and bzip2 mkdir common # build zlib wget http://downloads.sourceforge.net/project/libpng/zlib/1.2.10/zlib-1.2.10.tar… tar xaf zlib-1.2.10.tar.xz cd zlib-1.2.10/ ./configure make test make install prefix=../common cd .. # build bzip2 wget http://www.bzip.org/1.0.6/bzip2-1.0.6.tar.gz tar xaf bzip2-1.0.6.tar.gz cd bzip2-1.0.6/ make -f Makefile-libbz2_so make test make install PREFIX=../common cp -ar libbz2.so.* ../common/lib/ cd .. # build boost wget https://sourceforge.net/projects/boost/files/boost/1.63.0/boost_1_63_0.tar.… tar xaf boost_1_63_0.tar.bz2 cd boost_1_63_0/ mkdir boost-build cd tools/build/ ./bootstrap.sh gcc ./b2 install --prefix=../../boost-build cd ../../ export PATH=$PWD/boost-build/bin:$PATH Now, if I tell boost to pick zlib and bzip2 from the common directory with b2 -q \ -d+2 \ --build-dir=build \ --stage-dir=stage \ --disable-icu \ --without-atomic \ --without-chrono \ --without-container \ --without-context \ --without-coroutine \ --without-exception \ --without-graph \ --without-graph_parallel \ --without-locale \ --without-log \ --without-math \ --without-mpi \ --without-random \ --without-wave \ toolset=gcc \ link=shared \ threading=multi \ variant=release \ cxxflags="-std=c++11 -O2" \ -sBZIP2_INCLUDE=../common/include \ -sBZIP2_LIBPATH=../common/lib \ -sZLIB_INCLUDE=../common/include \ -sZLIB_LIBPATH=../common/lib \ stage it works just fine. However, if I try to build boost using two separate directories for bip2 and zlib, with b2 -q \ -d+2 \ --build-dir=build \ --stage-dir=stage \ --disable-icu \ --without-atomic \ --without-chrono \ --without-container \ --without-context \ --without-coroutine \ --without-exception \ --without-graph \ --without-graph_parallel \ --without-locale \ --without-log \ --without-math \ --without-mpi \ --without-random \ --without-wave \ toolset=gcc \ link=shared \ threading=multi \ variant=release \ cxxflags="-std=c++11 -O2" \ -sBZIP2_INCLUDE=../bzip2-1.0.6 \ -sBZIP2_LIBPATH=../bzip2-1.0.6 \ -sZLIB_INCLUDE=../zlib-1.2.10 \ -sZLIB_LIBPATH=../zlib-1.2.10 \ stage the build fails with Performing configuration checks - 32-bit : no - 64-bit : yes - arm : no - mips1 : no - power : no - sparc : no - x86 : yes - symlinks supported : yes - C++11 mutex : yes - Boost.Config Feature Check: cxx11_auto_declarations : yes - Boost.Config Feature Check: cxx11_constexpr : yes - Boost.Config Feature Check: cxx11_defaulted_functions : yes - Boost.Config Feature Check: cxx11_final : yes - Boost.Config Feature Check: cxx11_hdr_mutex : yes - Boost.Config Feature Check: cxx11_hdr_tuple : yes - Boost.Config Feature Check: cxx11_lambdas : yes - Boost.Config Feature Check: cxx11_noexcept : yes - Boost.Config Feature Check: cxx11_nullptr : yes - Boost.Config Feature Check: cxx11_rvalue_references : yes - Boost.Config Feature Check: cxx11_template_aliases : yes - Boost.Config Feature Check: cxx11_thread_local : yes - Boost.Config Feature Check: cxx11_variadic_templates : yes - zlib : yes error: at /home/fwyzard/src/boost/boost_1_63_0/tools/build/src/kernel/modules.jam:107 error: Name clash for '<pbuild/boost/bin.v2/standalone/ac/gcc-5.4.1/release/threading-multi>main.o' error: error: Tried to build the target twice, with property sets having error: these incompatible properties: error: error: - <dll-path>../zlib-1.2.10 <library-path>../zlib-1.2.10 <xdll-path>../zlib-1.2.10 error: - <dll-path>../bzip2-1.0.6 <library-path>../bzip2-1.0.6 <xdll-path>../bzip2-1.0.6 error: error: Please make sure to have consistent requirements for these error: properties everywhere in your project, especially for install error: targets. I get a similar error if I specify only one of bzip and bzip2. For example Now, if I tell boost to pick zlib and bzip2 from the common directory with b2 -q \ -d+2 \ --build-dir=build \ --stage-dir=stage \ --disable-icu \ --without-atomic \ --without-chrono \ --without-container \ --without-context \ --without-coroutine \ --without-exception \ --without-graph \ --without-graph_parallel \ --without-locale \ --without-log \ --without-math \ --without-mpi \ --without-random \ --without-wave \ toolset=gcc \ link=shared \ threading=multi \ variant=release \ cxxflags="-std=c++11 -O2" \ -sZLIB_INCLUDE=../common/include \ -sZLIB_LIBPATH=../common/lib \ stage results in Performing configuration checks - 32-bit : no - 64-bit : yes - arm : no - mips1 : no - power : no - sparc : no - x86 : yes - symlinks supported : yes - C++11 mutex : yes - Boost.Config Feature Check: cxx11_auto_declarations : yes - Boost.Config Feature Check: cxx11_constexpr : yes - Boost.Config Feature Check: cxx11_defaulted_functions : yes - Boost.Config Feature Check: cxx11_final : yes - Boost.Config Feature Check: cxx11_hdr_mutex : yes - Boost.Config Feature Check: cxx11_hdr_tuple : yes - Boost.Config Feature Check: cxx11_lambdas : yes - Boost.Config Feature Check: cxx11_noexcept : yes - Boost.Config Feature Check: cxx11_nullptr : yes - Boost.Config Feature Check: cxx11_rvalue_references : yes - Boost.Config Feature Check: cxx11_template_aliases : yes - Boost.Config Feature Check: cxx11_thread_local : yes - Boost.Config Feature Check: cxx11_variadic_templates : yes - zlib : yes error: at /home/fwyzard/src/boost/boost_1_63_0/tools/build/src/kernel/modules.jam:107 error: Name clash for '<pbuild/boost/bin.v2/standalone/ac/gcc-5.4.1/release/threading-multi>main.o' error: error: Tried to build the target twice, with property sets having error: these incompatible properties: error: error: - <dll-path>../common/lib <library-path>../common/lib <xdll-path>../common/lib error: - none error: error: Please make sure to have consistent requirements for these error: properties everywhere in your project, especially for install error: targets. As far as I can tell, this used to work with boost 1.57 - I haven't tried any version in between. Anybody has any suggestions ? Thank you, .Andrea -- Strategy is a system of expedients. Generalfeldmarschall Helmuth Karl Bernhard Graf von Moltke

4 11

Boost 1.63.0 with VS 2017 RC
by csiga.biga＠aol.com 09 Jan '17

09 Jan '17

Hi there! There were attempts at building 1.62.0 with the new v141 toolset of MSVC and some success. Did 1.63 make any progress in this regard? I see that the final release still builds libraries labeled v140 even when using a dev cmd prompt with the new compiler. Cheers, Máté

1 0

msvc-14.0 xp support
by Li Sheng 06 Jan '17

06 Jan '17

Hello Boost community, I'm trying to compile boost *1.61.0* with following command on Windows 10 machine: bootstrap.bat b2 headers b2 -a -j9 --stagedir=./stage/boost-xp *toolset=msvc-11.0,msvc-14.0* threading=multi link=static runtime-link=static address-model=64 *define=_USING_V110_SDK71_ define=BOOST_USE_WINAPI_VERSION=0x0501* -d+2 --debug-building As described above, my chosen compilers are MSVC 11 and 14, and I want to enable xp support on produced libraries. Library files compiled by VC11 work fine, but VC14 files are not. Then I added "--debug-building" to see what's actually going on. My console log shows "Command line free features: _USING_V110_SDK71_, BOOST_USE_WINAPI_VERSION=0x0501" when toolset is "msvc-14.0". This does not occur when toolset is "msvc-11.0". After all, can anyone kindly help me with compile boost on msvc-14.0 with xp support? Please forgive me for asking this weird usage. Warmest regards and gratitude, ZD

1 0

[operators] Move Semantics
by Helmut Zeisel 05 Jan '17

05 Jan '17

1 0

[Python] Custom __getattr__ with different CallPolicy for different return values
by H Krishnan 04 Jan '17

04 Jan '17

Hi, I would like to custom __getattr__ for my class such that for some attributes, an internal reference is returned, and for others, a new object is returned. I cannot hardcode the attributes as they are dynamic. I started writing something like: boost::python::object myGetAttr(boost::python::object obj, std::string const& attr) { MyObj* mobj = boost::python::extract<MyObj*>(obj)(); if (attr == "internal_ref") { MyRef* refObj = mobj->getInternalRef(); // how do I return this? } else { MyValueObj obj = mobj->getValueObj(); // how do I return this? } } class_<MyObj, boost::noncopyable> ("MyObj", no_init) .def ("__getattr__", &myGetAttr, ???what_should_I_put_here???) ; However, I am not able to get this to work. Any suggestions? Regards, Krishnan

1 0

How to reuse system_timer of boost::asio with async_wait?
by Bruce 03 Jan '17

03 Jan '17

Hi, I am doing a program that use asio to construct a timer thread. The thread use asio to async_wait many timers. These timers are constructed by boost::asio::system_timer. These timers are mostly designed to fire only once. When they are fired or canceled, I put them to a free queue to reuse. But when I reuse these times, I called expire_from_now() and async_wait() again on the timer, I got the timer fired immediately with error_code set to "ec == boost::asio::error::operation_aborted". And the timer are not fired after like 5 seconds late as I set to expire_from_now(seconds(5)). How to reuse the timer correctly? My codes basically are as follows: m_timer.cancel(); // I tried to cancel the timer first, but no use, I still get the timer fired at once m_timer.expires_from_now(std::chrono::seconds(nInterval)); m_timer.async_wait(boost::bind(&qmInnerTimer::Fire, this, _1)); Thanks. -Bruce

1 0

Boost.Context: SIGSEGV when switch a thread within a catch block
by Александр Пронченков 02 Jan '17

02 Jan '17

Hi, I get SIGSEGV on macOS when I switch an execution_context from a catch block with one thread and than switch back with another thread. Segmentation fault happens in libc++abi.dylib`__cxa_end_catch Does anyone have any ideas, what it could be? Is it legal to switch thread under an excecution_context while it's within a catch block? A minimal example is here: #include <boost/context/all.hpp> #include <future> using Context = boost::context::execution_context<void>; Context Main(Context context) { try { throw std::runtime_error("123"); } catch (const std::exception& ex) { context = context(); } return context(); } int main() { Context context(&Main); context = context(); std::thread([&] { context = context(); }).join(); return 0; } I see the problem on macOS with clang++ but can't reproduce on a linux with g++. $ uname -ap Darwin air.local 16.3.0 Darwin Kernel Version 16.3.0: Thu Nov 17 20:23:58 PST 2016; root:xnu-3789.31.2~1/RELEASE_X86_64 x86_64 i386 $ clang++ --version Apple LLVM version 8.0.0 (clang-800.0.42.1) Target: x86_64-apple-darwin16.3.0 Thread model: posix InstalledDir: /Library/Developer/CommandLineTools/usr/bin -- Thanks, Alexander Pronchenkov

3 4

[Stacktrace] Review Manager report
by Niall Douglas 02 Jan '17

02 Jan '17

Firstly happy new year to everyone, and many thanks to all those who participated in the peer review of proposed Boost.Stacktrace before Christmas. I received the following formal reviews: 1. Vladimir Batov (YES) 2. Klemens Morgenstern (YES) 3. Andrey Semashev (NO) 4. Nat Goodspeed (YES) 5. Artyom Beilis (NO) 6. Barrett Adair (YES) Extensive comments without a formal acceptance recommendation were also made by: 1. Peter Dimov 2. Edward Diener There was a very noticeable lack of consensus on the strengths and weaknesses of the proposed library which has made summarising all these reviews into a coherent report with concrete conditions for acceptance very difficult. Antony also responded to many of the concerns throughout the review, agreeing with some points, disagreeing with others. I think this lack of consensus stemmed from multiple opinions on what a stacktrace library ought to provide: Minimum viable product: A portable backtrace() and backtrace_symbols() implementation with a C++ STL idiomatic API. The stack backtrace capture implementation must not allocate memory and be "fast". Note that backtrace_symbols() does NOT symbolise into strings containing source file name and line number. It also allocates memory (backtrace_symbols_fd() does not). Opinion 1: Portable serialisation of the stacktrace with instruction pointers converted to offsets into SO/DLL binary paths so a later external process can use those in combination with their debug symbols to retrieve source file names and line numbers. Ideally the serialisation formats supported would include those consumable by well known tooling for this purpose e.g. addr2line or llvm-symbolizer. Opinion 2: That Stacktrace can in process use debug symbols to retrieve source file names and line numbers. This could be implemented by invoking addr2line or equivalent as a child process which has the enormous advantage of being async safe, but has disadvantages in almost every other area. Or it could be implemented by one's own DWARF parser but with the huge caveat that memory could be badly corrupted by now. Opinion 3: That Stacktrace be "async safe" in that backtraces can be safely captured and parsed from a signal handler or SEH/TEH handler callback or point of C++ exception throw. *This implies no memory allocation during the execution of the handler*. >From the reviews, I think there is consensus that the minimum viable product is achieved in Stacktrace though many mentioned that improvements are needed to the documentation and most of the negative feedback was on features not in the minimum viable product described above. I therefore recommend the ACCEPTANCE of proposed Boost.Stacktrace into Boost with the following conditions: 1. That the documentation more clearly set out what Stacktrace does and especially does not do, and that the default facilities provided are those of the Minimum Viable Product described above i.e. no source file nor line number discovery, no external dependencies on libraries not always provided on the target platform, no invocation of child processes, no async safety during stacktrace parsing. Stacktrace ought to be pared back to the minimum viable product featureset with no external dependencies at all. 2. That suitable predefined macros opt in to additional functionality such as serialisation of offset translated backtraces, retrieval of source code path and line number etc. Enabling these may introduce dependencies on third party libraries not always present on a system such as libunwind, or on third party helper utilities such as addr2line. I would recommend that the APIs the use of which are dependent on such third party items are entirely compiled out of Stacktrace when the macro enabling them is not defined. 3. That the documentation describe ONLY the default minimum viable product functionality in its tutorial and quick start. All extra functionality enabled by macros is to be relocated into a separate documentation section away from the main documentation e.g. "Advanced/optional features". 4. That APIs made available only when a given macro has turned them on are documented as such in the API reference docs. I'd like to congratulate Antony on getting yet another library of his into Boost. I hope some of his magic rubs off on me when I bring Outcome to review later this year! Regards, Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

2 1

[log] disabling logging at compile-time
by Igor R 02 Jan '17

02 Jan '17

Is it possible to completely eliminate the logging code (i.e. the run-time overhead) using some macro? Thanks.

2 1

Problems with nesting boost::container::small_vector and C++98
by Daniel Trstenjak 02 Jan '17

02 Jan '17

Hi, first the used versions: boost 1.63.0 gcc 5.4.0 20160609 Ubuntu 16.04.1 the example I've problems with is: #include <map> #include <boost/container/small_vector.hpp> using boost::container::small_vector; typedef std::pair<int, small_vector<float, 8> > mypair; typedef std::map<int, mypair> mymap; typedef std::map<int, small_vector<mypair, 8> > mymap2; typedef std::map<int, small_vector<small_vector<float, 8>, 8> > mymap3; int main() { mymap m; mypair& sv = m[0]; mymap2 m2; small_vector<mypair, 8>& sv2 = m2[0]; mymap3 m3; small_vector<small_vector<float, 8>, 8>& sv3 = m3[0]; } I can compile it without errors with 'g++ -std=c++11', but without '-std=c++11' I'm getting an error for the 'operator[]' on mymap2: dan@octa ~> g++ -Iboost_1_63_0/ small_vector_test.cpp In file included from boost_1_63_0/boost/container/small_vector.hpp:27:0, from small_vector_test.cpp:2: boost_1_63_0/boost/container/vector.hpp: In instantiation of ‘void boost::container::vector<T, Allocator>::assign(FwdIt, FwdIt, typename boost::move_detail::disable_if_or<void, boost::move_detail::is_same<typename boost::container::container_detail::version<Allocator>::type, boost::move_detail::integral_constant<unsigned int, 0u> >, boost::move_detail::is_convertible<InIt, typename boost::container::allocator_traits<Allocator>::size_type>, boost::container::container_detail::is_input_iterator<FwdIt> >::type*) [with FwdIt = boost::container::container_detail::vec_iterator<std::pair<int, boost::container::small_vector<float, 8ul> >*, true>; T = std::pair<int, boost::container::small_vector<float, 8ul> >; Allocator = boost::container::small_vector_allocator<boost::container::new_allocator<std::pair<int, boost::container::small_vector<float, 8ul> > > >; typename boost::move_detail::disable_if_or<void, boost::move_detail::is_same<typename boost::container::container_detail::version<Allocator>::type, boost::move_detail::integral_constant<unsigned int, 0u> >, boost::move_detail::is_convertible<InIt, typename boost::container::allocator_traits<Allocator>::size_type>, boost::container::container_detail::is_input_iterator<FwdIt> >::type = void]’: boost_1_63_0/boost/container/small_vector.hpp:563:7: required from ‘boost::container::small_vector<T, N, Allocator>::small_vector(const boost::container::small_vector<T, N, Allocator>&) [with T = std::pair<int, boost::container::small_vector<float, 8ul> >; long unsigned int N = 8ul; Allocator = boost::container::new_allocator<std::pair<int, boost::container::small_vector<float, 8ul> > >]’ /usr/include/c++/5/bits/stl_pair.h:113:31: required from ‘std::pair<_T1, _T2>::pair(const _T1&, const _T2&) [with _T1 = const int; _T2 = boost::container::small_vector<std::pair<int, boost::container::small_vector<float, 8ul> >, 8ul>]’ /usr/include/c++/5/bits/stl_map.h:487:23: required from ‘std::map<_Key, _Tp, _Compare, _Alloc>::mapped_type& std::map<_Key, _Tp, _Compare, _Alloc>::operator[](const key_type&) [with _Key = int; _Tp = boost::container::small_vector<std::pair<int, boost::container::small_vector<float, 8ul> >, 8ul>; _Compare = std::less<int>; _Alloc = std::allocator<std::pair<const int, boost::container::small_vector<std::pair<int, boost::container::small_vector<float, 8ul> >, 8ul> > >; std::map<_Key, _Tp, _Compare, _Alloc>::mapped_type = boost::container::small_vector<std::pair<int, boost::container::small_vector<float, 8ul> >, 8ul>; std::map<_Key, _Tp, _Compare, _Alloc>::key_type = int]’ small_vector_test.cpp:18:39: required from here boost_1_63_0/boost/container/vector.hpp:1261:15: error: binding ‘const std::pair<int, boost::container::small_vector<float, 8ul> >’ to reference of type ‘std::pair<int, boost::container::small_vector<float, 8ul> >&’ discards qualifiers *cur = *first; ^ In file included from /usr/include/c++/5/bits/stl_algobase.h:64:0, from /usr/include/c++/5/bits/stl_tree.h:63, from /usr/include/c++/5/map:60, from small_vector_test.cpp:1: /usr/include/c++/5/bits/stl_pair.h:96:12: note: initializing argument 1 of ‘std::pair<int, boost::container::small_vector<float, 8ul> >& std::pair<int, boost::container::small_vector<float, 8ul> >::operator=(std::pair<int, boost::container::small_vector<float, 8ul> >&)’ struct pair ^ Any ideas? Is 'boost:container::small_vector' supposed to work without C++11 support? Thanks! Greetings, Daniel

2 4

How to use Boost libs from projects?
by Andy 01 Jan '17

01 Jan '17

I have project compiled as Debug,x64. And boost 1.63 compiled by: - bootstrap.bat - b2 variant=debug,release link=static runtime-link=static address-model=64 In project I have cpp files with #include "boost/filesystem.hpp" and when compile project I have error: cannot open file 'libboost_filesystem-vc140-mt-gd-1_61.lib' whereas in c:\boost\stage\lib\ I have libboost_filesystem-vc140-mt-sgd-1_63.lib but not libboost_filesystem-vc140-mt-gd-1_61.lib and in VS2015 project I have no Boost lib. (sorry for my English)

4 13

How to compile boost-thread for Universal Windows?
by Petr Hudeček 31 Dec '16

31 Dec '16

Hello, I'm trying to use the boost-thread library in a Windows Runtime (UWP) project. The tutorial at https://blogs.msdn.microsoft.com/vcblog/2014/07/18/using-boost-libraries-in… doesn't seem to work anymore. If I compile boost sources, either from the primary distribution, or from https://github.com/wpbest/Boost, I can compile it as a static library but when I link to it from the project, I receive linker errors that four identifiers could not be found. These identifiers are API functions such as CreateWaitableTimer. Is there a way to use boost-thread from a UWP project despite this? Perhaps setting some preprocessor symbols? Thank you, Petr Hudecek

4 5

Echo AsyncReadStream/AsyncWriteStream example
by Adam Zegelin 30 Dec '16

30 Dec '16

Hi. Is there a boost::asio example that shows an implementation of AsyncReadStream/AsyncWriteStream that copies write->read? i.e. echoes what was sent I’m trying to emulate serial device that takes input commands separated by \r, executes them, and then outputs the result. i.e., I'm trying to create a fake asio::serial_port that can read input, process it and then provide output. The device I'm trying to emulate echoes the sent characters. Commands are executed once '\r' is received. Ideally the fake serial port also echos the received data immediately. I've got the read part working by implementing async_write_some(…), but I'm having trouble implementing the async_read_some(…) method so that it can be called at anytime, especially when io_service.run() is run from multiple threads. Thanks, Adam

1 0

Boost 1.63.0 is released!
by Marshall Clow 28 Dec '16

28 Dec '16

Release 1.63.0 of the Boost C++ Libraries is now available. These open-source libraries work well with the C++ Standard Library, and are usable across a broad spectrum of applications. The Boost license encourages both commercial and non-commercial use. This release contains numerous enhancements and bug fixes for existing libraries. For details, including download links, see http://www.boost.org/users/news/version_1.63.0 You can also download directly from SourceForge: https://sourceforge.net/projects/boost/files/boost/1.63.0/ To install this release on your system, see http://www.boost.org/doc/libs/release/more/getting_started/index.html Thanks to everyone who participated in this release. Thanks, --The Boost release team Vladimir Prus Rene Rivera Marshall Clow Eric Niebler Daniel James Beman Dawes

2 1

[msm] Transitions not atomic with orthogonal regions
by Stian Zeljko Vrba 26 Dec '16

26 Dec '16

I have spent almost an hour debugging the following issue. I have a "main" orthogonal region with a couple of states and an "error" region with RUNNING and PAUSED states, like this: struct PAUSED : public msm::front::interrupt_state<new_job> {}; struct RUNNING : public msm::front::state<> {}; There are transitions PAUSED --new_job--> RUNNING and RUNNING --failed--> PAUSED. Then, in one of the main states, I have struct MAIN_STATE1 : public msm::front::state<> { template<typename Event, typename FSM> void operator()(const Event& e, FSM& fsm) { if (error_condition) fsm.process_event(failed{}); } } and in the FSM I have using initial_state = mpl::vector<MAIN_STATE0, PAUSED>; Now, when new_job is processed and MAIN_STATE1::on_entry called, the machine is in (MAIN_STATE1, PAUSED). In this state, process_event(failed{}) is ignored and not queued because PAUSED is interrupt state. Then, the next orthogonal region is processed and the machine goes into (MAIN_STATE1, RUNNING), which destroys the rest of the logic around which the FSM is designed... Adding using active_state_switch_policy = msm::active_state_switch_after_exit; to the machine did not help in any way. Reordering the states in initial_state fixed the problem, but this is a rather unsatisfactory fix because it doesn't solve the underlying problem of transition atomicity. This is counter-intuitive, being on the border between a bug and lack of clear documentation. Specifically: 1. State switch should be atomic. By using active state switch after exit, I had *expected* that first all exit handlers of the current state are called, the state is atomically switched to the new state, and then all entry handlers are called. What I have *observed* is that on_entry handler of a state in one orthogonal region is called before the other orthogonal region has switched *its* state. Thus, with orthogonal regions, the machine can be in an "impossible" intermediate combination of states. Because not all side-effects have run, this has consequences for all actions/guards acting on state or FSM data, which is illustrated by the following: 2. process_event eagerly evaluates is_event_handling_blocked_helper for each orthogonal region in order. The event should be queued unconditionally, and only after the transition has completed, the machine should dequeue events and check whether processing is blocked etc. This, I think, would make at least process_event to appear atomic wrt state switch. Which leads me to conclude that MSM is seductively declarative on the surface, but that it's rather tricky operationally.

1 0

Boost 1.63.0 Release Candidate 2
by Marshall Clow 25 Dec '16

25 Dec '16

The second release candidate for the 1.63.0 release is now available at: https://sourceforge.net/projects/boost/files/boost/1.63.0/ The only changes are in Boost.Config, Boost.Context, and Boost.Fiber. The SHA256 checksums are as follows: 25db3956a8d58187ac7a0702cc917e9bab47ff90baafc35e4e789dca1ce5f423 boost_1_63_0_rc2.7z beae2529f759f6b3bf3f4969a19c2e9d6f0c503edcb2de4a61d1428519fcb3b0 boost_1_63_0_rc2.tar.bz2 fe34a4e119798e10b8cc9e565b3b0284e9fd3977ec8a1b19586ad1dec397088b boost_1_63_0_rc2.tar.gz 97ce6635df1f44653a597343cd5757bb8b6b992beb3720f5fc761e3644bcbe7b boost_1_63_0_rc2.zip As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. -- The release managers

2 2

cross compile application include boost-lib
by Thomas Schmiedl 24 Dec '16

24 Dec '16

Hello, I try to use this application https://github.com/repetier/Repetier-Server on my mips-based router. Sorry, I'm not a developer. Please could someone explain the cross-compiling steps for Linux (I use Ubuntu 14 x86 as build system). gcc/g++ toolchain for mips is from this project http://freetz.org/wiki/help/howtos/common/newbie.en. Do I have also to cross-compile boost for mips or only install the Ubuntu-source-package (sudo apt-get install libboost-all-dev). Thanks for your help and best regards, Thomas

2 2

[boost] [build] [context] Mac build options for Boost.Context
by Nat Goodspeed 23 Dec '16

23 Dec '16

Building Boost 1.57 on OS X 10.11, but building it *for* OS X down to 10.9, we're passing bjam a cxxflags command-line argument including -mmacosx-version-min=10.9. We see that switch (along with our other cxxflags switches) in most "g++" lines in the build output, since we also pass -d2 -d+4. But for the assembly source files in Boost.Context, the lines do not contain our requested -mmacosx-version-min switch. In fact they don't seem to contain any of our cxxflags switches. As a result, the assembly object files are built for the OS X version on the build host, 10.11. At executable link time, this produces messages: ld: warning: object file (packages/lib/Release/libboost_context-mt.a(jump_x86_64_sysv_macho_gas.o)) was built for newer OSX version (10.11) than being linked (10.9) ld: warning: object file (packages/lib/Release/libboost_context-mt.a(make_x86_64_sysv_macho_gas.o)) was built for newer OSX version (10.11) than being linked (10.9) (These are the only two .o files producing such messages.) I'm afraid that the resulting executable won't run on OS X 10.9 or 10.10, and that segment of the population is still interesting to us. How should we propagate the cxxflags=-mmacosx-version-min=10.9 feature to the assembler command for Boost.Context?

2 2

Units: Are scales for bytes really working?
by Akim Demaille 21 Dec '16

21 Dec '16

Hi, I’m trying to run the example given in http://www.boost.org/doc/libs/1_62_0/doc/html/boost_units/Examples.html#boo…, and the result is puzzling. $ cat foo2.cc #include <iostream> #include <boost/units/systems/information/byte.hpp> #include <boost/units/io.hpp> using namespace std; using boost::units::binary_prefix; using boost::units::engineering_prefix; using boost::units::no_prefix; using boost::units::symbol_format; using boost::units::quantity; using boost::units::information::byte_base_unit; int main() { cout << BOOST_VERSION << endl; // Don't forget that the units name or symbol format specification is persistent. cout << symbol_format; // Resets the format to the default symbol format. quantity<byte_base_unit::unit_type> b = 2048. * byte_base_unit::unit_type(); cout << b << endl; cout << engineering_prefix << b << endl; // 2.048 kb cout << symbol_format << binary_prefix << b << endl; // "2 Kib" } $ g++-6 foo2.cc -o foo2 $ ./foo2 106200 2048 B 16.384 kb 16 Kib What am I doing wrong? Note that there are a couple of issues in the example as given: - there should not be any endl after cout << symbol_format; - using ‘\n’ instead of endl, too many people seem to believe that std::endl is the normal way to end the lines - it would be nice to provide all the needed includes

1 0

Re: [Boost-users] [boost] Stacktrace library starts review today 14th Dec
by Antony Polukhin 17 Dec '16

17 Dec '16

2016-12-17 20:51 GMT+03:00 Andrey Semashev <andrey.semashev(a)gmail.com>: <...> > That said, I'd like to thank Antony for the submission and encourage > him to continue his work on the library. The library does try to fill > an important gap. Thank you for the review! I'd like to discuss a proposed changes. Thing that I'm worried about: * replacing class `stacktrace` with vector<void*> will definitely break the most common use case std::cout << get_stacktrace(); as it's not a good idea to overload ostream operator for vector of void pointers * addr2line and forking is not nice. Is linking with a GPL licensed code an acceptable alternative? * removing `basic_stacktrace` will definitely remove all the possibilities for optimization of caching COM/parsed DWARF. This is a quick win that later may turn into big troubles. -- Best regards, Antony Polukhin

1 0

Trouble with std::forward and fusion vector using clang 3.4 and older library - is there a workaround?
by Stan Blakey 16 Dec '16

16 Dec '16

I am trying to compile using clang 3.4 and -std=c++11. With the -std=c++11 it compiles and runs without any issue. With -std=c++11 I get boost_1_60_0\boost/fusion/container/vector/vector.hpp:126:20: error: no template named 'forward' in namespace 'std'; did you mean 'advance_detail::forward'? return std::forward<result>(*advance_c<N>(begin(seq))); and many similar errors. I have tried all the boost BOOST_NO_CXX11_HDR_ macros but not found one that resolves this issue. I don't have any choice about compiler version or library version due to other constraints. Any suggestions appreciated, Stan *#include* <boost/fusion/tuple.hpp> *#include* <boost/fusion/algorithm.hpp> *#include* <string> *#include <iostream>* *using* *namespace* boost::fusion; *struct* print { *template* <*typename* *T*> *void* *operator()*(*const* *T* &t) *const* { std::cout << std::boolalpha << t << '\n'; } }; *int* *main*() { *typedef* tuple<*int*, std::string, *bool*, *double*> tuple_type; tuple_type t(10,"Boost", *true*, 3.14); for_each(t, print()); }

2 4

Boost 1.63.0 Release Candidate 1
by Marshall Clow 15 Dec '16

15 Dec '16

The release candidates for the 1.63.0 release are now available at: https://sourceforge.net/projects/boost/files/boost/1.63.0/ The SHA256 checksums are as follows: 010699561c6011e6d97dc453d55bc5e54cdc192622434eff7b8ac559efac846c boost_1_63_0_rc1.7z d8ebf8d90b2f823f04119445ac339d704395d655907333ec0fab863040119d24 boost_1_63_0_rc1.tar.bz2 dabe29b8a60369ae7c15c20681a32717f3c37d97e5dc6d39c5b2fd0494b8e995 boost_1_63_0_rc1.tar.gz 1683942022fd63c205c48fb2155af6094c67ef67aa13ffb3e1db78fecf817efe boost_1_63_0_rc1.zip As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. -- The release managers

7 12