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

Aliased parameters in boost::bind
by Evgeny Fraimovitch 30 Mar '17

30 Mar '17

Hello, The tutorial for Boost.Bind library contains the following sample code: bind can handle functions with more than two arguments, and its argument substitution mechanism is more general: bind(f, _2, _1)(x, y); // f(y, x) bind(g, _1, 9, _1)(x); // g(x, 9, x) bind(g, _3, _3, _3)(x, y, z); // g(z, z, z) bind(g, _1, _1, _1)(x, y, z); // g(x, x, x) As can be seen, all but the first of those samples alias the placeholder parameters, substituting a single passed argument several times. What is not mentioned in the tutorial (http://www.boost.org/doc/libs/1_63_0/libs/bind/doc/html/bind.html#bind.purp ose.using_bind_with_functions_and_fu) is the fact that such aliasing for any parameter with nontrivial move behavior is unexpected and possibly dangerous. For instance, given the function: void f(std::string x, std::string y, std::string z) { std::cout << x << y << z; } One could following the tutorial naively write: auto triple = boost::bind(f, _1, _1, _1); and this would even work for: std::string x = "a pretty long string"; triple(x); printing the string thrice, as expected. However the result of triple(std::string("a pretty long string")) would be the string only printed once, which is hardly what the user would expect. In fact, any bind expression produced with aliasing would be sensitive to the type of the reference passed, violating the functor abstraction badly. So, the question is: 1. This behavior seems to match the std::bind behavior documented in the standard. However, would it not be wise to warn against parameter aliasing in the tutorial then? (instead of demonstrating it as one of the first samples) 2. Maybe we can actually check if a parameter is aliased in boost::bind and copy it if it is - I have not tried it, but I think it can be done. Would it be a more sensible thing to do? The sample above might seem contrived, but I've actually encountered this bug in production code when porting it to boost 1.63 (the boost.signals code implicitly does boost::forward, so it was not outwardly apparent what was happening) Sincerely yours, Evgeny

3 6

[asio] sslv3 handshake
by Jim Gallogly 30 Mar '17

30 Mar '17

I want to do a simple C++ web get similar to what is done by this curl command. I must use boost 1.49 > curl https://mysite.dev/api/v1/search?q=test -k --cert > C:\work\testCert.pem The server is requiring the client certificate. I started by using this as an example http://www.boost.org/doc/libs/1_49_0/doc/html/boost_asio/example/ssl/client… and I added modifications by adding calls to the context like ctx.set_options(boost::asio::ssl::context::default_workarounds); ctx.use_certificate_file("C:\\work\\testCert.pem", boost::asio::ssl::context_base::pem); My Request Looks like this: > GET /api/v1/search?q=test HTTP/1.0<br> > Host: mysite.dev <br> > Accept: \*/* but I keep getting messages like this > Error: sslv3 alert handshake failure Does any one know what other steps I need to do? Is this possible in boost 1.49? Thanks, Jim

3 3

[release] Boost 1.64.0 Beta 2
by Rene Rivera 30 Mar '17

30 Mar '17

Boost release 1.64.0 beta 2 is now available at: <https://dl.bintray.com/boostorg/beta/1.64.0.beta.2/source/> The SHA256 checksums are as follows: 49c9339b9d8e7b2610ef84c80cd10b2252c119ebcc3222450beb0e94c1da1522 boost_1_64_0_b2.7z 92ec7cbf1e77c5979e381dd8104f9e96c0471c276e566272751a981d806db94b boost_1_64_0_b2.tar.bz2 da5670892b7b572808491b13a7b9421c3c18e6f395d6a5ec822466db466eda01 boost_1_64_0_b2.tar.gz 61d6e29e8a1d7503fe13a9d916842f0ec2e625f733b98b29ce563f8cbdb482bf boost_1_64_0_b2.zip The notable difference from Beta 1 is the inclusion of the missing Boost Process library. For details of what's in the release, see < http://www.boost.org/users/history/version_1_64_0.html>. Note that the links to files on this web page are for the final release - use the link above to get the beta files. There are also pre-built libraries provided by Tom Kent here: <https://dl.bintray.com/boostorg/beta/1.64.0.beta.2/binaries/> Please download the beta, give it a try, and report any problems you encounter. Thanks, -- The Boost Release Team -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

2 1

[statechart] process_event vs post_event confusion
by Robert Dailey 30 Mar '17

30 Mar '17

I'm seeing a usability issue between post_event and process_event. Basically, depending on where in the state itself I'm executing an event, I need to know which to invoke. My understanding is that if I'm calling within the context of the constructor of the state, I need to use post_event. Anywhere else (outside of construction, but still within a member function of the state subclass itself), you invoke "context<MyMachine>().process_event()". Why can't post_event() automatically query the state machine to see if it is "busy" constructing a state, and if it isn't, forward the event to process_event() for me instead? Is there really a reason to have to make a logical decision between the two in code? I'm using boost::statechart in my code at $DAYJOB and this is a reasonable complaint I get in my code reviews, especially from those not very familiar with boost itself. They have an expectation for the implementation of the state machine to be intuitive and easy to understand.

2 2

Re: [Boost-users] [boost] Going insane trying to build boost for android
by Robert Dailey 29 Mar '17

29 Mar '17

On Wed, Mar 29, 2017 at 2:16 PM, John Maddock <jz.maddock(a)googlemail.com> wrote: > >> >> I'm still not able to build boost though. I did try switching to API >> 21 but I get a different set of errors: >> >> >> In file included from libs\atomic\src\lockpool.cpp:16: >> In file included from .\boost/config.hpp:57: >> In file included from .\boost/config/platform/linux.hpp:74: >> In file included from .\boost/config/posix_features.hpp:18: >> In file included from >> C:/android/ndk/platforms/android-21/arch-arm/usr/include\unistd.h:34: >> In file included from >> C:/android/ndk/platforms/android-21/arch-arm/usr/include\sys/select.h:35: >> C:/android/ndk/platforms/android-21/arch-arm/usr/include\signal.h:86:18: >> error: expected ';' at end of declaration list >> sighandler_t sa_handler; >> ^ >> >> C:/android/ndk/platforms/android-21/arch-arm/usr/include\asm/signal.h:101:22: >> note: expanded from macro 'sa_handler' >> #define sa_handler _u._sa_handler >> ^ > > > OK, so you suckered me into this.... just installed Android NDK and it's a > quoting problem, but what the issue is I'm not sure. For example b2 invokes > the compiler inside quotes like this: > > "clang++" <options> file.cpp > > Which produces: > > 'M:\data\boost\boost\\clang38++.exe' is not recognized as an internal or > external command, > operable program or batch file. > > So then you go into clang++cmd and maybe edit it so the %~dp0\ prefix is > removed from the clang38 invocation (that's what I did anyway). Now you > get: > > .\boost/config/stdlib/libstdcpp3.hpp:78:12: fatal error: 'unistd.h' file not > found > # include <unistd.h> > > Even though if you search the toolset it clearly does have several copies of > unistd.h there. > > And I'm wondering if you tried to fix that by editing the .cmd file and it's > further messed up clang's include paths? > > Note that at any time, invoking clang via: > > clang++ <options> file.cpp > > Will work just fine. > > So in effect Boost.Build has been broken by trying to do the right thing and > always quote the compiler invocation (and I'm pretty sure not quoting would > break a bunch of other toolsets). > > So... which libraries do you require.... since most of them are "just a > bunch of source files" the easiest solution would be to build what you need > manually with something like: > > clang++ -c -I. -O3 -shared libs/boost-library-you-want/src/*.cpp > mylibraryname.so > > which worked for me with a cursory test on Boost.System and Regex > > Hopefully someone can suggest a better long term fix, John. +boost again Yeah, you are running into the exact same set of issues I am. First I ran into the %~dp0 issue, removed that, then the next thing is it can't find any files. I started using the boost cmake build process here: https://github.com/Orphis/boost-cmake This works perfectly once I specify a toolchain file to CMake for the android NDK. Note that this solution means boost builds with the rest of my targets, which isn't ideal, but at least boost builds properly... I really hate bjam/b2 with a passion. For as great as boost is as a library, it's completely out of left field that they deal with this garbage build system.

1 0

boost::asio::asyn_accept handler
by Christopher Pisz 29 Mar '17

29 Mar '17

I am looking at the example source at http://www.boost.org/doc/libs/1_62_0/doc/html/boost_asio/tutorial/tutdaytim… It passes their own connection class to the accept handler as an argument: acceptor_.async_accept(new_connection->socket(), boost::bind(&tcp_server::handle_accept, this, new_connection, boost::asio::placeholders::error)); I tried to implement a class method that would do the same: void ServerSocketASIO::Listen(boost::asio::ip::tcp::acceptor & acceptor, std::function<void(ServerSocketASIO *, const boost::system::error_code &)> callback) { acceptor.async_accept(m_socket, callback); m_connectionState = LISTENING; } but when I compile, I get errors that say AcceptHandler type requirements not met If I implement it this way: void ServerSocketASIO::Listen(boost::asio::ip::tcp::acceptor & acceptor, std::function<void(const boost::system::error_code &)> callback) { acceptor.async_accept(m_socket, callback); m_connectionState = LISTENING; } It compiles, but I would need the actual connection to put into a collection of some kind, in order to keep track of connected clients. The documentation I found at http://www.boost.org/doc/libs/1_62_0/doc/html/boost_asio/reference/AcceptHa… Only shows the accept callback with a boost::system::error_code argument. So, my question is, is the example source wrong? and How do I pass my own connection class to the callback? -- View this message in context: http://boost.2283326.n4.nabble.com/boost-asio-asyn-accept-handler-tp4693107… Sent from the Boost - Users mailing list archive at Nabble.com.

4 13

[asio] Using ASIO to connect to https: using client certificate
by Jim Gallogly 29 Mar '17

29 Mar '17

I want to do a simple C++ web get, similar to what is done by this curl command. I must use boost 1.49 curl https://mysite.dev -k --cert C:\work\testCert.pem The server is requiring the client certificate. I started by using this example http://www.boost.org/doc/libs/1_49_0/doc/html/boost_asio/example/ssl/client… and I added modifications by adding calls to the context like ctx.set_verify_mode(boost::asio::ssl::verify_peer); ctx.set_options(boost::asio::ssl::context::default_workarounds); ctx.use_certificate_file("C:\\work\\testCert.pem", boost::asio::ssl::context_base::pem); ctx.use_private_key_file("C:\\work\\testKey.pem", boost::asio::ssl::context_base::pem); but I keep getting messages like Connect failed: No connection could be made because the target machine actively refused it Any help or direction would be very appreciated, Thanks, Jim

1 0

[crc] Current status of this library
by Bill Somerville 29 Mar '17

29 Mar '17

Hi all, I have a compile error using the Boost 1.63.0 CRC library with the g++ v6.3 compiler with the following usage: #include <boost/crc.hpp> extern "C" { short crc12 (unsigned char const * data, int length); bool crc12_check (unsigned char const * data, int length); } // assumes CRC is last 16 bits of the data and is set to zero // caller should assign the returned CRC into the message in big endian byte order short crc12 (unsigned char const * data, int length) { return boost::augmented_crc<12, 0xc06> (data, length); } bool crc12_check (unsigned char const * data, int length) { return !boost::augmented_crc<12, 0xc06> (data, length); } The error I get with g++ v6.3 is: In file included from /home/bill/wsjtx-prefix/src/boost/boost/config.hpp:61:0, from /home/bill/wsjtx-prefix/src/boost/boost/crc.hpp:12, from /home/bill/wsjtx-prefix/src/lib/crc12.cpp:1: /home/bill/wsjtx-prefix/src/boost/boost/crc.hpp: In instantiation of ‘const least boost::detail::mask_uint_t<12ul>::sig_bits’: /home/bill/wsjtx-prefix/src/boost/boost/crc.hpp:356:9: required from ‘const fast boost::detail::mask_uint_t<12ul>::sig_bits_fast’ /home/bill/wsjtx-prefix/src/boost/boost/crc.hpp:1075:16: required from ‘typename boost::uint_t<Bits>::fast boost::augmented_crc(const void*, std::size_t, typename boost::uint_t<Bits>::fast) [with long unsigned int Bits = 12ul; typename boost::uint_t<Bits>::fast TruncPoly = 3078u; typename boost::uint_t<Bits>::fast = short unsigned int; std::size_t = long unsigned int]’ /home/bill/wsjtx-prefix/src/boost/boost/crc.hpp:1092:41: required from ‘typename boost::uint_t<Bits>::fast boost::augmented_crc(const void*, std::size_t) [with long unsigned int Bits = 12ul; typename boost::uint_t<Bits>::fast TruncPoly = 3078u; typename boost::uint_t<Bits>::fast = short unsigned int; std::size_t = long unsigned int]’ /home/bill/wsjtx-prefix/src/lib/crc12.cpp:14:57: required from here /home/bill/wsjtx-prefix/src/boost/boost/crc.hpp:350:69: warning: left shift of negative value [-Wshift-negative-value] BOOST_STATIC_CONSTANT( least, sig_bits = (~( ~(least( 0u )) << Bits )) ); ~~~~~~~~~~~~~~~~~^~~~~~~~~ /home/bill/wsjtx-prefix/src/boost/boost/crc.hpp:350:69: error: left operand of shift expression ‘(-1 << 12ul)’ is negative [-fpermissive] I note that the develop branch of the CRC library has many changes since Boost 1.63.0 but none are in Boost 1.64.0-beta1. So far I have only tried the new library (devlop branch) with my code and it compiles ok but I need to test it further. Does anyone know why the latest CRC library is not in 1.64 and also does anyone know if the develop branch code is ok and usable by substitution into a 1.63 Boost? Regards Bill Somerville.

1 0

Re: [Boost-users] [boost] Going insane trying to build boost for android
by Robert Dailey 28 Mar '17

28 Mar '17

On Tue, Mar 28, 2017 at 3:10 AM, John Maddock <jz.maddock(a)googlemail.com> wrote: > > Right, but that built with g++ not clang++. Doh!!! I didn't even notice this. Crap. I'll fix it when I get home and try again. > I don't think Boost.Build does anything "different" here: it basically just > invokes clang++. > > Note that clang is not particular "bright" in the way it creates it's > wrapper headers around <cstdlib> etc. Last time I had an issue like this, > the missing symbol was present in the system headers, but guarded by a > #ifdef that clang didn't define (where as GCC did)..... OK had a quick > Google, and that symbol is protected by a clang "availability" macro: > > longdoublestrtold(constchar*,char**)__INTRODUCED_IN(21); > > So it seems to me that this error will be generated by any attempt to > compile with clang unless you specify some "magic" command line option to > tell it to use the latest android API? No idea what that would be, but > perhaps you do? > > Hopefully getting closer, John. Thanks for the help. Looks like it is expecting API 21 to be used. Which is strange because the NDK documentation says to use the minimum supported API you intend to use. Which is 15 for me. But just grepping the NDK source directories, I see that certain files/symbols are only defined starting at API 21. I think a lot of the C library headers are provided by the target API directory, and they change it with each version. I'll try 21 and see if I get different results, and hopefully it won't break when running on Android Ice Cream Sandwich (API 15)

1 1

Re: [Boost-users] [boost] Going insane trying to build boost for android
by Robert Dailey 28 Mar '17

28 Mar '17

On Sun, Mar 26, 2017 at 12:04 PM, John Maddock <jz.maddock(a)googlemail.com> wrote: > >> So outside of what I'm already doing, how do I properly tell boost >> about C++14? I'm sending down a cxxflag for it in my jam file right >> now, but that's the only way I've been able to find out how to do it >> from my Google searching. > > > Nod. As far as I know that's the only way, two questions: > > * If you build something with the -d2 option do you see all the -std= > compiler flags you expect to see? > * If you do so, can you build and *run* libs/config/test/config_info and > report the output? (Re-included the boost ML) With the -d2 option added, this is what I see: clang-linux.compile.c++.without-pth bin.v2\libs\atomic\build\clang-linux-android\release\link-static\target-os-android\threadapi-pthread\threading-multi\lockpool.o "C:/android/ndk/toolchains/llvm/prebuilt/windows-x86_64/bin/clang++" -c -x c++ -O3 -Wno-inline -Wall --sysroot=C:/android/ndk/platforms/android-15/arch-arm -IC:/android/ndk/sources/cxx-stl/llvm-libc++abi/include -IC:/android/ndk/sources/cxx-stl/llvm-libc++/include -IC:/android/ndk/sources/android/support/include -std=c++14 -fno-strict-aliasing -D__GLIBC__ -DANDROID -D__ANDROID__ -D_GLIBCXX__PTHREADS -D__arm__ -D_REENTRANT -O2 -DNDEBUG -DBOOST_ALL_NO_LIB=1 -DBOOST_ATOMIC_SOURCE -DBOOST_ATOMIC_STATIC_LINK=1 -DNDEBUG -I"." -o "bin.v2\libs\atomic\build\clang-linux-android\release\link-static\target-os-android\threadapi-pthread\threading-multi\lockpool.o" "libs\atomic\src\lockpool.cpp" In file included from libs\atomic\src\lockpool.cpp:16: In file included from .\boost/config.hpp:57: In file included from .\boost/config/platform/linux.hpp:15: C:/android/ndk/sources/cxx-stl/llvm-libc++/include\cstdlib:108:9: error: no member named 'strtold' in the global namespace using ::strtold; ~~^ C:/android/ndk/sources/cxx-stl/llvm-libc++/include\cstdlib:126:9: error: no member named '_Exit' in the global namespace using ::_Exit; ~~^ C:/android/ndk/sources/cxx-stl/llvm-libc++/include\cstdlib:149:9: error: no member named 'at_quick_exit' in the global namespace using ::at_quick_exit; ~~^ C:/android/ndk/sources/cxx-stl/llvm-libc++/include\cstdlib:150:9: error: no member named 'quick_exit' in the global namespace using ::quick_exit; ~~^ 4 errors generated. I do see -std=c++14 in there. As for your second question, why would building and running config_info help? I'm cross compiling, so I can't run this on Windows. If you could provide explicit instructions so I properly understand what you want me to do, that would be ideal. Thanks.

4 5