July 2008 - Boost-users - lists.boost.org

Problems with "to_lower" function on IBM AIX
by Frank Illenseer 03 Jul '08

03 Jul '08

Hello, We are currently trying to compile our code using the BOOST collection on IBM AIX 5.3 (POWER5). But we get errors: Using BOOST 1.33.0 ------------------ xlc++_r -I../../../../../include -q64 -qarch=ppc64 -qtune=balanced -qtls -D_POSIX_SOURCE -D_XOPEN_SOURCE -D_XOPEN_SOURCE_EXTENDED -D_LARGE_FILES -D_ALL_SOURCE -I../../../../../thirdparty/zlib/1.2.3/ibm_aix64/include -D_RECHNER_IBM_AIX64 -D_RECHNER_STR=\"IBM_AIX64\" -I../../../../../thirdparty/Boost/1_33_0 -I../../utilities/hardcore/include -I../../utilities/parser/include -I./include -D_RELEASE -D_GUI_LINUX -O3 -c ./src/common_OptimisationConstraintsParser.cpp "../../../../../thirdparty/Boost/1_33_0/boost/algorithm/string/detail/case_conv.hpp", line 36.44: 1540-0215 (S) The wrong number of arguments have been specified for "tolower(int)". "../../../../../thirdparty/Boost/1_33_0/boost/algorithm/string/detail/case_conv.hpp", line 31.23: 1540-0700 (I) The previous message was produced while processing "boost::algorithm::detail::to_lowerF<char>::operator()(char) const". "../../../../../thirdparty/Boost/1_33_0/boost/iterator/iterator_facade.hpp", line 515.41: 1540-0700 (I) The previous message was produced while processing "boost::iterator_core_access::dereference<boost::transform_iterator<boost::algorithm::detail::to_lowerF<char>,std::_Ptrit<char,long,const char *,const char &,char *,char &>,boost::use_default,boost::use_default> >(const transform_iterator<boost::algorithm::detail::to_lowerF<char>,std::_Ptrit<char,long,const char *,const char &,char *,char &>,boost::use_default,boost::use_default> &)". make[2]: *** [common_OptimisationConstraintsParser.o] Error 1 Then we tried, if maybe upgrading to the latest BOOST will solve, but unfortunately not: Using BOOST 1.35.0 ------------------ xlc++_r -I../../../../../include -q64 -qarch=ppc64 -qtune=balanced -qtls -D_POSIX_SOURCE -D_XOPEN_SOURCE -D_XOPEN_SOURCE_EXTENDED -D_LARGE_FILES -D_ALL_SOURCE -I../../../../../thirdparty/zlib/1.2.3/ibm_aix64/include -D_RECHNER_IBM_AIX64 -D_RECHNER_STR=\"IBM_AIX64\" -I../../../../../thirdparty/Boost/1_35_0 -I../../utilities/hardcore/include -I../../utilities/parser/include -I./include -D_RELEASE -D_GUI_LINUX -O3 -c ./src/common_OptimisationConstraintsParser.cpp "../../../../../thirdparty/Boost/1_35_0/boost/algorithm/string/detail/case_conv.hpp", line 37.50: 1540-0157 (S) The text ">" is unexpected. It may be that this token was intended as a template argument list terminator but the name is not known to be a template. "../../../../../thirdparty/Boost/1_35_0/boost/algorithm/string/detail/case_conv.hpp", line 32.23: 1540-0700 (I) The previous message was produced while processing "boost::algorithm::detail::to_lowerF<char>::operator()(char) const". "../../../../../thirdparty/Boost/1_35_0/boost/iterator/iterator_facade.hpp", line 515.41: 1540-0700 (I) The previous message was produced while processing "boost::iterator_core_access::dereference<boost::transform_iterator<boost::algorithm::detail::to_lowerF<char>,std::_Ptrit<char,long,const char *,const char &,char *,char &>,boost::use_default,boost::use_default> >(const transform_iterator<boost::algorithm::detail::to_lowerF<char>,std::_Ptrit<char,long,const char *,const char &,char *,char &>,boost::use_default,boost::use_default> &)". make[2]: *** [common_OptimisationConstraintsParser.o] Error 1 Can maybe anyone give some hints on how to get this compiling? If any further information is needed, please just state and I wil try to provide. Thanks in advance. Frank

2 1

Compilation Issues with 1.34.1 on gcc 4.3.1 / OpenSUSE 11
by Ruediger Berlich 02 Jul '08

02 Jul '08

Hi there, I recently posted a problem with the compilation of Boost 1.35 with gcc 4.3.1, on an OpenSUSE 11 box - the serialization library did not compile. I understand Robert's comments in a way that these problems will go away in 1.36 . I have thus in the meantime tried to use the current 1.36 trunk version, which fails in some other respects (which is fine, it is beta). OpenSUSE 11 comes with Boost 1.34.1 libraries installed, but does apparently not include the boost_system headers/library files. I have thus tried to revert to 1.34 by compiling my own Boost 1.34.1, but get the result: ...failed updating 180 targets... ...skipped 72 targets... ...updated 1263 targets... Not all Boost libraries built properly. Apart from a problem with bjam in OpenSUSE 10.3, which could be fixed thanks to comments on this list, Boost 1.34.1 compiled like a charm in the past. Now, with OpenSUSE 11 there does not seem to be any version of Boost that I can use. I have heard of similar issues with Gentoo, which also uses gcc 4.3.1, so I assume these are compiler-related problems. Apart from my own post on comp.lib.boost.user, however, there was no other reports on problems here. What am I missing ? Has anyone successfully compiled 1.34.1 (or 1.35, including serialization) on an OpenSUSE 11 box (64 bit) ? Thanks and Best Regards, Ruediger

1 1

[Boost.ASIO] and registering callbacks?
by James Vanns 02 Jul '08

02 Jul '08

Hi, I don't quite understand how to register an async call back once with an IO service? Consider this model: * Server accepts connection * connection object *registers* it's read callback (to be executed whenever there is a read available from the underlying epoll) * connection object is doing some processing or conversing with the main server etc. * epoll (or IO service) says there is data to be read * call back gets executed * the processing resumes (based on the newly received data) I'm prettu sure that is normal. But how do I actually achieve this with ASIO? I mean, I have the main server that runs io_service::run() so it knows when there is data to be read on a socket - that socket is of course associated with this client connection object and it is this object that acts on the data (and may also send of course). However, using the async_read methods I appear to have to keep re-registering the callback in the actual callback: client::register_read_callback() { m_socket.async_read_some(buffer(m_inbound, p_max), boost::bind(&base_io::read_callback, shared_from_this(), placeholders::error, placeholders::bytes_transferred)); } client::read_callback(error_code &ec, size_t bytes) { // I have to call register_read_callback() again here! } client::start() // I get called when I am established between two endpoints { register_read_callback(); do_some_processing(); } client::do_some_processing() { } I'm trying to build a library that transparently buffers messages received from an endpoint. THe library silently receives the data (through the epoll based io service and the asio tcp sockets), builds the messages from the received data and puts them in a stack ready to be poppped off by an application calling pop_msg(). I'm having difficulty in separating these two operations. Regards, Jim Vanns -- James Vanns Systems Programmer Framestore CFC Ltd.

2 1

LibBoost link error when building from source 1.35.0
by Robin 02 Jul '08

02 Jul '08

Hi, I'm encountering a libboost_system-vc80-mt-1_35.lib not found link error after recompiling boost using vc8 sp1. I have set BOOST_ALL_DYN_LINK and I link in the dll version of boost file system like so #pragma comment (lib, "boost_filesystem-vc80-mt-1_35.lib") My steps for building are. 1. Open VC console. 2. Run "bjam --build-type=complete". I tried "bjam --toolset=msvc" too 3. Compile is fine. Dlls and libs are created properly but the libboost_system-vc80-mt-1_35.lib is not in output folder. I noticed it's is built in the link static threading-multi portion but was copied to the output lib folder, so why is the dll version trying to link that? Robin

2 2

Boost support for consoles? (X360, PS3)
by JF Charron 02 Jul '08

02 Jul '08

I am currently developping software for Windows. I'd like to use boost as I have used it before and liked it. However, in the near future, I expect to also develop for the Xbox 360 and the PlayStation 3 consoles. Currently I make different projects when doing non portable Windows stuff. However, if boost can be used on the X360 and PS3, it could be really usefull as I could do, I suppose, pretty portable code between the 3 platforms with minimum changes. So my question is: Is Boost working on Xbox 360 and PlayStation3? Thanks! JFC

6 10

Constructing a time_iterator in a class
by Carlos Alvarez 02 Jul '08

02 Jul '08

Hi, I have a compilation issue with time_iterator from date_time library First, I use this object in a struct, it worked fine typedef struct{ ... time_iterator _iter; ... }Timer; But when I needed to translate my struct into a class to have more functionality, I wasn't able to compile it : class Timer{ public: Timer(); Timer( time_iterator & ); ~Timer(); private: time_iterator _iter; }; ... Timer::Timer( time_iterator &it ) { _iter = it; } I got these errors : In constructor `Timer::Timer()' error: no matching function for call to `boost::date_time::time_itr<boost::posix_time::ptime>::time_itr()'| note: candidates are: boost::date_time::time_itr<boost::posix_time::ptime>::time_itr(const boost::date_time::time_itr<boost::posix_time::ptime>&)| note: boost::date_time::time_itr<time_type>::time_itr(time_type, typename time_type::time_duration_type) [with time_type = boost::posix_time::ptime] It seems to mean that a cannot construct a time_iterator without parameters... But I don't know how to fix that.

2 2

bjam, mingw and cross-compiling
by Paul 02 Jul '08

02 Jul '08

Hi, I find the bjam documentation and configuration files very frustrating... it seems to almost do what you want, but it doesn't finish the job and tell you how to actually do it correctly. I have been trying to cross-compile boost 1.35.0 for windows, on a linux host. The gcc.jam file mentions cross-compiling, but does not mention how to get it done. I almost got it to work with a user-config.jam: using gcc : mingw : /home/paul/mingw/bin/mingw32-g++ : <flavor>mingw32 <root>"/home/paul/mingwj/" ; and then hacking away at the gcc.jam file, which checks the os.system variable to decide how to link object files and use the -fPIC flag... around line 703: if [ os.on-windows ] change to... if [ modules.peek : UNIX ] to force it to use -mthreads instead of -pthreads if [ os.name ] != CYGWIN && [ os.name ] != NT change to if [ os.name ] = CYGWIN && [ os.name ] != NT to correct -fPIC it almost did the right thing, except for the extensions... i assume if i hack these lines, I'll get what i want... type.set-generated-target-suffix STATIC_LIB : <toolset>gcc <target-os>cygwin : a ; type.set-generated-target-suffix IMPORT_LIB : <toolset>gcc <target-os>cygwin : dll.a ; type.set-generated-target-prefix IMPORT_LIB : <toolset>gcc <target-os>cygwin : lib ; BUT then I'll have to fight to use the mingw32-ar and mingw32-windres and other programs. Isn't there an easier way? *please* I'm giving up on 1.35 as I've just read about the serialization vector bug, so I'll attempt to use 1.34.1 thanks, Paul

6 10

python example build newbie problem
by David Philp 02 Jul '08

02 Jul '08

Hi I'm hoping this is a reasonable place to ask the question but please point me elsewhere otherwise. I'm trying and failing to build the Boost Python examples. In the directory boost-1_35/libs/python/example/tutorial I am running "bjam". It dies with the following errors: /sw/lib/boost-1_35/tools/build/v2/build/project.jam:266: in find- jamfile from module project error: Unable to load Jamfile. error: Could not find a Jamfile in directory '../../../..'. error: Attempted to find it with pattern '[Bb]uild.jam [Jj]amfile.v2 [Jj]amfile [Jj]amfile.jam'. Here's my configuration: * Intel Mac OS X, 10.5.3 * boost 1.35 * installation is via fink unstable, everything is up to date * I have made unrelated changes (I think) to site-config.jam, in my home directory * I have put the following into my user-config.jam: using darwin ; using python ; bjam --debug-configuration indicates that python is happily found. I'm inclined to suspect that the boost fink package is incorrectly or incompletely configured, but I can't figure out why bjam (correctly configured or not) would look for a Jamfile in that directory. Thanks in advance... David

3 4

[Asio] Buffer size on async_receive_from
by Ryan McConnehey 01 Jul '08

01 Jul '08

All, What happens if the "async_receive_from" receives data that's larger than the buffer size that it was pasted. In Windows, if the data coming in the socket is larger than the buffer than the data is discarded. Is this how "async_receive_from" handles it? Ryan

2 1

Re: [Boost-users] [Asio] async_receive_from
by Igor R 01 Jul '08

01 Jul '08

You can use tcp::socket::cancel(), but note that on most Windows platforms it won't work. http://www.boost.org/doc/libs/1_35_0/doc/html/boost_asio/reference/basic_st… On the other hand, you can just close() your socket - this cancels any outstanding operations on it. 2008/7/1, Ryan McConnehey <mccorywork(a)gmail.com>: > All, > > Async_receive_from waits on data from a socket. I have this socket in > another thread with the sole function of listening to the socket, putting > it's data into a queue and then going back to listening on the socket. Is > there a way to end the function early, while it's still waiting for data, so > I can destroy the socket? > > Ryan > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

1 0