- Boost - lists.boost.org

[thread] Interrupt callback?
by Preston A. Elder 26 Aug '08

26 Aug '08

I want to be able to get an interruption callback (as, say a boost::function) for the current thread. Here is the deal - I have a boost::thread_group, now each thread is creating a series of messages. This message passed from thread to thread, and as it passes to a new thread, each thread needs to put into the message a callback function that can be used to interrupt the thread that owns said message. This is mainly so that if I want to do something with the message mid-'flight' (the request will come from a different thread processing the message), and it happens to be waiting on a condition, I can interrupt that thread and handle the change request on the message. Here is the problem - there is no way to get the callback function from within the thread itself, or indeed. The best I can do is create the thread then set the function in some kind of data structure AFTER the thread has been created, and use some kind of condition or something to 'hold' the new thread until this function has been set. This is kind of dodgy imho. It also requires me to keep a secondary structure pretaining to a thread outside the thread_group. At which point I may as well do the thread_group by hand (which, given the simplicity of thread_group is not a huge amount of code, but still). So, is there a solution that will satisfy the code poet in me? or do I just roll my own thread group? PreZ :)

1 0

[filesystem] Linux copy_file api with wpath
by Joshua Peterson 26 Aug '08

26 Aug '08

It seems that Boost.FileSystem does not implement some methods accepting wpath arguments when compiled with GCC 3.4.3 on Linux. This occurs because BOOST_FILESYSTEM_NARROW_ONLY is defined, which according to the comment in the code occurs for "badly broken compilers or libraries". Ideally, I would like to use the boost::filesystem::copy_file() method on both Windows and Linux, passing wpath arguments in both cases. Is the lack of wpath support simply a problem with GCC 3.4.3, or is wpath never supported for copy_file() with GCC? I can see from the implementation of boost::filesystem::detail::copy_file_api() that open(), read(), and write() are used to copy the file. Since open() accepts only a const char* this may be the limitation. If this is the case, has there been any discussion of providing a copy_file() method which accepts wpath arguments and converts from wchar_t to char before calling copy_file_api(). I do not have a lot of experience with locales and character encodings, so I may be missing something. If this has already been discussed on this mailing list or elsewhere, please point me to the correct discussion, as I have been unable to find it. Thanks. Josh -- Rejoice always. Pray without ceasing. In all circumstances give thanks, for this is the will of God for you in Christ Jesus. -- 1 Thes. 5:16-18 NAB -- I am the living bread that came down from heaven; whoever eats this bread will live forever; and the bread that I will give is my flesh for the life of the world. -- John 6:51 NAB

2 6

[function_types] member_ccs / GCC test annotations
by David Abrahams 25 Aug '08

25 Aug '08

With x86 linux gcc-4.0.1, I get this for the member_ccs test: member_ccs.cpp:17:5: error: #error "test not supported with this compiler/platform" and the test is annotated thusly: Not all compilers/platforms implement nonstandard calling conventions. With GCC/x86 this failure reflects http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29328 . But the nonme the nonmember_ccs test *also* fails with the same error, and this annotation: Not all compilers/platforms implement nonstandard calling conventions. i.e. there is no reference to a GCC bug. Points: * The error message for member_ccs is inconsistent with the note about the gcc bug. * The function types library documentation barely mentions "calling conventions" and doesn't explain what "nonstandard" means in this context. For example, does extern "C" count? * Since I am running x86/GCC, the note about the GCC bug much more specific than the other part of the note, so I assumed that it applied. That would mean that my platform implements a nonstandard calling convention, but a compiler bug is causing the test to fail. * However, that interpretation would be inconsistent with the annotation for nonmember_ccs. * I suspect that my platform does not support nonstandard calling convention, and the bug only applies on platforms where GCC supports nonstandard calling conventions (e.g. Windows), and does not apply to me. If so, the annotation should be more specific about the platforms on which the note applies. * The error message should be better, e.g.: "this platform does not support nonstandard calling conventions; the test is irrelevant here" * A test that always succeeds when the platform doesn't support nonstandard calling conventions would yield far more reliable results than the overly general wildcard matching of failure annotations you have now, which makes a failure on any platform look as though it were expected in the testing chart. -- Dave Abrahams BoostPro Computing http://www.boostpro.com

1 0

Problems with boost 1.36 and serialization library with VS2005 SP1
by Hansi 25 Aug '08

25 Aug '08

Hello, at the moment I wont to switch from boost 1.34.1 to 1.36. But now I have a compile error witch depends from the serialization error. I have a managed application and for that there is also a definition array in the cli namespace. The new version of boost seems to define somewhere the namespace boost globally and so it is not anymore possilbe to compile the application. I have searched where this problem exists but I don't have found the header... I have found only that the problem exists when this 2 headeres are included #include <boost/serialization/detail/shared_ptr_132.hpp> #include <boost/archive/basic_binary_iprimitive.hpp> Has anyone an idea how I can solve the problem? Best regards Hansjörg

3 5

Cannot compile the debug version.
by john chen 25 Aug '08

25 Aug '08

Dear All, I wouldn't be able to compile the debug version of boost_1.360 (cannot find *-gd* under boost/lib). I tried following bjam arguments: bjam ^ --prefix=c:\boost install ^ --build-dir=c:\boost ^ --sBUILD=debug ^ --toolset=msvc ^ --threading=multi ^ --link=static and bjam ^ --prefix=c:\boost install ^ --build-dir=c:\boost ^ --sBUILD=debug release ^ --toolset=msvc ^ and it still doesn't build the debug version. My environment: ms vista 32 ms vc9 sp1 boost 1.360 bjam 03.1.17 Can someone tell me how to get the debug version? Thanks very much. John

2 4

[math][libs] Header-only status wrong for Boost.Math
by Daryle Walker 25 Aug '08

25 Aug '08

Looking at <http://www.boost.org/doc/libs>, I see 14 entries using separate source, but I have 15 source directories still in my all- Boost library project. It seems that the Special Function section of the Math library has separate source for the (optional) C99 and TR1 math functions, but the library page doesn't list that. (The source files are mentioned in the Math-SpecFunc docs.) Can that be changed, please? -- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT hotmail DOT com

3 4

[test] More on formatting log entries for XCode (by default when applicable)
by Daryle Walker 25 Aug '08

25 Aug '08

This is based on the <https://svn.boost.org/trac/boost/ticket/2059> ticket I entered a couple months ago. I just tried out the actual change needed and attached it to the ticket. (I just made a project that compiles all the mandatory Boost source code, like the unit-test framework, into various static libraries I can include in my other projects without repeating the compilation instructions [or wasting disk space]. Centralization is fun.) I still don't know if my __APPLE_CC__ preprocessor test is the official way to check for Xcode, or any compatible, setup. I heard that the current code is supposedly formatted for Visual C++'s error message handling. If so, and the class I changed is _not_ where the per-compiler-product changes are supposed to go, then move the code as needed. (I guess there would be parallel visual-c++ and xcode default formatting classes.) -- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT hotmail DOT com

2 1

Re: [boost] New libraries should provide forward declarations? (Benoit))
by rtsweng-sw＠yahoo.com 25 Aug '08

25 Aug '08

>>>That would make it possible to easily separate the compilation of >>>boost classes (or at least their methods) from that of user-defined >>>classes. >> No, unfortunately, it wouldn't, because you have to pull in the >> definitions to use the template class, what you want is precompiled >> headers. >This is not exactly true. I already use such an approach in my own >developments. >The idea is to use concepts to make sure that the right member >functions are called inside a cpp file and therefore will be compiled >and usable everywhere.That would mean pulling in all related >implementation details... But only for one compilation unit (created >with that purpose in mind). Other compilation units would only have >access to template class definitions. For that matter, i personally >use the .h extension for template class definitions and .hpp for >template class method definitions (but that's only an example of what >could be done). >Note that i agree that it would also bring a readability benefit to >the source code. >BenoÃ®t This sounds exactly like what I am trying to do, I already split the declarations into .h files and the definitions into .hpp. Can you give a brief example of using a single .cpp file for compilation? Is this done like MSVC stdafx files? Thanks, Robert ________________________________

4 3

[Release] Supported platforms are...?
by Gennaro Prota 24 Aug '08

24 Aug '08

In <http://www.boost.org/development/tests/release/developer/summary.html> I just see seven platform/compiler combinations. Are these the only ones officially supported for release(s)? -- Genny

2 2

Lag of trunk tests?
by Gennaro Prota 24 Aug '08

24 Aug '08

I'm watching the trunk regression summary for several days and almost all the tests are reportedly stale at 19-21 August. If I go, for instance, to <http://www.boost.org/development/tests/trunk/Sandia-sun.html> it says that tests are run *nightly* (incidentally, it also doesn't tell anything about the gcc invocation). What's wrong? -- Genny

2 1