June 2009 - Boost-users - lists.preview.boost.org

[bind maybe?] find_if of map mapped_type.
by Robert Jones 04 Jun '09

04 Jun '09

If I'm iterating over a map with, say find_if, and my predicate function accepts the mapped_type of the map, is there a mechanism to pick out the pair.second value inline? class C; map< int, C > m; bool pred( const C & ); // correct version of this intent. find_if( m.begin( ), m.end( ), pred( second ) ); Thanks, - Rob.

2 2

dev_poll_reactor vs. strand and async_write
by Chris Hite 04 Jun '09

04 Jun '09

In a project using asio I noticed performance issues on Sun vs. Linux. It turned out that dev_poll_reactor is doing tones of system calls always pairs of write() and ioctl(). Stranger still is that the calls to write() are for 0 bytes. I was able to avoid the issue by switching from strand to mutex. I think there is some bug or logic error in dev_poll_reactor so I wrote a test program to narrow down when it happens: // Christopher Hite RTS Realtime Systems Software GmbH, Rembrandtstrasse 13, D-60596 Frankfurt am Main T: +49.69.61009.0 / F: +49.69.61009.181 Sitz: Frankfurt am Main - HRB 84467 Amtsgericht Frankfurt am Main Geschäftsführer: Steffen Gemuenden, Igor Sluga www.rtsgroup.net This email and any attachments are for the exclusive and confidential use of the intended recipient. If you are not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, please do not read, distribute or take action in reliance upon this message. If you have received this in error, please notify me immediately by return email and promptly delete this message and its attachments from your computer system. ------------------------------------------------------------------------------- #include <boost/asio.hpp> #include <iostream> using namespace boost::asio; io_service service; io_service::strand strand_(service); local::stream_protocol::socket s0(service); local::stream_protocol::socket s1(service); void wh(const boost::system::error_code& ec,const std::size_t){ if(ec) std::cout<<ec; //never happens } void sendAByte(){ static char sendBuf[256]; async_write(s0,buffer(sendBuf,1),wh); } int main(){ local::connect_pair(s0,s1); std::cout<<"dispatching\n"; for(int i=0;i<100000;i++) #if 1 strand_.dispatch(sendAByte); #else sendAByte(); //doesn't cause the sys calls #endif std::cout<<"waiting on run to finish\n"; service.run(); return 0; } //------------------------------------------------------------------------------- Here are some dtraces to help analyze what's going on $XX is the executable. This one shows all the calls to write() with size=0. async_write uses sendmsg() so all these are attributable to dev_poll_reactor. dtrace -n 'syscall::write:entry/pid==$target&&arg2==0/{ ustack(); }' -c $XX 0 5467 write:entry libc.so.1`_write+0x8 strandtest`__1cFboostEasioGdetailQdev_poll_reactor4b_Drun6Mb_v_+0x1b4 strandtest`__1cFboostEasioGdetailPtask_io_service4n0CQdev_poll_reactor4b___Gdo_one6Mrn0CLscoped_lock4n0CLposix_mutex___pn0EQidle_thread_info_rn0AGsystemKerror_code__I_+0x204 strandtest`__1cFboostEasioGdetailPtask_io_service4n0CQdev_poll_reactor4b___Drun6Mrn0AGsystemKerror_code__I_+0xa8 strandtest`main+0xe8 This one counts the system calls. What's very disturbing is that the longer you let it run the more sys calls are made. Why doesn't the system block waiting for the socket to become write-able again (which it won't)? dtrace -n 'syscall:::entry/pid == $target/ { @num[probefunc] = count(); }' -c $XX fcntl 6 brk 2474 sendmsg 21505 ioctl 60194 write 60195 lwp_sigmask 120382 Similar testing on Linux(x86) using strace doesn't show so many system calls (except a few for locking mutexes). Perhaps Christopher Kohlhoff or whoever implemented the thing can take a look. I'm guessing there some event which keeps getting detected but isn't processed. Would someone please take a look or do I need to debug it myself?

1 0

POD 3d vectors
by gtsml owevwr 04 Jun '09

04 Jun '09

Hi, I need fast low dimensional vector (2, 3 and 4 dimension let say). How shall I implement them with boost? Off course ublas is a possibility but I feel it's more designed for higher dimension. Ideally, low dimension vector are better as POD, without any pointer indirection for instance. Maybe there's a way to enforce no indirection with ublas? Off course, boost::array would do the trick but it doesn't have any of the basic arithmetic operators. I could easily add them with the operators library, but maybe there's a better way? And off course, could be great to have POD small matrices as well... I know this request is probably over specialized but maybe I'm missing something already there Thanks, Frank

8 7

[asio] thread-safe async_write with streambuf
by Rutger ter Borg 04 Jun '09

04 Jun '09

Hi all, I'm looking into thread-safe writing of data using Asio, using a streambuf. My current implementation boils down to something like the following. // posting Something from >1 thread(s) void async_send( Something s ) { m_strand.post( async_send_something(s) ); } // Serialized execution // Protects resources socket and streambuf void async_send_something( Something s) { std::ostream os( &m_streambuf ); os << s << std::flush; asio::async_write( m_socket, m_streambuf, transfer_all(), bind( async_complete ) ); } void async_complete( error e ) { } My main confusion is on how to do this with a streambuf. What happens between async_write and the completion handler? It could be that async_send_something is called from one thread before a completion handler of an earlier operation finishes in another thread, and in that case, I'm not sure if it is defined what happens with the streambuf. For the situation sketched above, I couldn't find guarantees around the access pattern of other io_service threads, so my question would be if the implementation above is really thread-safe, and if not, how to best achieve that. Thanks, Kind regards, Rutger

1 0

Using different variables based on template type
by Ryan McConnehey 04 Jun '09

04 Jun '09

I have a couple of functions that provide the same functionality for different types. I'd like to combine them to reduce code duplication. An example of two functions is listed below. std::map<std::string, uint16> m_Uint16; std::map<std::string, uint32> m_Uint32; void putUint16(std::string const& variableName, uint16 const& value) { m_Uint16[variableName] = value; } void putUint32(std::string const& variableName, uint32 const& value) { m_Uint32[variableName] = value; } The functions can't be combined into a template since they are the interface for the programmer. Is there a boost library that would let me call a template or common function that would use a different variable based on the variable type? Ryan

4 13

[string_algo] replace string with tokens
by Michael Schuster 03 Jun '09

03 Jun '09

Hello, I'm new to string_algo and have the following problem: I have string like a+3*sin(b) and want to replace a by a different string e.g. Alpha and b by Beta and sin by cos so that Alpha+3*cos(Beta) is the result. So tokens would be "+*()" The replace command works with spaces, but I think this should be possible with string_algo, too? Thanks for any hint Michael Schuster

3 2

[BGL] A bug in const bundle_property_map or in Readable Property Map Concept ?
by Dmitry Bufistov 03 Jun '09

03 Jun '09

Hi there, The attached test case fails to compile. If it is a bug I would be happy to open a ticket for it! Could somebody please confirm? Thank you, Dmitry

3 5

[exception] What's wrong with boost::get_error_info?
by Markus Schweiß 03 Jun '09

03 Jun '09

Dear boost community, I'm really new to using the boost library. Whatever, the first part I want to use is boost::exception. But already there start my problems: I'm using 'boost::get_error_info<boost::throw_file>( myBoostException )' to retrieve the file name where my exception was thrown. The problem is that the returned 'const char*' pointer contains several wild characters at its end. It seems that the null-termination get lost somehow, somewhere. The same is with 'boost::throw_function' and the string returned by 'boost::diagnostic_information( myBoostException )'! For example, that is what I get from boost::diagnostic_information: ".\myFile.cppÍýýýý««««««««îþ(134): Throw in function MyClass::DoSomething()ÍÍÍÍÍÍ^ñ6 Dynamic exception type: class MyBoostException" That's really strange because the strings are correctly stored to the boost::exception class attributes 'throw_function_' and 'throw_file_'. So what do I wrong??? I would be so thankful when anyone could help me! Following some detailed information: - plattfrom: Windows XP SP3 - compiler: MS VC++ 8 - character set: multi-byte - boost version: Release 1.37.0 code snippet for throwing my exception: throw MyBoostException() << boost::throw_file( __FILE__ ) << boost::throw_line( __LINE__ ) << boost::throw_function( "MyClass::DoSomething()" ); code snippet for catching my exception: catch( MyBoostException& myBoostException ) { std::string info = boost::diagnostic_information( myBoostException ); const char* string = info.c_str(); } My best regards, Markus

2 5

[Program Options]
by Björn Pollex 03 Jun '09

03 Jun '09

Hi everyone ! Is there a simple way to make the command line parser throw an error when an option is not specified ? This way i would not have to check this in my code. My program requires the user to specify an input file, and if he doesn't, an error should be displayed. So, is there any support for mandatory parameters (they wouldn't really be options if they are mandatory) ? wbr Bjoern Pollex

2 1

[ASIO] Where to use CoInitizeEx
by Daniele Barzotti 03 Jun '09

03 Jun '09

Hi, in a library I have a factory function that creates objects: ---- Main Library File ---- // Main io_service static boost::asio::io_service io_service; static boost::scoped_ptr<boost::thread> io_service_thread; static boost::asio::io_service::work* p_work; static BOOL thread_started; EXTERN_C EUROATLIB_API TETRAMotoHndl CALL GetTETRAMotoObj() { if (!thread_started) { try { // create the work object on the heap p_work = new boost::asio::io_service::work(io_service); // run the IO service as a separate thread io_service_thread.reset( new boost::thread( boost::bind(&boost::asio::io_service::run, &io_service) ) ); thread_started = !thread_started; } catch (boost::thread_resource_error e) { // Failed to create the new thread return 0; } } // create the new object TETRAMoto* pTETRAMoto = new TETRAMoto(io_service); // Assign a reference to io_service work object: // This reference is used to counting the objects created. // When the last TETRAMoto object is deleted the main work object // is destroyed and the io_service thread ends. pTETRAMoto->_io_work_ptr.reset(p_work); // return the object return pTETRAMoto; }; --------------------------------- These TETRAMoto objects creates another object, ATDispatcher, that use a SerialPort (which is a wrapper around asio::serial_port) to continuously read and write to RS232. When the SerialPort receive data, pass it to the ATDispatcher that raises a signal to the TETRAMoto object that parses the data and raise the event (always a signal) to the extern: asio::serial_port -> SerialPort -> ATDispatcher -> TetraMoto -|-> EXTERN Now I have to encapsulate this lib into an ActiveX DLL to be used by VB6 and I have to call CoInitizeEx in the thread that "raise" events.. My question is: where I have to call CoInitizeEx? Thanks in advance, Daniele.

5 14