December 2008 - Boost-users - lists.preview.boost.org

ublas vector resize std::bad_alloc
by Fabian Lim 06 Dec '08

06 Dec '08

Im having a memory management problem. I have 2 ublas vectors Vector A = vector<V1 > Vector B = vector<vector<V2 > > : first vector level size is 2^22 where the data types both V1=vector<int> and V2 are vector<vectors<int>> and the size |V1| <= 2 and the first level size of V2 <=2 and the second level size <= 10. A grows as my program runs. B also grows but only the second level. I run in windows and when i check the memory usage, it seems to be around 90,000K when the exception throws. I dont really have a good feel for these figures, so I compared to my firefox browser, which seems to occupy approximately the same amount of memory, hence i concluded that not alot of memory is used. My computers have around 1/2 Gb RAM. I do not see the typical OS slowdown which accompanies alot of paging.

1 0

boost::thread issue
by danielsabin＠comcast.net 05 Dec '08

05 Dec '08

Hi, I started working on a project a couple of months ago and i decided to use boost::thread. I like the library, very nicely done and easy to use - great job. However, I ran into an issue and although i have already spent a significant amount of time trying to figure it out, I was not able to make any progress so far. The project is written in pure C++ (no managed code) on Windows XP/SP3. I am using Visual Studio 2005 with SP1 as the C++ compiler. I am using the pre-built binaries for boost 1_36_0 provided by boost consulting. My main program simply creates several boost::thread objects which do various things. Everything works fine except for one thing: occasionally when I start the program I receive the following error: "The exception unknown software exception (0xc0000409) occurred in the application at location 0x00421cd3". In general exception 0xc0000409 means that the system detected (or thinks it detected) an overrun of a stack-based buffer. The code is compiled with the /GS switch (Buffer security check) enabled. The debug stack trace when the exception is thrown shows 2 threads, the exception being thrown by thread2, as follows: (thread 1) 5752 __tmainCRTStartup __CxxSetUnhandledExceptionFilter Normal 0 --------------------------------------------------------------------------------- ntdll.dll!_KiFastSystemCallRet@0() ntdll.dll!_ZwWaitForSingleObject@12() + 0xc bytes ntdll.dll!_RtlpWaitForCriticalSection@4() + 0x8c bytes ntdll.dll!_RtlEnterCriticalSection@4() + 0x46 bytes ntdll.dll!_LdrLockLoaderLock@12() + 0x6b bytes kernel32.dll!7c80b43f() [Frames below may be incorrect and/or missing, no symbols loaded for kernel32.dll] ntdll.dll!_ZwQueryVirtualMemory@24() + 0xc bytes kernel32.dll!7c80b9ed() kernel32.dll!7c8638bd() kernel32.dll!7c8449b0() ntdll.dll!_RtlpAllocateFromHeapLookaside@4() + 0x31 bytes kernel32.dll!7c80262a() ntdll.dll!_NtQueryInformationProcess@20() + 0xc bytes msvcr80.dll!7814a84c() MyService.exe!__CxxSetUnhandledExceptionFilter() Line 90 C++ msvcr80.dll!78131762() > MyService.exe!__tmainCRTStartup() Line 494 + 0xf bytes C kernel32.dll!7c817007() ntdll.dll!_RtlpAllocateFromHeapLookaside@4() + 0x31 bytes (thread 2) 7976 __report_gsfailure __report_gsfailure Time Critical 0 ------------------------------------------------------------------------- MyService.exe!__crt_debugger_hook() Unknown > MyService.exe!__report_gsfailure() Line 298 + 0x7 bytes C MyService.exe!boost::detail::create_once_mutex() + 0x63 bytes C++ MyService.exe!boost::call_once<void (*)(void)>() + 0x36 bytes C++ MyService.exe!boost::`anonymous namespace'::run_thread_exit_callbacks() + 0x34 bytes C++ MyService.exe!boost::thread::interrupt() + 0xdc bytes C++ ntdll.dll!_LdrpCallInitRoutine@16() + 0x14 bytes ntdll.dll!_LdrpCallTlsInitializers@8() + 0x12fad bytes ntdll.dll!_LdrShutdownThread@0() + 0xf4ae bytes kernel32.dll!7c80c0c6() [Frames below may be incorrect and/or missing, no symbols loaded for kernel32.dll] csauser.dll!10103b7f() kernel32.dll!7c80b6b9() Since I did not get too far on my own I really need your help. Did anybody else run into anything similar or does anybody have any idea about how to debug this? Thanks, --daniel

1 0

[lambda] Can't quite get the hang of it!
by Robert Jones 05 Dec '08

05 Dec '08

Starting with this... for( j = joff; j < joff + 8; j++) for( i = ioff; i < ioff + 8; i++) img->m7[j][i] = iClip1(img->max_imgpel_value, (img->m7[j][i]+(long)mpr[j][i])); I'd like to move as much of the loop invariance out of the loop as I can. So I tried this... using boost :: lambda :: _1; using boost :: lambda :: _2; int ilast = ioff + 8; int jlast = joff + 8; boost :: function< int( int &, int ) > clip = _1 = boost :: lambda :: bind( iClip1, img -> max_imgpel_value, _1 + _2 ); for( j = joff; j != jlast; ++ j ) for( i = ioff; i != ilast; ++ i ) clip( img -> m7[j][i], mpr[j][i] ); Which works, but I'd like the nested clip function to be fn( i, j ), so I tried this... boost :: function< void( int, int ) > clip2 = boost :: lambda :: bind( clip, img -> m7[ _2 ][ _1 ], mpr[ _2 ][ _1 ] ); for( j = joff; j != jlast; ++ j ) for( i = ioff; i != ilast; ++ i ) clip2( i, j ); Which doesn't work. I think I need some boost::refs around the arrays I'm subscripting, but I can't quite seem to get it right. Error is... loop.cpp:50: error: no match for 'operator[]' in 'img->img_par::m7[boost::lambda::<unnamed>::_2]' loop.cpp:50: error: no match for 'operator[]' in 'mpr[boost::lambda::<unnamed>::_2]' loop.cpp:50: error: `bind' undeclared (first use this function) Any thoughts? Thanks, Rob.

2 2

shared_ptr to stack-allocated object
by Jason Cipriani 05 Dec '08

05 Dec '08

I'm refactoring parts of an existing application to use shared_ptrs instead of raw pointers. What is the best way to deal with code like this (please note in this situation passing a pointer to a stack-allocated object to function is acceptable and safe): void function (shared_ptr<Object> ptr) { ... } void function2 () { Object o; function(&o); // <--- need to change this } It seems I have two choices, I am not sure which is better, or if there are other common techniques. I can either use a no-op delete function: void noop (Object *) {} void function2 () { Object o; function(shared_ptr<Object>(&o, noop)); } Or I can allocate it on the heap instead of the stack, and let shared_ptr deal with the cleanup (this seems to be the more convenient solution): void function2 () { shared_ptr<Object> o(new Object); function(o); } What's the normal practice here? Thanks, Jason

4 5

Re: [Boost-users] [review][constrained_value] Review of Constrained Value Library begins today
by jesseperla＠gmail.com 05 Dec '08

05 Dec '08

Robert: >Sorry, but this makes no sense to me in the context of constrained values. If >the current bounds are [0, 1] and you have r1 == .5 and r2 == .9, then what >changing the bounds for the whole type to [.8, .99] would mean for r1? How the >type is supposed to know that there is an instance that is not valid according >to the new bounds? The type should keep handles to all its instances? Or the >bounds should change anyway, leaving the instance in invalid state? So what >would be the point of bounds checking, if you can't guarantee that once you >assign a valid value then the object will remain valid? The problem for me is that the change_bounds function at runtime operators on an instance rather than for the type itself (which is what the compile time bounds mean, and what I really want). I was using this as an example of why I believe the semantics are different and what I really want is something like what I listed next: >typedef bounded<double, "INTEREST RATE"> interest_rate; >typedef bounded<double, "RISK_AVERSION"> risk_aversion; >interest_rate r; >risk_aversion crra ; >change_bounds(r, .8. .99); //Statically change it >change_bounds(crra, .5, 10); //Statically. >interest_rate r_guess; //want the right bounds But even this is a little shady since I kept the change_bounds called on an instance to show the similarity to what is listed. What I would really be looking for is a way to change the bounds for the type itself, which would change static bounds. Maybe something like the following: >typedef bounded<double, "INTEREST RATE"> interest_rate; interest_rate::change_bounds(.8, .99); //Changes static data for the type? Does that help you make sense of my previous questions and comments? Now, you may say that this doesn't work because how could it throw errors on existing variables of type interest_rate that may not fulfill the bounds? The short answer is that it couldn't and it doesn't matter. In practice, you are changing the bounds here prior to creating instances of an item. It is purely a question of being able to dynamically create bounds for a type and then query them later... this makes the library far, far more than just a debugging and bounds checking utility and gives real semantics that can better match mathematics. >What you ask for is not possible in C++. You cannot have one type and use it >exactly the same way as other type. I'll give you some examples (having >constrained<string> cs): > const char * s = cs.c_str(); // not possible >Type constrained<string> can't have members that string has. > char c = cs[0]; // not possible Yes, I see now why it can't work for wrapping classes. But in my mind I am only thinking of the wrapping of intrinsic, built in types like doubles and ints. Here I would never want to call member functions on them... But it sounds like with no implicit conversions for template instantiation, we could run into problems... But how often would this be in practice? My knowledge of generic programming is not sufficient to understand the story here, but it sounds like the problem with std::max is that it is forcing the two parameters to be of the same type and that is the failure. Otherwise if std::max was written to take two different types with compatible comparison operators, that this could work fine? And if the constrained_value<double> had an implicit conversion to double, then we would have no problem? So is std::max anomalous here, and this wouldn't come up in practice for most usage for most numeric work? Thanks Robert. I really like this library and think it has great promise for those of us programming mathematics written on a blackboard.

2 1

noncopyable
by Hicham Mouline 05 Dec '08

05 Dec '08

AFAIU, deriving from boost::noncopyable disables _accidental_ object copying. Is it possible to implement a clone() const member function of the object to _explicitly_ get a copy of the object? Perhaps deriving boost::noncopyable makes it totally impossible to ever get a copy of an object without resorting to byte copy. Regards,

5 6

shared_ptr to const object
by Jason Cipriani 05 Dec '08

05 Dec '08

How do I represent "const Object *" as a shared_ptr? Using "const shared_ptr<Object>" is not right, it makes the shared_ptr const, not the Object. Thanks, Jason

3 3

scoped_ptr with destructor function
by Jason Cipriani 04 Dec '08

04 Dec '08

Does boost have something lightweight like a scoped_ptr but that lets you specify a custom delete function, or do I have to use a shared_ptr for that? Thanks, Jason

3 4

Simple use case for boost exceptions
by Yang Zhang 04 Dec '08

04 Dec '08

With boost::exception, can I simply construct an exception with a custom what() message (without subclassing exception)? Thanks. -- Yang Zhang http://www.mit.edu/~y_z/

2 1

[asio] sending multiple messages in the same soocket
by Henrique Aparício 04 Dec '08

04 Dec '08

Hi! I'm trying to implement an asynchronous client to communicate with a server using asio. To do this, I'm following the example of the http client ( http://tenermerx.com/Asio/asio-1.3.1/src/examples/http/client/async_client.… http://tenermerx.com/Asio/asio-1.3.1/src/examples/http/client/async_client.… ), but I'm trying to make it so that it can be used for a more general purpose. So in my implementation I have the same methods that appear in the example, but with only one method for reading the response. My main function calls the constructor of the client which is responsible for initializing variables and calling the chain of asynchronous methods. After this, the main function calls a method send(message), that puts the message I want to send in the variable request and runs the io_service. The problem is that I want to be able to call this send function several times, but it only works the first time I call it. For example: Client c(host, port); //remember that the constructor calls the asynchronous methods c.send("msg to send"); c.send("msg to send"); The message is actually sent the first time, but the rest of the times nothing happens. What am I missing? Can you help me? If you need more information to be able to help, please ask me. Henrique Aparício -- View this message in context: http://www.nabble.com/-asio--sending-multiple-messages-in-the-same-soocket-… Sent from the Boost - Users mailing list archive at Nabble.com.

2 5