- Boost-users - lists.boost.org

Re: [Boost-users] Does shared_array's get() method introduceruntimeoverhead?
by Sohail Somani 25 Oct '06

25 Oct '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of > Sohail Somani > > I have the following two code segments. I'm wondering if > there is any > > run time overhead of the get() method. Shall I use the latter one > > instead of the former one? > > > > > > > > boost::shared_array a(new int[100]); > > for(int i = 0; i < 100; ++ i) > > a[i] = i; > > > > > > > > boost::shared_array a(new int[100]); > > int a_ptr = a.get(); > > for(int i = 0; i < 100; ++ i) > > a_ptr[i] = i; > > I'd guess not (from shared_array.hpp): > > template<class T> class shared_array > { > ... > T * get() const // never throws > { > return px; > } > ... > > Unless that doesn't get inlined of course. > [snip smart-ass comment] Argh, of course I meant to paste: T & operator[] (std::ptrdiff_t i) const // never throws { BOOST_ASSERT(px != 0); BOOST_ASSERT(i >= 0); return px[i]; }

3 2

Re: [Boost-users] Does shared_array's get()methodintroduceruntimeoverhead?
by Sohail Somani 25 Oct '06

25 Oct '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Peng Yu > Sent: Tuesday, October 24, 2006 6:31 PM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] Does shared_array's > get()methodintroduceruntimeoverhead? > > On 10/24/06, Sohail Somani <s.somani(a)fincad.com> wrote: > > > -----Original Message----- > > > From: boost-users-bounces(a)lists.boost.org > > > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Peng Yu > > > Sent: Tuesday, October 24, 2006 5:33 PM > > > Does it mean that code segment 1 is as fast as segment 2, but code > > > segment 3 is the slowest? > > > > > > ///////1 > > > boost::shared_array a(new int[100]); > > > for(int i = 0; i < 100; ++ i) > > > a.get()[i] = i; > > > > > > ////////2 > > > boost::shared_array a(new int[100]); > > > int a_ptr = a.get(); > > > for(int i = 0; i < 100; ++ i) > > > a_ptr[i] = i; > > > > > > ////////3 > > > boost::shared_array a(new int[100]); > > > for(int i = 0; i < 100; ++ i) > > > a[i] = i; > > > > All I can say is profile, profile, profile! > > > > PS: Profile. > > PPS: I'd guesstimate with a proper C++ compiler + optimizer, you > > wouldn't notice much of a difference between 1 and 3. > > Could you recommend a profiler on linux x86? I recall using something called callgrind. There is a gui called kcachegrind. I hope that should give you enough search keywords. Its been a while since I profiled on linux. You also always have the gcc option -pg which I've used quite often.

1 0

Re: [Boost-users] Does shared_array's get() methodintroduceruntimeoverhead?
by Sohail Somani 25 Oct '06

25 Oct '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Peng Yu > Sent: Tuesday, October 24, 2006 5:33 PM > Does it mean that code segment 1 is as fast as segment 2, but code > segment 3 is the slowest? > > ///////1 > boost::shared_array a(new int[100]); > for(int i = 0; i < 100; ++ i) > a.get()[i] = i; > > ////////2 > boost::shared_array a(new int[100]); > int a_ptr = a.get(); > for(int i = 0; i < 100; ++ i) > a_ptr[i] = i; > > ////////3 > boost::shared_array a(new int[100]); > for(int i = 0; i < 100; ++ i) > a[i] = i; All I can say is profile, profile, profile! PS: Profile. PPS: I'd guesstimate with a proper C++ compiler + optimizer, you wouldn't notice much of a difference between 1 and 3.

2 1

Re: [Boost-users] Does shared_array's get() method introduce runtimeoverhead?
by Sohail Somani 24 Oct '06

24 Oct '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Peng Yu > Subject: [Boost-users] Does shared_array's get() method > introduce runtimeoverhead? > > Hi, > > I have the following two code segments. I'm wondering if there is any > run time overhead of the get() method. Shall I use the latter one > instead of the former one? > > > > boost::shared_array a(new int[100]); > for(int i = 0; i < 100; ++ i) > a[i] = i; > > > > boost::shared_array a(new int[100]); > int a_ptr = a.get(); > for(int i = 0; i < 100; ++ i) > a_ptr[i] = i; I'd guess not (from shared_array.hpp): template<class T> class shared_array { ... T * get() const // never throws { return px; } ... Unless that doesn't get inlined of course. PS: Use the source :)

1 0

GIL Review
by Christian Henning 24 Oct '06

24 Oct '06

Hi Here's my GIL review: > Please always explicitly state in your review, whether you think the > library should be accepted into Boost. I vote to accept GIL into boost. > - What is your evaluation of the design? I like the design and the use of modern C++ paradigms. The design seems to well thought through and taken the whole discourse so far it also seems to hold against heavy criticism. I don't think I'm the position or having the skills to criticize the design. I'm a boost user and happy to have found that place. > - What is your evaluation of the implementation? Seems fine to me. Although I'm a bit anal when code lines that are longer than 80 characters. But that is more minor. I'm not sure the extensions are on review as well. If so, using plain old C code for reading and writing to files, feels outdated. > - What is your evaluation of the documentation? This my biggest problem so far. I like a lot of documentation. My favorite is the one from ASIO. And in this way GIL is loosing. I do understand that documentation is hard but I think it should be treated as source code. For example, I was trying to figure out how to implement indexed images using GIL. Since, it's not there yet it should be easy to add. If I would have known through the documentation the image views have struct add_deref then I might have been able to figure out how to do it. Also I think the documentation should not just enumerate but also explain plus examples. As I said I'm a user and so I don't need to understand all the details on how and why. > - What is your evaluation of the potential usefulness of the library? There are millions of things the lib can do and hopefully will. Image processing is just the start! I'm think of 3D, loading Quake models and so on ... . Would be cool if researchers can use GIL to test new algorithms quickly in an efficient way. > - Did you try to use the library? With what compiler? Did you have > any problems? Yes, I'm using it and have send the creators extensions to read and write bmp files. An early version. Sorry, Lubomir, haven't had a single chance to work on it. I also played around with adding SDL to GIL. Using SDL would make it easy to display a window with the actual image. I'm using VC2003 on Windows XP SP2. > - How much effort did you put into your evaluation? A glance? A quick > reading? In-depth study? I know about GIL for quite a while now. I have tried some things and was usually successful in implementing. But I haven't done an in-depth study. > - Are you knowledgeable about the problem domain? I would be the last to call myself an expert in image processing. I know basics and eager to learn more. The library will certainly help me. Regards, Christian

2 1

Does shared_array's get() method introduce runtime overhead?
by Peng Yu 24 Oct '06

24 Oct '06

Hi, I have the following two code segments. I'm wondering if there is any run time overhead of the get() method. Shall I use the latter one instead of the former one? boost::shared_array a(new int[100]); for(int i = 0; i < 100; ++ i) a[i] = i; boost::shared_array a(new int[100]); int a_ptr = a.get(); for(int i = 0; i < 100; ++ i) a_ptr[i] = i; Thanks, Peng

1 0

Deferred Function calls
by Matt Davies 24 Oct '06

24 Oct '06

I am implementing a signal & slots implementation with a twist. Normally when a signal is fired the connected slots are called immediately. I will need to defer the actual calls to be invoked later. Basically I need to store the function reference (via pointer or functor) and the accompanying parameter values to a 'composite' object. This composite object will be stored on a buffer. Later another processor will pick this object of the buffer and then invoke the actual function call. I am not sure how to do this, or if it can be done. I've been looking at functors and binding functionality in Boost but I am not sure if I can store a functor with parameter values in a generic container only to invoke it at a later time. The motivation is that I want a signal & slots solution where slots are executed on different processors in a multi-processor environment. Any ideas would be very welcome. Thanks, Matt Davies Software Engineer Action Pants Inc.

2 1

Re: [Boost-users] Deferred Function calls
by Sohail Somani 24 Oct '06

24 Oct '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Matt Davies > > I am implementing a signal & slots implementation with a > twist. Normally > when a signal is fired the connected slots are called > immediately. I will > need to defer the actual calls to be invoked later. > Basically I need to > store the function reference (via pointer or functor) and the > accompanying > parameter values to a 'composite' object. This composite > object will be > stored on a buffer. Later another processor will pick this > object of the > buffer and then invoke the actual function call. > > I am not sure how to do this, or if it can be done. I've > been looking at > functors and binding functionality in Boost but I am not sure > if I can store > a functor with parameter values in a generic container only > to invoke it at > a later time. The motivation is that I want a signal & slots > solution where > slots are executed on different processors in a > multi-processor environment. I think you're on the right track. Basically, you want the interface to be `void fn()' Binding should help you: deferred_fn_t fn=boost::bind(the_function,1,2,3); And have a vector of deferred_fn_t's? HTH! Sohail

1 0

[asio][serialization] read_until end of archive, which delimiter
by Christian Rössel 24 Oct '06

24 Oct '06

Dear all, I want to receive a boost::archive::text_iarchive using the boost::asio::read_until () function. boost::asio::streambuf response; boost::asio::read_until(socket, response, boost::regex("\r\n")); std::istream responseStream(&response); boost::archive::text_iarchive responseArchive (responseStream); responseArchive >> retVal; The data is sent using code linke this: boost::asio::streambuf response; std::ostream commandStream (&response); boost::archive::text_oarchive responseArchive (commandStream); const ReturnValue retVal (42); responseArchive << retVal; boost::asio::write(socket, response); If I use boost::regex("\r\n") or a character that does not occur in the sent archive, read_until throws an eof exception. If I use a delimiter that is contained in the archive, read_until succeeds and the streambuf does contain the entire archive, not only up to the delimiter. What is the correct way to read the entire archive without throwing an exception? Thanks and regards, Christian

3 5

[serialization] Serialization of derived classes
by Christian Henning 24 Oct '06

24 Oct '06

Hi there, I'm having problems when serializing derived classes from a non-polymorphic base class. I have read the section "Pointers to Objects of Derived Classes". But for some reasons I cannot make these tricks to work in my application. I have created a small sample application that shows my problem: class base { public: base() : _type( 0 ) {} base( unsigned int type) : _type( type ) {} unsigned int type() { return _type; } private: friend boost::serialization::access; template< class ARCHIVE > void serialize( ARCHIVE& ar, const unsigned int version ) { ar & _type; } private: unsigned int _type; }; class derived1 : public base { public: derived1() : base( 1 ), _value( 2.1 ) {} private: friend boost::serialization::access; template< class ARCHIVE > void serialize( ARCHIVE& ar, const unsigned int version ) { ar & boost::serialization::base_object<base>( *this ); ar & _value; } private: double _value; }; class derived2 : public base { public: derived2() : base( 2 ), _value( 8 ) {} private: friend boost::serialization::access; template< class ARCHIVE > void serialize( ARCHIVE& ar, const unsigned int version ) { ar & boost::serialization::base_object<base>( *this ); ar & _value; } private: short _value; }; BOOST_CLASS_EXPORT( derived1 ); BOOST_CLASS_EXPORT( derived2 ); int _tmain(int argc, _TCHAR* argv[]) { std::string data; { std::ostringstream archive_stream; boost::archive::binary_oarchive archive( archive_stream ); derived2 d; archive & d; data = archive_stream.str(); } { // deserialize std::string archive_data( &data[0], data.size() ); std::istringstream archive_stream( archive_data ); boost::archive::binary_iarchive archive( archive_stream ); // I don't know what derived class was serialized. base* b; archive & b; if( b->type() == 1 ) { derived1* d1 = static_cast<derived1*>( b ); } else { derived2* d2 = static_cast<derived2*>( b ); } } return 0; } Any help is more than welcome. Thanks ahead, Christian

4 12