C++Talk.NET C++ language newsgroups   Archives   FAQ   Search   Memberlist   Usergroups   Register   Profile   Log in to check your private messages   Log in  volatile-qualified primitive types          C++Talk.NET Forum Index -> C++ Language (Moderated) View previous topic :: View next topic   Author Message Christopher D. Russell Guest Posted: Thu Mar 17, 2005 8:18 am    Post subject: volatile-qualified primitive types Hello, I'm reading the Alexandrescu article "Generic<Programming>: volatile - Multithreaded Programmer's Best Friend Volatile-Correctness or How to Have Your Compiler Detect Race Conditions for You" (http://www.cuj.com/documents/s=7998/cujcexp1902alexandr/) - an excellent expose on the volatile qualifier. However, for the life of me, I can't figure out how to volatile-qualify a primitive type so that the compiler emits an error (or warning even) if I operate on the variable without const_cast. Alexandrescu writes "Simply encapsulate the primitive data that you use in higher-level structures and use volatile with those structures. Paradoxically, it's worse to use volatile directly with built-ins, in spite of the fact that initially this was the usage intent of volatile!" So I tried this: template <typename PrimitiveT> class PrimitiveWrapper { private: // disallow default construction PrimitiveWrapper() {} public: // construction/destruction PrimitiveWrapper(PrimitiveT const& _Primitive) : Primitive(_Primitive) {} ~PrimitiveWrapper() {} public: // implementation PrimitiveT Primitive; }; ..... The idea is that I should be able to, for example, declare a volatile PrimitiveWrapper<int> as opposed to a volatile int in order to get the compiler to emit an error if I attempt to access its Primitive member sans const_cast. No luck however. Even volatile-qualify the declaration of Primitive in my wrapper seems to make no difference. Would someone please suggest a way that I can accomplish this task. The corollary question is why on earth is this standard behavior? I'm sure there's some good reason (there always is). But it's not at all obvious to me what it is. Any assistance is appreciated. - Regards Chris [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Andrei Alexandrescu (See Guest Posted: Fri Mar 18, 2005 1:12 pm    Post subject: Re: volatile-qualified primitive types Christopher D. Russell wrote: Quote: Hello, I'm reading the Alexandrescu article "Generic<Programming>: volatile - Multithreaded Programmer's Best Friend Volatile-Correctness or How to Have Your Compiler Detect Race Conditions for You" (http://www.cuj.com/documents/s=7998/cujcexp1902alexandr/) - an excellent expose on the volatile qualifier. Thanks. You should be aware, however, that the introductive example on volatile's role is mistaken, which has led many to misunderstand the article and think it relies on some specific semantics of volatile. To see why, you may want to check: http://moderncppdesign.com/publications/DDJ_Jul_Aug_2004_revised.pdf Quote: However, for the life of me, I can't figure out how to volatile-qualify a primitive type so that the compiler emits an error (or warning even) if I operate on the variable without const_cast. Alexandrescu writes "Simply encapsulate the primitive data that you use in higher-level structures and use volatile with those structures. Paradoxically, it's worse to use volatile directly with built-ins, in spite of the fact that initially this was the usage intent of volatile!" So I tried this: template <typename PrimitiveT> class PrimitiveWrapper { private: // disallow default construction PrimitiveWrapper() {} public: // construction/destruction PrimitiveWrapper(PrimitiveT const& _Primitive) : Primitive(_Primitive) {} ~PrimitiveWrapper() {} public: // implementation PrimitiveT Primitive; }; .... The idea is that I should be able to, for example, declare a volatile PrimitiveWrapper<int> as opposed to a volatile int in order to get the compiler to emit an error if I attempt to access its Primitive member sans const_cast. No luck however. Even volatile-qualify the declaration of Primitive in my wrapper seems to make no difference. This is just a simple error. When you encapsulate data, you must as a matter of course make that data private and make it accessible via functions only. Otherwise, no matter how you play the constructors/destructors game, client code that accesses that data will bypass all of your fortress. Andrei [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Christopher D. Russell Guest Posted: Sat Mar 19, 2005 10:37 am    Post subject: Re: volatile-qualified primitive types [ In reference to: http://tinyurl.com/6htof ] Andrei, thank you for the response. Quote: ... When you encapsulate data, you must as a matter of course make that data private and make it accessible via functions only. Otherwise, no matter how you play the constructors/destructors game, client code that accesses that data will bypass all of your fortress. Okay. That's clear enough. I was hoping I missed something that would allow me to accomplish this without resorting to access methods and private data. Oh well. Expressing my code so that others don't necessarily have to understand the nuances of the threading model is more important than elegance and performance in the grand scheme of things. The cited article http://moderncppdesign.com/publications/DDJ_Jul_Aug_2004_revised.pdf was very illuminating. In 10 [Sidebar] volatile: A Brief History, you discuss the more restrictive semantics imposed on the volatile concept by Java and .NET CLI and conclude by stating "We know of no similar work being done on C's or C++'s volatile." Maybe this is because most people are like I was a few days ago - blissfully unaware of the incorrectness of their multithreaded code and ignorant of the complexities of fixing it? Generalizing the notion of "abstract machine" to multiple threads and multiple processors is truly daunting. Still, given that multiple core processors and SMP machines are more and more common, I'm surprised to learn that it's not a front burner issue in the C++ community. The ability to succinctly and portably express concurrency seems _so_ vital. It sure would be nice to have standard language support for something syntactically similar to namespaces that could be used to (a) preserve sequence ordering for a block of statements (b) explicitly express cache coherency constraints. Maybe someday. - Thanks again and regards Chris Russell http://encapsule.com // http://hyperworx.org // http://chrisrussell.net [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Andrei Alexandrescu (See Guest Posted: Sun Mar 20, 2005 11:38 pm    Post subject: Re: volatile-qualified primitive types Quote: The cited article http://moderncppdesign.com/publications/DDJ_Jul_Aug_2004_revised.pdf was very illuminating. In 10 [Sidebar] volatile: A Brief History, you discuss the more restrictive semantics imposed on the volatile concept by Java and .NET CLI and conclude by stating "We know of no similar work being done on C's or C++'s volatile." Maybe this is because most people are like I was a few days ago - blissfully unaware of the incorrectness of their multithreaded code and ignorant of the complexities of fixing it? Generalizing the notion of "abstract machine" to multiple threads and multiple processors is truly daunting. Still, given that multiple core processors and SMP machines are more and more common, I'm surprised to learn that it's not a front burner issue in the C++ community. The ability to succinctly and portably express concurrency seems _so_ vital. Fortunately, there are good news. Just like you, seeing that (astonishingly) nobody seemed to consider that adding threading support to C++ is vital for the community and for the survival of the language itself, yours truly has decided to spur an effort in negative time and to convince a small team of experts (notably Hans Boehm) to contribute: http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2004/n1680.pdf http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1777.pdf Andrei [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Display posts from previous: All Posts1 Day7 Days2 Weeks1 Month3 Months6 Months1 Year Oldest FirstNewest First         C++Talk.NET Forum Index -> C++ Language (Moderated) All times are GMT Page 1 of 1   Jump to: Select a forum C/C++----------------C LanguageC Language (Moderated)C++ Language (Moderated)C++ language (comp.lang.c++)C++ language, library and standards Non-English C/C++ Newsgroups----------------C++ (German)C++ (French)  You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum Powered by phpBB © 2001, 2006 phpBB Group SEO toolkit © 2004-2006 webmedic.