 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Java Forum / First Aid / July 2006Making JAVA methods non-virtual
We know that, In Java every method is virtual by nature, i.e. always late binding is performed. But early binding helps in improving performance. Is it possible to change this behaviour to early binding. i.e. can Java methods be made non-virtual ? > We know that, In Java every method is virtual by nature, i.e. always > late binding is performed. > But early binding helps in improving performance. A modern JVM will be happy to inline your virtual methods. Even in C++ that cost of a virtual function is effectively zero. Look for optimisations (if necessary) at a higher level. Tom Hawtin  SignatureUnemployed English Java programmer http://jroller.com/page/tackline/ > Look for optimisations (if necessary) at a higher level. Actually more than optimisation I am eager to know whether this can be done or not ? If yes, then how ? I was asked this question in an interview. And, the interviewer told me that it is possible to do so as well as asked me to explore it. >> Look for optimisations (if necessary) at a higher level. > [quoted text clipped - 3 lines] > I was asked this question in an interview. And, the interviewer told me > that it is possible to do so as well as asked me to explore it. Take a look at final, i.e. public final void finalFunc() > >> Look for optimisations (if necessary) at a higher level. > >> [quoted text clipped - 6 lines] > > Take a look at final, i.e. public final void finalFunc() C++ virtual methods implement dynamic binding. Non-virtual methods implement static binding. AFAIK all Java methods are dynamically bound, even final ones, since I can't see how to implement them as dynamically bound in superclasses of the class but statically bound in the "final" class. I think I would have flunked that interview as I would have argued that the interviewer was wrong (:-), but I'd like to know the right answer :-) m. >> Look for optimisations (if necessary) at a higher level. > [quoted text clipped - 3 lines] > I was asked this question in an interview. And, the interviewer told me > that it is possible to do so as well as asked me to explore it. Was this an interview for a Java related position? Because AFAIK, Java itself has no concept of "virtual" methods. Rather, it's a term borrowed from C++. What I'm getting at, I guess, is that I suspect it's conceivable for someone to be the worlds greatest Java programmer, and to have no idea what "virtual methods" are. - Oliver >>> Look for optimisations (if necessary) at a higher level. >> [quoted text clipped - 11 lines] > > - Oliver I'd word that differently. All java methods are virtual and can't be none-virtual, and so no keyword is needed. >> Was this an interview for a Java related position? Because AFAIK, >> Java itself has no concept of "virtual" methods. Rather, it's a term >> borrowed from C++. What I'm getting at, I guess, is that I suspect >> it's conceivable for someone to be the worlds greatest Java >> programmer, and to have no idea what "virtual methods" are. > I'd word that differently. Yup, it's a common case of an interview question depending on incorrect terminology. > All java methods are virtual and can't be none-virtual, and so no > keyword is needed. Point of pedanty: Not all of them. Private methods, calls to super methods, static methods, static initialisers, as well as constructors use different, non-virtual bytecode (invokestatic and invokespecial instead of invokevirtual and invokeinterface, IIRC). At runtime anything (consistent with the spec) can happen. Even interface methods implemented in dynamically loaded classes can be inlined. Tom Hawtin SignatureUnemployed English Java programmer http://jroller.com/page/tackline/ >>> Was this an interview for a Java related position? Because AFAIK, >>> Java itself has no concept of "virtual" methods. Rather, it's a term [quoted text clipped - 18 lines] > > Tom Hawtin Cheers, I was a bit lazy there... I was simply referring to public instance methods, rather than the other types. > A modern JVM will be happy to inline your virtual methods. Even in C++ > that cost of a virtual function is effectively zero. An old article talked about the fact that the cost of virtual methods in C++ is basically equal to the cost of adding a parameter. Not to many people have trouble with adding a parameter to a method but so many people worry over the "cost" of virtual methods. SignatureDale King Free MagazinesGet these publications absolutely FREE for up to 12 months. There are no hidden fees and no obligation. Simply choose a title, complete the application form and submit it. Read more ...
|
Reply to this Message
|
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.