InvokeDynamic
From APIDesign
| Line 1: | Line 1: | ||
| + | When I was younger I believed that having with [[invokeDynamic]] instruction in [[HotSpot]] [[VM]] can be beneficial. [[I]] even argued that the instruction should not be used just for languages like [[Ruby]] but rather by the core [[Java]] to implement [[Closure|lambdas]]. Now, after spending time to implement [[Closure|lambdas]] in my [[Bck2Brwsr]] [[VM]] and seeing things from the other side I have to admit I was wrong. [[invokeDynamic]] is wrong idea (especially for implementation of [[Closure|lambdas]]). | ||
| + | |||
| + | === Benefits === | ||
| + | |||
Implementing different languages on top of [[HotSpot]] virtual machine is of different complexity. The most problematic thing is to properly and effectively dispatch methods calls. Not every language uses the [[Java]] rules. Some support type conversions, implicit arguments. Some can dynamically alter the existing dispatch target or strategies. More about that in an excellent summary [http://blogs.sun.com/jrose/entry/vmil_paper_on_invokedynamic Bytecodes meet Combinators]. | Implementing different languages on top of [[HotSpot]] virtual machine is of different complexity. The most problematic thing is to properly and effectively dispatch methods calls. Not every language uses the [[Java]] rules. Some support type conversions, implicit arguments. Some can dynamically alter the existing dispatch target or strategies. More about that in an excellent summary [http://blogs.sun.com/jrose/entry/vmil_paper_on_invokedynamic Bytecodes meet Combinators]. | ||
| Line 4: | Line 8: | ||
The proposed improvements to the [[HotSpot]] virtual machine may help the [[JDK]] to support different languages, but first and foremost they open the door to effective implementation of [[Closures]]. | The proposed improvements to the [[HotSpot]] virtual machine may help the [[JDK]] to support different languages, but first and foremost they open the door to effective implementation of [[Closures]]. | ||
| + | |||
| + | |||
| + | === Drawbacks === | ||
| + | |||
| + | [[TBD]] | ||
Revision as of 06:45, 17 August 2014
When I was younger I believed that having with invokeDynamic instruction in HotSpot VM can be beneficial. I even argued that the instruction should not be used just for languages like Ruby but rather by the core Java to implement lambdas. Now, after spending time to implement lambdas in my Bck2Brwsr VM and seeing things from the other side I have to admit I was wrong. invokeDynamic is wrong idea (especially for implementation of lambdas).
Benefits
Implementing different languages on top of HotSpot virtual machine is of different complexity. The most problematic thing is to properly and effectively dispatch methods calls. Not every language uses the Java rules. Some support type conversions, implicit arguments. Some can dynamically alter the existing dispatch target or strategies. More about that in an excellent summary Bytecodes meet Combinators.
To address all these different needs the new invokeDynamic bytecode operand does not hardcode the actual invocation, but delegates it to software controllable MethodHandles. What is a method handle? A pointer to method of some signature (for example plus would take two ints and return their sum as an int) and an object - a receiver to call the method on. However this is nothing else than a closure.
The proposed improvements to the HotSpot virtual machine may help the JDK to support different languages, but first and foremost they open the door to effective implementation of Closures.
Follow
