Re: proposal: VMStackBrowser

[Archie Cobbs]

Steven Augart wrote:
> I propose adding a new class that may be optionally implemented by the
> VM, gnu.classpath.VMStackBrowser.  I say that it is "optional" because
> I have here a reference implementation for it that uses the existing
> gnu.classpath.VMStackWalker to do all the work:
>
> /* VMStackBrowser.java -- Reference implementation of OPTIONAL
>    VM hooks for stack access.
>    
>    [The Classpath copyright notice can go here]
> */
>
> package gnu.classpath;
>
> /**
>  * This optional class provides access to the classes on the Java stack.
>  * Use this class to explore the stack.  It is sometimes necessary to
>  * find out the current context class loader, and other things like that.
>
>  * <P>
>  * This reference implementation works on top of the mandatory VMStackWalker
>  * class.  So you can leave it completely alone if you want to.
>  * You may be able to improve your VM's efficiency by implementing
>  * VMStackBrowser yourself.  It's your choice.
>  */

Looks good. Presumably the idea here is to optimize this in such a
way that you don't have to crawl the entire stack in the constructor
(as is done by the default implementation). In that case, to make such
implementations easier, we should specify that the VMStackBrowser
object is only valid until the instantiating method returns, and that
the instantiating method is required to call release() before returning
(e.g., using try { ... } finally { x.release(); } or whatever).

Then simple implementations that store the current position as a
frame pointer (or whatever) are possible and don't need to worry
about the stack changing out from under them.

-Archie

__________________________________________________________________________
Archie Cobbs      *        CTO, Awarix        *      http://www.awarix.com