RE: [cp-patches] RFC: RMI Class loader fix

[Jeroen Frijters]

Andrew Haley wrote:
> All of this is fairly ill-specified. 

As usual :-( I just re-read the relevant part of the 1.5 API spec and
JLS (which it refers to) and I think you can actually interpret it to
prohibit loading array classes via ClassLoader.loadClass (indirectly
based on the definition of "binary name").

> ClassLoader.loadClass() can to
> cope with "[I", for example, and there is extant application code that
> requires it.  That's how I discovered the same incompatibility that
> you've found.  This may be a bug, or perhaps just a side-effect of a
> particular implementation.

It's probably a side effect from the fact that findLoadedClass looks in
the hashtable that contains all classes associated with a class loader.
Once the array classes get created they're presumably stuffed into the
hashtable (it's what I do at least) and subsequently they show up in the
ClassLoader.loadClass() path.

> Is there a strong reason to prefer fixing all the clients of
> loadClass() ?  There doesn't seem to be any strong specification
> reason to do so.

I don't have any strong arguments for doing that, but it feels like it's
the right thing to do to me.

Regards,
Jeroen