Re: Speeding up method searches

[Bryce McKinlay <bryce at albatross dot co dot nz>]

Tom Tromey wrote:

> Bryce> Right. Actually, it seems to me that the collector already
> Bryce> doesn't consider the pointers in the intern table for marking,
> Bryce> since it is allocated with _Jv_AllocBytes which is for
> Bryce> "pointer-free" memory. However, there is nothing to remove a
> Bryce> collected string from the hashtable, so we currently get
> Bryce> "random" crashes when calling intern()[*].
>
> Thanks.
> What if we replace the hash table in natString.cc with a
> java.util.Hashtable?

I don't think so. Interned strings should still be collectable, right?
Maybe we can put a special finalizer in natString.cc and call
_Jv_RegisterFinalizer from String.intern()?

> Does interning have to be very fast?  Would the added performance hit
> of this approach be acceptable?

I would say that Interning should be as fast as possible if we are going
to be calling it on every method name and signature of every initialized
class.

regards

  [ bryce ]