Bryce McKinlay <[EMAIL PROTECTED]> writes: > Cierniak, Michal wrote: > > >Brian Jones writes: > > > >>Bryce McKinlay <[EMAIL PROTECTED]> writes: > >> > >>>I don't think it is neccessary to make it native. Method is already > >>>part of the VM interface, so VMs can modify it or make it native as > >>>they wish. I think the more conservative approach should be taken by > >>>the default implementation. > >>> > >>It may be necessary to consider including an ORP vm directory > >>or ask that the ORP team distribute these particular classes > >>specific to ORP with ORP. > >> > > > >But I thought that Bryce indicated in his message that he also liked the > >patch submitted by Gansha. It sounds like this modification is useful for > >both ORP and GCJ since they both have their own internal data structures to > >represent a method. > > > > Right. Gansha's implementation should be the one that goes in > vm/reference. It may not be optimal for all VMs, but it is always > correct and most accurately matches the JDK spec, and it is the > implementation required for both ORP and GCJ.
Sorry, I mis-read that; I think this answers Mark's question. Brian -- Brian Jones <[EMAIL PROTECTED]> _______________________________________________ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath
