Re: PR java/19870: Generate Synthetic Accessors For Static Private Field Accesses Across Nested Classes

[Ranjit Mathew <rmathew at gmail dot com>]

On 26 May 2005 12:21:59 -0600, Tom Tromey <tromey@redhat.com> wrote:
> >>>>> "Ranjit" == Ranjit Mathew <rmathew@gmail.com> writes:
> 
> Ranjit>   Here is a patch for PR java/19870. This patch generalises
> Ranjit> the current synthetic accessor generation support in GCJ for
> Ranjit> private fields in outer classes accessed from nested classes
> Ranjit> to additionally support:
> 
> I think we also need similar treatment for private methods.
> Or is this working ok?

No, it isn't. The patch was only for making static fields 
work across nested class boundaries.

I think we should go for fixing that under the same
PR since it just talks of private member access.


> Ranjit> Tested on i686-pc-linux-gnu with Jacks included. The Jacks
> Ranjit> testresults were rather disappointing:
> Ranjit>   FAIL: 8.5.2-accessible-static-member-usage-3
> Ranjit>   FAIL: 15.8.4-static-2
> 
> I wouldn't expect to see much improvement on jacks simply because we
> accepted this code fine, and then generated invalid bytecode.  Most of
> the jacks tests are purely for semantic analysis and don't try
> executing the result.

True.

Thanks,
Ranjit.

-- 
Ranjit Mathew      Email: rmathew AT gmail DOT com

Bangalore, INDIA.    Web: http://ranjitmathew.hostingzero.com/