Re: Patch: java.net win32 - removing unused code

[Mohan Embar <gnustuff at thisiscool dot com>]

Hi Michael,

>I would like to commit the attached patch to trunk. It removes the
>code enclosed by the DISABLE_JAVA_NET defines. This code is unused
>because when --disable-java-net is specified these files are not used
>during build time (then the *NoNet.cc versions are used).

THE PROBLEM

As we talked about before, I have already done this (and much, much more)
here:

http://gcc.gnu.org/ml/java-patches/2003-q3/msg00111.html

But this patch is in a wait state pending this one:

http://gcc.gnu.org/ml/java-patches/2003-q3/msg00390.html

...which Bryce said he is okay with with one exception:

http://gcc.gnu.org/ml/java-patches/2003-q3/msg00391.html

...but for which I am getting discrepancies with his theory as to
which is faster:

http://gcc.gnu.org/ml/java-patches/2003-q3/msg00392.html

:( :( :(

I would also need to reconcile the cleanup patch with this one:

http://gcc.gnu.org/ml/java-patches/2003-q3/msg00283.html

MOHAN'S PRESCRIPTION

- someone approve these two as-is:

  - http://gcc.gnu.org/ml/java-patches/2003-q3/msg00283.html
  - http://gcc.gnu.org/ml/java-patches/2003-q3/msg00390.html

    We can always quibble about whether we use _Jv_AllocBytes()
    or _Jv_Malloc() later.

- then I redo this in terms of the above two and resubmit:

  - http://gcc.gnu.org/ml/java-patches/2003-q3/msg00111.html

- then someone approves the above

- then I can check these in and DISABLE_JAVA_NET will be
  gone forever. I'd need to slightly modify my cleanup patch
  before checking it in.

-- Mohan
http://www.thisiscool.com/
http://www.animalsong.org/