This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
Re: RMI problem
- From: Radu Racaru <radu dot racaru at gmail dot com>
- To: Andrew Haley <aph at redhat dot com>
- Cc: Amir Bukhari <ufz6 at rz dot uni-karlsruhe dot de>, java at gcc dot gnu dot org
- Date: Fri, 27 May 2005 15:40:11 +0300
- Subject: Re: RMI problem
- References: <17045.40866.782931.982155@zapata.pink> <200505261524.j4QFOwKc032340@mx3.redhat.com> <17046.3662.108285.285784@zapata.pink>
- Reply-to: Radu Racaru <radu dot racaru at gmail dot com>
yep, this comes from the static linking Win GCJ4 does.
To avoid this you should try to include references to classes that are
loaded with classforname mechanism.
i.e.
class rmiTest{
//force the linker to include this class in your app
java.rmi.server.RemoteStub _fakestub;
// your code.....
}
On 5/26/05, Andrew Haley <aph@redhat.com> wrote:
> Amir Bukhari writes:
> >
> > I am using GCJ 4.0 build for MinGW for Windows.
>
> Ah, Windows. There are problems with gcj for Windows, because it
> doesn't support building DLLs, and you really need functional shared
> library semantics to make this sort of stuff work in general.
>
> There are some hacks that the Windows people use to make this work,
> and perhaps if one of them is listening they could shout out.
>
> Andrew.
>