This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
Re: One step closer to getting Mohans Test JVM to work
- From: Øyvind Harboe <oyvind dot harboe at zylin dot com>
- To: gnustuff at thisiscool dot com
- Cc: java at gcc dot gnu dot org
- Date: Mon, 10 Nov 2003 14:15:24 +0100
- Subject: Re: One step closer to getting Mohans Test JVM to work
- Organization: Zylin AS
- References: <2YY9HCMIKGC787ECYU1UVSLFB6VRGE.3faf8bab@p733>
On Mon, 2003-11-10 at 13:59, Mohan Embar wrote:
> Hi Ãyvind,
>
> The JVMTest sample works as-is in my latest build, but there
> are no DLLs involved.
Hmmm... I saw some patch messages, but I didn't see any commits. Perhaps
my local GCC build is stale.
> I'm not sure how the DLL in your example relates to this problem.
> Is send.dll doing something with JNI?
It must be a seperate problem then.
What I'm trying to do is to write a DLL with a C++ interface for use
from e.g. Visual C++, hence I need to launch/shutdown the VM as in your
example. No other symbols should(or needs) to be exported than the C++
API. The rationale behind this approach is to sneak in some Java code
under the radar for this legacy C++ Windows app.
The "interface" consists of a single exported from send.dll:
/* returns success failure*/
bool sendMail(
/* null terminated list of recipients */
wchar_t **recipients,
wchar_t *subject,
wchar_t *body,
/* null terminated list of Win32 paths to attachments */
wchar_t *attachments);
Ãyvind