Is JNI functional at this point?

[Mike Swingler]

JNI does work (it is how HotSpot gets any real work done out the built-in class libraries).

Have you identified the closest point in your code where the crash is occurring? Are there type warnings which could indicate a mis-matched expectation of sizes? Have you run this code under 64-bit on Windows? Have you tried linking against the JavaVM.framework and using the built-in Java SE 6 to try to reproduce the problem in a different configuration?

There are any number of problems you could be having, and we'd really need more info to provide any better advice.

Regards,
Mike Swingler
Java Engineering
Apple Inc.

On Aug 16, 2011, at 6:31 PM, David Smith-Uchida wrote:

> I've got quite a bit of JNI code running so I would say you should expect it to work.
> 
> Cheers,
> Dave Smith
> 
> On Aug 17, 2011, at 12:31 AM, Mark Roos wrote:
> 
>> I am porting some JNI code from a Windows JDK7 testbed to OSX 10.6.8 with 
>> a recent JDK7 OSX  build from
>> Google code.  I have it compiled and linked but it segfaults doing a 
>> JNI_ArgumentPusherVaArg:... operation. when
>> it starts.
>> 
>> Should I expect it to work?  or am I too early?
>> 
>> thanks 
>> 
>> mark
>