Re: Bug in java_nio_VMDirectByteBuffer.c

On Tue, Mar 15, 2005 at 10:25:50PM -0600, Archie Cobbs wrote:
> There is a bug in java_nio_VMDirectByteBuffer.c in the function
> Java_java_nio_VMDirectByteBuffer_init().
> 
> This function sets "classRawData", which is a static variable of type
> jclass, from the result of calling (*env)->FindClass(), and then returns,
> and then other JNI functions try to use this variable from within different
> JNI invocations.
> 
> This is broken because the local native reference that was created by
> (*env)->FindClass() goes away when Java_java_nio_VMDirectByteBuffer_init()
> returns (not to mention that it could be used from a different thread).
> 
> The solution is to put a global native reference around "classRawData".
> This fixes the problem for me. I'll commit the attached patch unless
> there are issues.
> 
> 2005-03-15  Archie Cobbs  <archie@xxxxxxxxxxxx>
> 
>          * native/jni/java-nio/java_nio_VMDirectByteBuffer.c: use
>          global native reference to wrap persistent jclass variable.

Thanks, Looks good. Can you point me to some explaining links on the web
for this? All I found via goolge was not very well explaining. I think
there are much more of these bugs in out JNI code and more in my
upcoming new code. Perhaps I can fix them before commiting new code when
I understand this more.


Michael
-- 
Java Trap: http://www.gnu.org/philosophy/java-trap.html