More information on JNI Win32 heap corruption

Øyvind Harboe
Fri May 23 16:21:00 GMT 2003

Perhaps it is reproducable under Linux?

I suppose the code below is free and it shouldn't be too hard to
take it for a spin.

>>However, this might be a GCJ JNI Win32 issue, and not just a buggy
.dll from Sun. 
>It doesn't look like a JNI issue per se to me...

The reason why I think it might be is that it seems like half the
time it is callbacks from Suns JNI code that show up in the debugger.

>If it bugs you that much, see if you can poke around
>the GC sources to figure out the cause of this

I've been taking considerable flak because my app has randomly crashed,
and I now I'm hunting down everything to restore confidence...

Heap corruption in the debugger irks me, though I can't see any negative

effects on the application.

I don't speak GCC development skills too good, so I'll be inspecting
GCC source and running Mohans build(soon with debug info, I hope :-) for
while still.


More information about the Java mailing list