On Wed, 23 Nov 2005 12:01:34 -0800, arunmozhian wrote:
> Hello,
> I'm trying to debug an issue with a Java program which involves/uses
[quoted text clipped - 10 lines]
>
> Any resources/pointers on JNI debugging will also be helpful ?
AFAIK, the JDK does not use any system signals it does not provide its own
handlers for. The fact that the signal is raised seems to me like an issue
between the JNI and whatever native libraries used. Take a look at the
crash trace of the java process, and make sure that the signal is not
raised in your (or whatever native libraries you use) code.

Signature
You can't run away forever,
But there's nothing wrong with getting a good head start.
--- Jim Steinman, "Rock and Roll Dreams Come Through"
arunmozhian@gmail.com - 28 Nov 2005 16:22 GMT
There's no crash generated. All that is generated is a User Signal 2.
And I've verified that
my native libs do not handle/raise SIGUSR2. I believe the JVM uses
SIGUSR2 for Garb. coll'n.
-- Arun
Thomas Fritsch - 28 Nov 2005 17:01 GMT
> AFAIK, the JDK does not use any system signals it does not provide its own
> handlers for.
I don't agree here.
The JDK contains some signal-related classes (Signal, SignalHandler,
NativeSignalHandler in package sun.misc) which map pretty straight to
the C-functions raise() and signal(). This fact makes me think that Sun
very well may raise/handle signals somewhere in its Java code.
See also
http://www.google.com/search?q=%22Signal.java%22+%2B%22package+sun.misc%22

Signature
"Thomas:Fritsch$ops:de".replace(':','.').replace('$','@')