[jcifs] Possible deadlock?

[Michael B Allen]

On Sat, 19 Aug 2006 11:54:58 -0400
Jake Goulding <goulding at vivisimo.com> wrote:

> Thanks for the good information!
> 
> However, I happen to be running this particular implementation  
> through the JVM invocation interface. The crawler itself is written  
> in C, and I call through to the Java code.

If you're using C, libsmbclient might be better.

> I'm fairly certain that  
> the CTRL-\ will not work through that layer.

A threaddump is triggered when a SIGQUIT is received. Ctrl-\ is just
the term escape for sending a sigquit. If you're using C just do kill(0,
SIGQUIT).

> Could this possibly cause my problems?

If the C crawler is using threads that translate into threads in Java
then I would think there should be no difference in behavior. If the
C crawler is creating multiple processes it's possible that you could
simply be exhausting JVM memory. In general, with JNI you open up a whole
new possibility for non-deterministic behavior. But I don't know of any
specific reason why it would cause a problem.

Also, try one of the crawlers in the examples directory to see if you
get the same behavior.

Mike

-- 
Michael B Allen
PHP Active Directory SSO
http://www.ioplex.com/