On Fri, Oct 24, 2008 at 5:40 PM, Dave Jones <[EMAIL PROTECTED]> wrote:

> A good suggestion, Tom and Scott, but a NUXCMAP shows just one instance of
> DRPC and no DDR....

Most likely the nucleus extension is written to be not reentrant and
uses a static anchor or flag to protect itself against that. If it
loads as a pipeline filter package, then it may be an attempt to
register itself for the 2nd time). If you're not doing that kind of
things (not even sure you could), then I would say that last time the
code was terminate through a code path that did not reset that flag,
and thus makes it think the previous call has not terminated yet. Does
it maybe load the nucleus extension with the SYSTEM attribute and did
you fight your way out with HX last time?

If this is nuclues extension is the only trail it left, NUCXDROP would
remove it and next invocation will load a fresh copy with no
recollection. Unless it also hooks itself into CMS, in which case you
may need to re-IPL CMS to clear up things.

-Rob (pure guessing, I have not looked at what it does)

Reply via email to