The nightly service_sdk_baseline testing that uses the runregress script
has been having some problems on Linux with premature termination.
It appears that jtreg/javatest is getting into trouble when running the
test/com/sun/jdi tests, and after that the runregress script seems to be
causing the jbuild script to die terminating the entire nightly build/test
run.

This is only happening on Linux as far as I can tell, both Linux AMD64
(SUSE) 'quoddy' and Linux X86 (RedHat) 'fyi'.

I'm playing with runregress to try and prevent it from terminating it's
parent process, if that's what is happening. I'm trying to allow the
nightly build/test to continue, then investigate this jtreg/javatest crash
as a separate issue.

###@###.### 2005-07-05 10:13:52 GMT