It is a good cleanup and the change looks okay to me. I'll count on
your testing to verify if this fixes the intermittent problem :)
StreamPipe.run() catches InterruptedIOException and IOException and
ignores the error. Is it safe? Are the tests expected to fail in some
other way (missing output?). Should it (or waitFor method) throw a
RuntimeException instead if any such exception is thrown?
Just curious - Is this intermittent test failure easy to reproduce?
Thanks
Mandy
On 1/3/2013 6:39 PM, Stuart Marks wrote:
Hi all,
Please review these additional RMI test cleanups:
http://cr.openjdk.java.net/~smarks/reviews/7187882/webrev.0/
in service of fixing the following bug:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=7187882
The basic problem here is that tests were waiting for the JVM
subprocess to exit, but not waiting for all output from the subprocess
to be collected. The fix includes a bit of new infrastructure in RMI's
test library and adjustment of several other tests to use it.
Thanks,
s'marks
