Hi Mandy,
Yes, I still did need a reviewer for this, thanks. Yeah, it was a pain to
debug, but once I figured out the solution it really was fun. "Oh, I don't need
this piece anymore, or this piece, or this piece. Delete, delete, delete."
Negative code FTW!
Pushed:
http://hg.openjdk.java.net/jdk8/tl/jdk/rev/c18f28312c49
s'marks
On 1/21/13 2:15 PM, Mandy Chung wrote:
Stuart,
I think you still need a reviewer for this fix. This change looks okay to me.
The details in the bug report are useful and you must have lots of fun to
diagnose this :) - definitely not easy.
Mandy
On 1/8/2013 3:50 PM, Stuart Marks wrote:
Hi all,
Please review this fix:
http://cr.openjdk.java.net/~smarks/reviews/8005646/webrev.0/
for this bug:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8005646
Many gory details are in that bug report. TL;DR the old test jumped through
several hoops attempting to deactivate all the activated objects, but this
sometimes failed. The revised test uses a better deactivation technique. This
improves the reliability of the test, and it makes unnecessary a bunch of
infrastructure that supported the old deactivation technique.
I've also removed the RMI stubs from the checked-in sources, since they have
been generated at runtime since Java 5.
Thanks,
s'marks
