Hi Roger,
thanks for the comments. I added more information to the comments.
-Felix
On 2016/12/6 23:45, Roger Riggs wrote:
Hi,
Looks fine, rmid will be destroyed in the finally clause.
!! There is no description in the bug and the open comments are
unenlightening.
Attaching one of the failing logs would be have been useful.
Roger
On 12/6/2016 2:32 AM, Hamlin Li wrote:
I see, looks fine. But I'm not a reviewer.
Thank you
-Hamlin
On 2016/12/6 11:38, Felix Yang wrote:
Hi Hamlin,
as stated in the bug, the timeout is more-likely a test setup
issue that small timeout factor together with "-Xcomp".
But in theory, if not put rmid.start() in finally, it indeed
possibly leaves orphaned processes. So that is still a minor problem
to fix.
Thanks,
Felix
On 2016/12/6 11:26, Hamlin Li wrote:
Hi Felix,
As the issue is timeout at rmid.start(), so it does not resolve the
issue to just move rmid.start() to try block.
I saw the issue happened last in 2015/6, maybe we could just close
it as "Not Reproduced"?
Thank you
-Hamlin
On 2016/12/6 10:08, Felix Yang wrote:
Hi,
please review the small fix to avoid orphaned processes left.
Bug:
https://bugs.openjdk.java.net/browse/JDK-8081390
Webrev:
http://cr.openjdk.java.net/~xiaofeya/8081390/webrev.00/
Thanks,
Felix
