Review Request for CR : 7144861 RMI activation tests are too slow (original) (raw)

Olivier Lagneau olivier.lagneau at oracle.com
Thu May 10 22:16:55 UTC 2012

Olivier Lagneau said on date 5/11/2012 12:13 AM:

Please find the second webrev with your remarks applied here: http://cr.openjdk.java.net/~olagneau/7144861/webrev.01/

In addition to the way we have agreed (see below) to handle InterruptedException in this fix, I have applied all the other requests for change. Regarding RMID.start() with outer and inner timer loops, I have kept the current structure of the code, but for sure this should be totally rewritten in a further cleanup of the code. Thanks, Olivier.

Stuart Marks said on date 5/10/2012 6:28 AM: On 5/9/12 8:26 AM, Olivier Lagneau wrote: Given that we want to push the code quickly, I don't think I should go for such a large IE cleanup for this fix, which is meant to provide better exceution speed only.

I suggest to follow Stuart's proposal first (i.e. reassert and return immediately in my code changes) , and create a dedicated new CR regarding proper handling of IE in all the rmi tests (low priority). Do you agree with this ? Yes, this seems like a sensible approach. The new CR might also cover the cleaning/unification of all the retry-repeatedly-with-time-limit loops that are spread through this code. Thanks. s'marks

More information about the core-libs-dev mailing list