RFR (XS): 8140777: Make Adaptive IHOP logging information the same as JFR logging

Jon Masamitsu jon.masamitsu at oracle.com
Tue Feb 9 18:28:49 UTC 2016


Thomas,

For _last_unrestrained_young_size the title is "additional buffer size".

Is there anything more descriptive that occurs to you?  Even
(if correct) "current eden size" might be more helpful.  Let the
reader make the connection between eden size and  "additional
buffer".   I'm making this up as I go so if it doesn't make sense,
do spend time on trying to make sense of it.

Jon


On 2/9/2016 7:50 AM, Thomas Schatzl wrote:
> Hi,
>
> On Tue, 2016-02-09 at 10:03 -0500, Tom Benson wrote:
>> Hi,
>> Looks OK to me, except that durations usually seem to have an 'ms'
>> suffix (line 51).  And the copyright.
> Updated in place because all relevant lines changed anyway.
>
> Also added other missing units.
>
> http://cr.openjdk.java.net/~tschatzl/8140777/webrev
>
> Note that I am not really sure if using ms for durations in this case
> is the best way. Marking cycles are typically measured in seconds if
> they are actually interesting.
>
> The other related change is that I feel that bytes/ms is very unwieldy,
> so I kept bytes/s.
>
> Thanks,
>    Thomas




More information about the hotspot-gc-dev mailing list