RFR: 8078901: Add trace event for G1 MMU information

Stefan Johansson stefan.johansson at oracle.com
Tue Jun 30 11:14:34 UTC 2015


Thanks Erik!

Stefan

On 2015-06-30 10:58, Erik Helin wrote:
> On 2015-06-26, Stefan Johansson wrote:
>> Thanks Erik for looking at this,
>>
>> On 2015-06-18 10:44, Erik Helin wrote:
>>> Hi Stefan,
>>>
>>> sorry for not getting to this earlier, thanks a lot for this patch!
>>>
>>> On 2015-04-30, Stefan Johansson wrote:
>>>> Hi,
>>>>
>>>> Please review this change to add a trace event for MMU tracking:
>>>> https://bugs.openjdk.java.net/browse/JDK-8078901
>>>>
>>>> Webrev:
>>>> http://cr.openjdk.java.net/~sjohanss/8078901/hotspot.00/
>>> I have two comments:
>>> - instead of using multiple inheritance, can you instead have the
>>>    G1MMUTracer as a field in G1NewTracer and G1OldTracer? Then you can
>>>    have a getter to the field.
>>> - in G1MMUTracer, can you have the function report_mmu not take a
>>>    G1MMUTracker* but instead have two additional double parameters, one
>>>    for time_slice and one for max_gc_time? This way the tracers don't
>>>    need to know about G1MMUTracker.
>> Erik and I have discussed this a little more off-line and come to the
>> conclusion to have an all-static class handling the the reporting of the MMU
>> event. This avoids the multiple inheritance and other problems of having a
>> G1MMUTracer as part of the other tracers. This solution slightly differ from
>> previously uses of the "tracers" but still keeps the code separated in a
>> good way.
>>
>> New webrev:
>> http://cr.openjdk.java.net/~sjohanss/8078901/hotspot.01/
> Looks good, Reviewed. As we discussed offline, please revert the changes
> touching G1OldTracer and CMSTracer (no need to re-review that change).
>
> Thanks,
> Erik
>
>> Thanks,
>> Stefan
>>> Thanks!
>>> Erik
>>>
>>>> Summary:
>>>> When adding pause information to the G1MMUTracker, it now reports a
>>>> trace-event with the latest MMU information.
>>>>
>>>> add_pause() previously took a bool argument that was unused, but said
>>>> whether or not it was a concurrent-cycle pause. I removed the bool argument
>>>> and added G1MMUTracer* argument. The G1MMUTracer is an interface that both
>>>> the stw-tracer and the cm-tracer now implement to make sure the correct GC
>>>> id is provided with the event.
>>>>
>>>> Thanks,
>>>> Stefan



More information about the hotspot-gc-dev mailing list