RFR: Refactoring ShControlThread::run_service() method
Aleksey Shipilev
shade at redhat.com
Thu Oct 25 18:46:41 UTC 2018
On 10/25/2018 07:06 PM, Zhengyu Gu wrote:
> On 10/25/2018 11:25 AM, Aleksey Shipilev wrote:
>> On 10/25/2018 04:55 PM, Zhengyu Gu wrote:
>>> Webrev: http://cr.openjdk.java.net/~zgu/shenandoah/cleanup_ctrl_thr/webrev.00/index.html
>>
>> Massaged patch, apply on top:
>> http://cr.openjdk.java.net/~shade/shenandoah/sh-control-thread-1.patch
>>
> Okay, as I said, it should not have functional differences. If I missed something or need further
> polish, I am all for it.
Apply the patch above over your changeset?
> This is unpolished and half tested patch on top of my original patch, just give you some ideas.
>
> http://cr.openjdk.java.net/~zgu/shenandoah/metadata_gc/webrev.01/
Okay, interesting. It does not look like requiring the refactoring, am I right? Moreover,
refactoring hides stuff from it, which you then need to reintroduce with e.g. "ShenandoahHeuristics*
heuristics = heap->heuristics();". It really looks you can introduce the metaspace OOM handling first.
-Aleksey
More information about the shenandoah-dev
mailing list