RFR(XS): 8240592: HeapRegionManager::rebuild_free_list logs 0s for the estimated free regions before

Ivan Walulya ivan.walulya at oracle.com
Fri Mar 6 08:38:04 UTC 2020


Thanks kim!
> 
> But there's also the question of why the log message mentions the
> number of free regions at all, since the number of pre-existing free
> regions isn't important because of the abandonment.

I will remove the number of free regions from the log entry and then set back the previous ordering. 

> On 6 Mar 2020, at 02:38, Kim Barrett <kim.barrett at oracle.com> wrote:
> 
>> On Mar 5, 2020, at 5:37 AM, Ivan Walulya <ivan.walulya at oracle.com> wrote:
>> 
>> Hi all,
>> 
>> Please review a small modification to fix logging during HeapRegionManager::rebuild_free_list.
>> 
>> JBS: https://bugs.openjdk.java.net/browse/JDK-8240592
>> Webrev: http://cr.openjdk.java.net/~iwalulya/8240592/00/
> 
> I think I'd prefer the old ordering, but capture num_free_regions()
> into a variable before the abandon, and use that variable in the logging.
> 
> But there's also the question of why the log message mentions the
> number of free regions at all, since the number of pre-existing free
> regions isn't important because of the abandonment.
> 




More information about the hotspot-gc-dev mailing list