CRR (XS): 7091032: G1: assert failure when NewRatio is used

Bengt Rutisson bengt.rutisson at oracle.com
Thu Sep 15 13:24:15 PDT 2011


Tony,

On 2011-09-15 22:18, Tony Printezis wrote:
> Bengt,
>
> Thanks for looking at it. Yes, I can easily make the change you 
> suggested. I also changed "overrides" -> "override" as Ramki 
> suggested. Latest webrev is here:
>
> http://cr.openjdk.java.net/~tonyp/7091032/webrev.1/

Looks great!

Bengt

>
> Tony
>
> Bengt Rutisson wrote:
>>
>> Tony,
>>
>> Thanks for fixing this.
>>
>> The change looks good to me too. I don't know how far you have gotten 
>> with the push for this, but if there is still time you could make 
>> G1YoungGenSizer::size_to_region_num() private again. I had to make it 
>> public to use it in the one place that you remove now. ;-)
>>
>> Not a big deal if you have already pushed it...
>>
>> Thanks,
>> Bengt
>>
>> On 2011-09-15 20:42, Tony Printezis wrote:
>>> All set thanks to Jon and Jonh Cuthbertson.
>>>
>>> Tony
>>>
>>> Jon Masamitsu wrote:
>>>> Looks good.
>>>>
>>>> On 09/15/11 08:59, Tony Printezis wrote:
>>>>> Hi all,
>>>>>
>>>>> I'd like to get a couple of code review for this very small change 
>>>>> to avoid an assert failure when NewRatio is explicitly set 
>>>>> (size_to_region_num() takes a size parameter in bytes, so passing 
>>>>> a region number is incorrect):
>>>>>
>>>>> http://cr.openjdk.java.net/~tonyp/7091032/webrev.0/
>>>>>
>>>>> I also changed a tty->print_cr statement into a warning to be 
>>>>> consistent with what we do in similar situations.
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Tony
>>>>>
>>



More information about the hotspot-gc-dev mailing list