RFR: 8148755: -XX:+HeapDumpAfterFullGC creates heap dump both before and after Full GC

Claes Redestad claes.redestad at oracle.com
Wed Feb 3 13:11:01 UTC 2016


Hi Bengt,

On 2016-02-03 09:27, Bengt Rutisson wrote:
>
> Hi Claes and Sangheon,
>
> On 2016-02-03 01:04, Claes Redestad wrote:
>> Hi Sangheon,
>>
>> On 2016-02-03 00:05, sangheon wrote:
>>> Hi Claes,
>>>
>>> Looks good 
>>
>> thanks!
>
> The change looks good to me too.
>
>>
>>> and I have minor comments.
>>>
>>> 1. Line 598: 'before' should be 'after'.
>>>
>>>  597 void CollectedHeap::post_full_gc_dump(GCTimer* timer) {
>>>  598 full_gc_dump(timer, false /* before */);
>>>  599 }
>>
>> "before" here is the name of the parameter, which I thought was the 
>> convention when calling out parameter values.
>
> I know this is a convention in the code, but it is also a bit 
> confusing in this case. Since all the full_gc_dump() method is right 
> above the pre_full_gc_dump() and post_full_gc_dump() methods I would 
> suggest to just skip the /* before */ comments in this case.
>
> But I am also fine with keeping it. Your call.
>
> Reviewed.

I'll remove the comments, fix copyright and push.

Thanks!

/Claes

>
> Bengt
>
>>
>>>
>>>
>>> 2. Copyright year update.
>>
>> Deal!
>>
>> /Claes
>>
>>>
>>> Thanks,
>>> Sangheon
>>>
>>>
>>> On 02/02/2016 07:02 AM, Claes Redestad wrote:
>>>> Hello,
>>>>
>>>> this patch fixes an issue where heap is dumped both before and 
>>>> after full GC if either one of -XX:+HeapDumpAfterFullGC or 
>>>> -XX:+HeapDumpBeforeFullGC is specified.
>>>>
>>>> Bug: https://bugs.openjdk.java.net/browse/JDK-8148755
>>>> Webrev: http://cr.openjdk.java.net/~redestad/8148755/webrev.01
>>>>
>>>> Testing: JPRT, manual verification
>>>>
>>>> Thanks!
>>>>
>>>> /Claes
>>>
>>
>




More information about the hotspot-gc-dev mailing list