RFR (S): 8026047: add regression test for DisableExplicitGC flag

[Tao Mao]

On 10/9/13 9:53 AM, Filipp Zhinkin wrote:
> Hi Bengt,
>
> thank you for looking at it!
>
> On 10/09/2013 04:12 PM, Bengt Rutisson wrote:
>>
>> Hi Filipp,
>>
>> A couple of comments.
>>
>> The @bug 8026047 tag refers to:
>>
>> JDK-8026047: [TESTBUG] add regression test for DisableExplicitGC flag
>>
>> Which is just the bug to implement this test. I thought the @bug tag 
>> was normally used to identify the bug which was filed to report a 
>> failure that the test is designed to verify regressions against. So, 
>> in this case I guess there is no such bug, right? There are no issue 
>> with DisableExplicitGC that I am aware of (which makes me wonder why 
>> it is important to invest time in this right now, but that is a 
>> different topic.). I would suggest to not have the @bug in this case.
> Yes, I agree.
>>
>> Why do you do 100 System.gc() calls? Isn't one enough?
> Just to be sure that particular GC didn't avoid single System.gc() for 
> some reason.
> While such case does not looks realistic at the moment, who knows what 
> could happen. :)
If that is the case, you should be glad that your test's caught 
something. :)

>>
>> Why do you need this:
>>
>>   options = options.replaceAll("-Xloggc:[^ ]*", "");
>>
>> Are you worried that the OutputAlayzer won't work if output is piped 
>> to a file?
> Yes, it is done to guarantee that GC log will be in stdout.
>> In that case shouldn't that functionality be in ProcessTools rather 
>> than in the test?
> That sounds reasonable.
>
> Thanks,
> Filipp.
>>
>> Thanks,
>> Bengt
>>
>> On 10/8/13 5:30 PM, Filipp Zhinkin wrote:
>>> Hi,
>>>
>>> I would like to get couple reviews on regression test for 
>>> DisableExplicitGC flag.
>>>
>>> It verifies that System.gc() calls did not trigger garbage 
>>> collection if DisableExplicitGC is turned on and vice versa.
>>>
>>> Bug ID: https://bugs.openjdk.java.net/browse/JDK-8026047
>>> Webrev: http://cr.openjdk.java.net/~kshefov/8026047/webrev.00/
>>> Testing: manual on local host, automated on various platforms.
>>>
>>> Thanks,
>>> Filipp.
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/hotspot-gc-dev/attachments/20131010/26ffe378/attachment.htm>