Request for review (S): 8006398: Add regression tests for deprectated GCs

Jon Masamitsu jon.masamitsu at oracle.com
Thu Feb 14 20:06:22 UTC 2013


On 2/14/2013 6:32 AM, Bengt Rutisson wrote:
>
> Hi Jon,
>
> Are you ok with me pushing this? We can always add more tests later.

Your call.  I myself would do it now so I don't have to think about it
again but that's because I so easily forget about things :-)

Ship it.

Jon
>
> Thanks,
> Bengt
>
> On 2/13/13 9:11 AM, Bengt Rutisson wrote:
>>
>> Hi John,
>>
>> Thanks for looking at this!
>>
>> On 2/13/13 3:40 AM, Jon Masamitsu wrote:
>>> Bengt,
>>>
>>> Changes look good.
>>>
>>> You include tests for UseG1GC and UseParallelGC (should
>>> not print warnings) but not UseConcMarkSweepGC nor
>>> UseSerialGC.  Why the former but not the latter?
>>
>> The TestSerialGC.java is running with UseSerialGC. Am I missing 
>> something or is this the test you meant?
>>
>> For CMS I think you have a point. I was kind of thinking that the 
>> test TestParNewCMS was anyway testing that GC combination, but we 
>> should really have a separate CMS test as well. I just added 
>> TestCMS.java to test UseConcMarkSweepGC explicitly. Here is an 
>> updated webrev:
>>
>> http://cr.openjdk.java.net/~brutisso/8006398/webrev.01/
>>
>>>
>>> Do you want to test
>>>
>>> -XX:+UseConcMarkSweepGC -XX:-CMSIncrementalMode -version
>>>
>>> to check that no warning is emitted?
>>
>> Hm. I don't know. If we start combining all the flags we will get 
>> into an explosion of tests. I think this is a valid combination to 
>> test. But in that case so are things like "-XX:+UseConcMarkSweepGC 
>> -XX:-UseSerialGC". I like the way you think about this testing, I 
>> just think I would like to keep these tests to be smoke testing the 
>> more commonly used combinations.
>>
>> Make sense?
>>
>> Thanks again for looking at this!
>> Bengt
>>
>>>
>>> Jon
>>>
>>> On 2/12/2013 4:56 AM, Bengt Rutisson wrote:
>>>>
>>>> Hi again,
>>>>
>>>> Christian has pushed his test library change and it has propagated 
>>>> to the hotspot-gc repository. So, this change is ready to be 
>>>> pushed. Just need some reviews. Any takers? :)
>>>>
>>>> Thanks,
>>>> Bengt
>>>>
>>>> On 1/16/13 1:58 PM, Bengt Rutisson wrote:
>>>>>
>>>>> Hi all,
>>>>>
>>>>> Could I have a couple of reviews for this change?
>>>>> http://cr.openjdk.java.net/~brutisso/8006398/webrev.00/
>>>>>
>>>>> Recently we deprecated some GC combinations. Those should now 
>>>>> print a warning at startup. Other GC combinations should not print 
>>>>> any warnings.
>>>>>
>>>>> With the new process handling support that Christian Törnqvist is 
>>>>> adding to the JTREG tests for hotspot it is very easy to write 
>>>>> test that start a VM and checks the output.
>>>>>
>>>>> This changes makes use of Christian's testlibrary to verify that 
>>>>> warnings are printed as expected.
>>>>>
>>>>> I'm also adding the "gc" keyword to JTREG to make it possible to 
>>>>> filter out GC tests. We should probably use this for all test in 
>>>>> the the /gc folder, but I think that should be done as a separate 
>>>>> change.
>>>>>
>>>>> The webrev above is based on Christian's webrev to add the 
>>>>> testlibrary:
>>>>> http://cr.openjdk.java.net/~brutisso/8006413/webrev.00/
>>>>>
>>>>> Thanks,
>>>>> Bengt
>>>>>
>>>>
>>>>
>>>
>>
>




More information about the hotspot-gc-dev mailing list