RFR(XS): 8144536: Clean up Unified Logging test directory

Rachel Protacio rachel.protacio at oracle.com
Thu Dec 3 18:05:38 UTC 2015


Thank you, David!

On 12/3/2015 1:03 AM, David Holmes wrote:
> Hi Rachel,
>
> On 3/12/2015 6:45 AM, Rachel Protacio wrote:
>> Hi,
>>
>> I have a short cleanup of the test/runtime/logging directory,
>> consolidating two options' tests into one file each. Also improved the
>> @run directives as per the recent emails on the topic.
>
> Test consolidation seems fine.
>
> What is the story with the use of "driver"? IIRC it means that none of 
> the VM args supplied to the jtreg invocation are passed through - 
> right? Which suggests to me that we may not be testing the logging 
> adequately - for example if the default VM is server and we never pass 
> through -client, and there are log statements specific to client code 
> paths (or a specific gc or something else controlled by a VM option) 
> then we never test them. May not be applicable for safepoint logging 
> or VMOperation logging, but it means we need to carefully consider 
> when "driver" make sense to be used.
>
That fits with my understanding of it. I'll make sure to be mindful of 
the context with future logging options.
R
> Thanks,
> David
>
>> Open webrev: http://cr.openjdk.java.net/~rprotacio/8144536/
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8144536
>> Testing: The tests still succeed with jtreg.
>>
>> Thank you!
>> Rachel



More information about the hotspot-runtime-dev mailing list