[10] RFR (XS): 8180366: [TESTBUG] gc/g1/humongousObjects/TestHumongousClassLoader should not be run with class unloading disabled

Kevin Walls kevin.walls at oracle.com
Thu Aug 10 11:57:49 UTC 2017


Hi Muthu -

There is a catch of course!  8-)  There we are making it require the 
explicit option -XX:+ClassUnloading, and would this make it skip the 
test if that option isn't set, even though it might be the default 
behaviour.  I think we really want:

43  * @requires !(vm.opt.ClassUnloading  == false)

..which jtreg understands, skipping the test if we explicitly disabled 
ClassUnloading, running if we don't specify it, or if we explicity 
enable ClassUnloading.

Thanks
Kevin

The operators are mentioned here: 
http://openjdk.java.net/jtreg/tag-spec.html#requires_names




On 09/08/2017 12:36, Muthusamy Chinnathambi wrote:
> Hi,
>
> Please review this trivial change for the fix of bug 'JDK-8180366 [TESTBUG] gc/g1/humongousObjects/TestHumongousClassLoader should not be run with class unloading disabled'
>
> Summary:
> This testcase(TestHumongousClassLoader.java) checks that unreachable classes and unreachable humongous class loader are unloaded after GC.  Hence we need to have class unloading enabled for the correctness of the test. The fix is to prohibit execution of the test if class unloading is disabled.
>
> jdk10 bug: https://bugs.openjdk.java.net/browse/JDK-8180366
> webrev link: http://cr.openjdk.java.net/~shshahma/Muthu/8180366/webrev.00/
>
> Testing: I have manually confirmed that jtreg will no longer run the test if -XX:-ClassUnloading is used.
>
> Regards,
> Muthusamy C




More information about the hotspot-gc-dev mailing list