RFR: 8022547: [verifier] move DefaultMethodRegressionTests.java to jdk

Kumar Srinivasan kumar.x.srinivasan at oracle.com
Sun Aug 11 00:25:06 UTC 2013


Hi Karen,

> Kumar,
>
> Looks good. And thank you - we definitely needed the URLClassLoader so that we
> actually test the verifier, unless we had a force -Xverify:all flag. Thank you for
> fixing the major version from 0x33 ->0x34.
You are welcome.

During jprt testing there was an issue with some debugging code I had in 
there, this turned
out to be a problem on Windows, as I was not closing the 
DirectoryStream. I have removed
that code and the updated webrev is here:
http://cr.openjdk.java.net/~ksrini/8022547/webrev.1/

to make it easier, the delta webrev showing only the changes since the 
last is here:
http://cr.openjdk.java.net/~ksrini/8022547/webrev.1/webrev.delta/index.html

>
> I presume the push will also delete the file from the current location :-)

Of course :-) , however since those are in langtools repository and 
already in ignored state,
I will get that done separately.

Thanks
Kumar

>
> thanks,
> Karen
>
> On Aug 7, 2013, at 8:25 PM, Kumar Srinivasan wrote:
>
>> Hello,
>>
>> Testing this functionality in langtools does not seem to be the appropriate location,
>> and the teams have decided to move it to jdk/test/vm/verifier, which seems to be the
>> logical place.
>>
>> I have modified the test to remove testng dependencies, also added a variant which
>> involves loading classes via an URLClassLoader which exposed a test bug wrt. the
>> class.version.major not being correct, therefore it is useful to have this variant anyway.
>>
>> The webrev:
>> http://cr.openjdk.java.net/~ksrini/8022547/webrev.0/
>>
>> The bug:
>> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8022547
>>
>> Thanks
>> Kumar
>>
>>
>>
>>




More information about the core-libs-dev mailing list