RFR(s) 8245060: remove copying of s.h.WB$WhiteBoxPermission in cds/appcds tests
David Holmes
david.holmes at oracle.com
Fri Jun 12 01:40:14 UTC 2020
On 12/06/2020 11:08 am, Calvin Cheung wrote:
> Hi David,
>
> Thanks for your review.
>
> On 6/11/20 5:04 PM, David Holmes wrote:
>> Hi Calvin,
>>
>> On 12/06/2020 8:43 am, Calvin Cheung wrote:
>>> JBS: https://bugs.openjdk.java.net/browse/JDK-8245060
>>>
>>> webrev: http://cr.openjdk.java.net/~ccheung/jdk16/8245060/webrev.00/
>>>
>>> Most changes involve removing the
>>> sun.hotspot.WhiteBox$WhiteBoxPermission from the following:
>>>
>>> @run driver ClassFileInstaller sun.hotspot.WhiteBox
>>> sun.hotspot.WhiteBox$WhiteBoxPermission
>>
>> That bit seems fine.
>>
>> Your copyright updates are wrong:
>>
>> - * Copyright (c) 2019, Oracle and/or its affiliates. All rights
>> reserved.
>> + * Copyright (c) 2020, Oracle and/or its affiliates. All rights
>> reserved.
>>
>> They should be
>>
>> - * Copyright (c) 2019, Oracle and/or its affiliates. All rights
>> reserved.
>> + * Copyright (c) 2019, 2020, Oracle and/or its affiliates. All rights
>> reserved.
> I did it too quick. I'll correct them.
>>
>>> Two additional cleanups:
>>> - remove the unneeded SkippedException related code from the tests
>>> under dynamicArchive/methodHandles;
>>
>> This is completly unrelated cleanup.
> Can I leave those in? Or do you prefer to have those changed in a
> separate bug?
I'd prefer to see it separated out.
>>
>>> - remove the JarBuilder.addInnerClasses method since its purpose is
>>> to add the sun.hotspot.WhiteBox$WhiteBoxPermission inner class.
>>
>> I'm not clear how this relates either. Don't we still need to list all
>> classes to be added to the jar file? If this code is unneeded then it
>> must have been unneeded for quite some time now.
>
> IIUC, the function only adds sun/hotspot/WhiteBox$WhiteBoxPermission to
> the list of classes if the list contains sun/hotspot/WhiteBox. Although
> all appcds tests passed without the function, I will revert the
> JarBuilder.java change to keep it consistent with ClassFileInstaller and
> we may need it in the future.
I think there may be something to look at here, but that would be a
separate issue.
Thanks,
David
> thanks,
>
> Calvin
>
>>
>> Thanks,
>> David
>> -----
>>
>>> Passed tier1, 2 tests.
>>
>>
>>
>>> thanks,
>>>
>>> Calvin
>>>
More information about the hotspot-runtime-dev
mailing list