RFR 8003421: NPG: Move oops out of InstanceKlass into mirror

[Christian Thalinger]

On May 21, 2013, at 5:24 AM, Coleen Phillimore <coleen.phillimore at oracle.com> wrote:

> 
> On 05/21/2013 01:12 AM, David Holmes wrote:
>> On 21/05/2013 2:50 PM, Dean Long wrote:
>>> Right, unless there is a way to inject them conditionally.
>> 
>> You can inject them conditionally based on the JRE version, but then why bother moving to the Java level if you will still need the code in the VM ?
> 
> Absolutely.   That's why I don't want to move them to the Java level.   I'd have to add 90% of the same code and then they'd be available to Java but they're used as much in the VM than in Java. And I think protection_domain shouldn't be visible to the Java code for security reasons.   So I see no point to doing it.
> 
>> 
>>> On 5/20/2013 9:02 PM, Ioi Lam wrote:
>>>> But if you move these fields into Class.java (in JDK8), then hsx25
>>>> will not run on JDK7 anymore, unless these fields are also added in
>>>> Class.java in JDK7.
>> 
>> As I understand it hs25 is already limited in the versions of jdk7 that it can run in.
> 
> Yes, I think it still works for now.

Correct but it's not a supported configuration.  It can start to fail anytime and very likely already fails for particular applications.

-- Chris

> 
> Coleen
>> 
>> David
>> -----
>> 
>>>> 
>>>> - Ioi
>>>> 
>>>> On 05/20/2013 05:42 PM, Dean Long wrote:
>>>>> It seems like you could take this opportunity to make these declared
>>>>> fields of java.lang.Class,
>>>>> allowing, for example, getProtectionDomain0() to be a simple Java
>>>>> method instead of a native method.
>>>>> 
>>>>> dl
>>>>> 
>>>>> On 05/20/2013 03:39 PM, Coleen Phillimore wrote:
>>>>>> Summary: Inject protection_domain, signers, init_lock into
>>>>>> java_lang_Class
>>>>>> 
>>>>>> Net footprint change is zero except that these fields are in Java
>>>>>> heap rather than metaspace.  This helps a little with InstanceKlass
>>>>>> size which is in fixed size space with UseCompressedKlassPointers.
>>>>>> Included serviceability because there were SA changes to code that I
>>>>>> don't know is used.
>>>>>> 
>>>>>> Future work is to remove the signers field and the unused
>>>>>> SetProtectionDomain function.
>>>>>> 
>>>>>> open webrev at http://cr.openjdk.java.net/~coleenp/8003421/
>>>>>> bug link at http://bugs.sun.com/view_bug.do?bug_id=8003421
>>>>>> 
>>>>>> Tested with vm.quick.testlist, JPRT, jtreg java/security tests and
>>>>>> jck8 tests.
>>>>>> 
>>>>>> Thanks,
>>>>>> Coleen
>>>>> 
>>>> 
>>> 
>