RFR: 7093328: JVMTI: jvmtiPrimitiveFieldCallback always report 0's for static primitives

Daniel D. Daugherty daniel.daugherty at oracle.com
Wed Aug 29 10:25:06 PDT 2012


On 8/29/12 11:17 AM, serguei.spitsyn at oracle.com wrote:
> On 8/29/12 7:25 AM, Daniel D. Daugherty wrote:
>> . . . <skip>
>>
>> Very nice find:
>>
>>> I did some research (archeology), the change that introduced this:
>>>
>>> changeset:   2223:c7f3d0b4570f
>>> user:        never
>>> date:        Fri Mar 18 16:00:34 2011 -0700
>>> summary:     7017732: move static fields into Class to prepare for 
>>> perm ten removal
>>
>> I had forgotten that static fields moved from the instanceKlass
>> into Class. The original code from 2005 did indeed work as expected
>> way back then. In fact, there were three such uses in jvmtiTagMap.cpp
>> back then and it looks like the other two uses were updated to use
>> the Java mirror and one was missed.
>
> Right. This explains how it did happened.

Yup. I just read through 7017732. I've added a note to this bug (7093328)
and added 7017732 to the "See also" list.

Rickard, you need to make yourself the RE for the bug and update the
evaluation...

Dan


>
> Thanks,
> Serguei
>>
>> Dan
>>
>


More information about the hotspot-runtime-dev mailing list