review (XS) for 7029509: nightly failures after static fields in Class

[Daniel D. Daugherty]

On 3/21/2011 11:38 AM, Tom Rodriguez wrote:
> http://cr.openjdk.java.net/~never/7029509
>   
Thumbs up.

src/share/vm/prims/jvmtiRedefineClasses.cpp
    It looks like this fix would be applicable even in the
    absence of moving static fields.

src/share/vm/prims/jvmtiTagMap.cpp
    No comments.


src/share/vm/services/heapDumper.cpp
    No comments.

Dan


> 7029509: nightly failures after static fields in Class
> Reviewed-by:
>
> There were a few failures in hotspot-gc after the static fields in
> Classes change.  Two failures are because of missed places where the
> address of static fields were computed.  The third is because only
> real Java klasses have a mirror now so they should be skipped when
> updating the class redefinition statistics.  Tested with nsk hprof,
> heapdump and the nsk/jvmti/scenarios tests.
>