RFR: 8079080: ConcurrentMark::mark_stack_push(oop) is unused

Kim Barrett kim.barrett at oracle.com
Thu Apr 30 18:31:06 UTC 2015


On Apr 30, 2015, at 3:57 AM, Thomas Schatzl <thomas.schatzl at oracle.com> wrote:
> 
> Hi Kim,
> 
> On Wed, 2015-04-29 at 19:42 -0400, Kim Barrett wrote:
>> Please review this cleanup of ConcurrentMark and CMMarkStack.
>> 
>> The following functions were unused, so are being removed:
>> 
>> - ConcurrentMark::mark_stack_push(oop) - the other overload is used and retained.
>> - CMMarkStack::push(oop)
>> - CMMarkStack::par_push(oop)
>> - CMMarkStack::par_adjoin_arr(oop*, int)
>> 
>> In addition, CMMarkStack::pop() is only used in the implementation of
>> CMMarkStack::drain(), so is being made private.
>> 
>> CR:
>> https://bugs.openjdk.java.net/browse/JDK-8079080
>> 
>> Webrev:
>> http://cr.openjdk.java.net/~kbarrett/8079080/webrev.00/
>> 
> 
> while cleaning up there, could you also remove the following places:
> 
> - "friend class CMGlobalObjectClosure" - there is no
> CMGlobalObjectClosure class
> 
> - CMMarkStack::_max_stack can be removed. While the member is updated in
> non-product builds, the value is never used anywhere.
> 
> - CMMarkStack::isFull() is not referenced
> 
> - maybe in the declaration of ConcurrentMark::_remark_mark_times and
> ConcurrentMark::_remark_weak_ref_times the extra space could be removed.

Yay!  More code to delete!

Webrev:
http://cr.openjdk.java.net/~kbarrett/8079080/webrev.01/

Incremental webrev:
http://cr.openjdk.java.net/~kbarrett/8079080/webrev.01.incr/




More information about the hotspot-gc-dev mailing list