|
Quick Lists
|
|
Bug ID:
|
6511756
|
|
Votes
|
0
|
|
Synopsis
|
forte_is_valid_method() should call CollectedHeap::is_in_reserved() rather than is_in()
|
|
Category
|
hotspot:jvmpi
|
|
Reported Against
|
b02
|
|
Release Fixed
|
hs11(b07),
hs10(b23) (Bug ID:2158807)
, 6u7(b01) (Bug ID:2171808)
, 7(b21) (Bug ID:2177043)
|
|
State
|
10-Fix Delivered,
Needs Verification,
bug
|
|
Priority:
|
3-Medium
|
|
Related Bugs
|
6463133
,
6500866
,
6603919
,
6641053
,
6826295
|
|
Submit Date
|
10-JAN-2007
|
|
Description
|
See comments section.
Posted Date : 2007-01-10 20:38:27.0
|
|
Work Around
|
N/A
|
|
Evaluation
|
Note: the assert affects only the GenCollectedHeap side of things.
Current GenCollectedHeap collectors do not actually have a performance
problem (despite the pessimistic assert). The assert really belongs
in CollectedHeap or specific sub-classes, for example G1CollectedHeap(?),
that have non-convex committed regions for which is_in() may be more
expensive to compute in general.
Posted Date : 2007-07-24 19:09:06.0
The assert referred to above has still been left in place,
but the is_valid_method() has been moved into a CollectedHeap
interface. The need for is_in() or is_in_reserved() from this
method is gone, and we just rely on is_permanent().
Posted Date : 2007-09-13 16:27:22.0
|
|
Comments
|
PLEASE NOTE: JDK6 is formerly known as Project Mustang
|
|
|
|
