The program in the comments section, courtesy of ###@###.###,
throws an OOM when it should not. The bug seems to have first appeared
in 6u18 (hs16), and is present in all subsequent releases. As far as
I can tell the bug is not present in releases prior to and including
6u17 (hs14.3).
I am adding another program, courtesy of ###@###.### / ###@###.###,
where WeakRefs are not cleared when their referents are not reachable. The program is in the
comments section. In this case, reference processing reports that the referents are unreachable
and that it has cleared the referents, yet the subsequent get() returns a non-NULL reference.
throws an OOM when it should not. The bug seems to have first appeared
in 6u18 (hs16), and is present in all subsequent releases. As far as
I can tell the bug is not present in releases prior to and including
6u17 (hs14.3).
I am adding another program, courtesy of ###@###.### / ###@###.###,
where WeakRefs are not cleared when their referents are not reachable. The program is in the
comments section. In this case, reference processing reports that the referents are unreachable
and that it has cleared the referents, yet the subsequent get() returns a non-NULL reference.
- relates to
-
JDK-6858496 Clear all SoftReferences before reaching the GC overhead limit.
-
- Closed
-