[8u60] RFR of backport for 8075215: SATB buffer processing found reclaimed humongous object

Kim Barrett kim.barrett at oracle.com
Sat May 9 22:34:29 UTC 2015


Please review this 8u60 backport of the fix for JDK-8075215.  The 9
change didn't apply cleanly to 8u60, so a re-review is needed.

This set of changes are applied to the result of the in-review
backports for JDK-8048179, JDK-8076265, JDK-8069367, JDK-8075466,
JDK-8078021, and JDK-8078023.
http://comments.gmane.org/gmane.comp.java.openjdk.hotspot.gc.devel/13205

There was minor fuzzing of the changes for satbQueue.cpp and
concurrentMark.inline.hpp.

There were patch failures for concurrentMark.cpp, due to unrelated
changes in surrounding code between 8u60 and 9.

- G1RemarkThreadsClosure had _is_par member, initialization, and use
  eliminated in 9, where it is treated as always true.

- G1RemarkThreadsClosure _thread_parity member initialization was
  changed in 9, due to the elimination of SharedHeap.

CR:
https://bugs.openjdk.java.net/browse/JDK-8075215
[Note: For 9 fix, see https://bugs.openjdk.java.net/browse/JDK-8079236.]

Webrev:
http://cr.openjdk.java.net/~kbarrett/8u60-eager-reclaim/8075215/webrev.00/

Differences between JDK9 and JDK8 patches
http://cr.openjdk.java.net/~kbarrett/8u60-eager-reclaim/8075215/8075215.9to8.00.txt

JDK9 review thread:
http://comments.gmane.org/gmane.comp.java.openjdk.hotspot.gc.devel/13184



More information about the hotspot-gc-dev mailing list