RFR (S): 8151920: Region liveness printing is broken

Thomas Schatzl thomas.schatzl at oracle.com
Tue Mar 15 20:02:16 UTC 2016


Hi all,

  can I have some reviews for a change that removes code that is not
necessary any more, and actually asserts now.

In some earlier change we made humongous and regular regions work the
same in terms of what is stored in the used/marked_bytes members.

The code for printing heap region liveness with -Xlog:gc+liveness=trace
still assumed that so some assert there fails now.

The change simply removes that code, and adds a test.

CR:
https://bugs.openjdk.java.net/browse/JDK-8151920
webrev:
http://cr.openjdk.java.net/~tschatzl/8151920/webrev/
Testing:
local compilation, testing using the test case, (checking that the
printed values with -Xlog:gc+liveness=trace are sensible).

Thanks,
  Thomas


More information about the hotspot-gc-dev mailing list