-
Enhancement
-
Resolution: Fixed
-
P4
-
14
-
b26
Currently a remembered set's occupancy is calculated from its components. This is problematic for a few reasons:
- slow on remembered sets with lots of components, which causes lots of additional code to avoid recalculation or mitigations
- may cause the total count going down temporarily when calculating during coarsening; attempting to account for this adds additional code
Particularly for updating predictions at mutator time such a temporary wrong value may cause unexpected prediction results.
- slow on remembered sets with lots of components, which causes lots of additional code to avoid recalculation or mitigations
- may cause the total count going down temporarily when calculating during coarsening; attempting to account for this adds additional code
Particularly for updating predictions at mutator time such a temporary wrong value may cause unexpected prediction results.
- blocks
-
JDK-8231579 G1's incremental calculation of region elapsed time always uses the same age group for prediction
-
- Resolved
-
- is blocked by
-
JDK-8233997 Some members of HeapRegion are not cleared in HeapRegion::hr_clear()
-
- Resolved
-
-
JDK-8233998 New young regions registered too early in collection set
-
- Resolved
-
- relates to
-
JDK-8237500 Increase the G1 eager reclaim candidate selection threshold
-
- Open
-
-
JDK-8267836 Separate eager reclaim remembered set threshold from G1RSetSparseRegionEntries
-
- Resolved
-
-
JDK-8017163 G1: Refactor remembered sets
-
- Resolved
-
(1 relates to)