Details
-
Enhancement
-
Resolution: Fixed
-
P4
-
11-shenandoah, 14, 15
-
b13
Backports
Issue | Fix Version | Assignee | Priority | Status | Resolution | Resolved In Build |
---|---|---|---|---|---|---|
JDK-8241538 | 14.0.2 | Aleksey Shipilev | P4 | Resolved | Fixed | b01 |
Description
It is currently statically driven by:
experimental(uint, ShenandoahParallelSafepointThreads, 4, \
"Number of parallel threads used for safepoint prolog/epilog") \
This is troublesome either way:
*) For small machines or containers where ergonomics figures out ParallelGCThreads=1, this trashes safepoint cleanup with excessive threads;
*) For large machines where ergonomics figures out ParallelGCThreads >> 4, this holds back parallelism needed at cleanup.
Should be just driven by ParallelGCThreads. ZGC drives this with ParallelGCThreads. G1 implicitly does the same.
experimental(uint, ShenandoahParallelSafepointThreads, 4, \
"Number of parallel threads used for safepoint prolog/epilog") \
This is troublesome either way:
*) For small machines or containers where ergonomics figures out ParallelGCThreads=1, this trashes safepoint cleanup with excessive threads;
*) For large machines where ergonomics figures out ParallelGCThreads >> 4, this holds back parallelism needed at cleanup.
Should be just driven by ParallelGCThreads. ZGC drives this with ParallelGCThreads. G1 implicitly does the same.
Attachments
Issue Links
- backported by
-
JDK-8241538 Shenandoah: parallel safepoint workers count should be ParallelGCThreads
-
- Resolved
-
- relates to
-
JDK-8225229 Shenandoah: trim down default number of GC threads
-
- Resolved
-