GC Benchmarks Suite
----------------------------------------------------------------------------------------------------------
# JMH 1.19 (released 173 days ago)
# JDK 10-internal, VM 10-internal+0-adhoc.rkennke.shenandoah-jdk10
# Linux, amd64, 3.10.0-693.5.2.el7.x86_64
These tests assess the garbage collector performance. These tests implicitly test
other aspects of system under test, including hardware, OS and JVM tuning. Some tests
may implicitly test the testing infrastructure itself, although the benchmark code
tries to reduce the infrastructural impact as much as possible.
If you are sharing this report, please share it in full, including the JVM version,
OS flavor and version, plus some data on used hardware.
Use -h to get help on available options.
Available tests:
alloc.peak.object
alloc.peak.intarray
alloc.peak.refarray
alloc.uninit.intarray
retain.array
retain.linkedlist
retain.tree
retain.hashmap
fragger.array
fragger.tree
fragger.linkedlist
sieve.object
roots.strings
roots.synchronizers
roots.locals
runtime.reads.plain
runtime.reads.volatile
runtime.writes.plain
runtime.writes.volatile
Matching tests:
runtime.reads.volatile
===== runtime.reads.volatile
Estimates the runtime overhead for reads. These costs include all the infrastructural
overheads too, and therefore the score differences matter, not their absolute values.
This test is single-threaded. Reads volatile Java fields.
Test dimensions:
Mode: available JVM modes
Type: Java types
Mode Type Performance Cycles Instructions Branches
-Xint none 588,06 ± 19,46 ns/op - - -
-Xint boolean 587,80 ± 5,77 ns/op - - -
-Xint byte 585,89 ± 17,44 ns/op - - -
-Xint short 590,48 ± 13,53 ns/op - - -
-Xint char 590,67 ± 11,62 ns/op - - -
-Xint int 590,78 ± 21,20 ns/op - - -
-Xint float 593,47 ± 16,92 ns/op - - -
-Xint long 595,52 ± 21,51 ns/op - - -
-Xint double 600,57 ± 60,99 ns/op - - -
-Xint Object 595,50 ± 18,99 ns/op - - -
-XX:TieredStopAtLevel=1 none 6,92 ± 0,79 ns/op - - -
-XX:TieredStopAtLevel=1 boolean 7,64 ± 0,05 ns/op - - -
-XX:TieredStopAtLevel=1 byte 7,70 ± 0,41 ns/op - - -
-XX:TieredStopAtLevel=1 short 8,45 ± 6,99 ns/op - - -
-XX:TieredStopAtLevel=1 char 7,60 ± 0,16 ns/op - - -
-XX:TieredStopAtLevel=1 int 7,61 ± 0,09 ns/op - - -
-XX:TieredStopAtLevel=1 float 7,57 ± 0,34 ns/op - - -
-XX:TieredStopAtLevel=1 long 7,75 ± 0,27 ns/op - - -
-XX:TieredStopAtLevel=1 double 7,71 ± 1,19 ns/op - - -
-XX:TieredStopAtLevel=1 Object 7,66 ± 0,07 ns/op - - -
-XX:TieredStopAtLevel=4 none 8,50 ± 0,50 ns/op - - -
-XX:TieredStopAtLevel=4 boolean 8,87 ± 0,41 ns/op - - -
-XX:TieredStopAtLevel=4 byte 9,02 ± 1,23 ns/op - - -
-XX:TieredStopAtLevel=4 short 8,80 ± 0,02 ns/op - - -
-XX:TieredStopAtLevel=4 char 8,91 ± 0,92 ns/op - - -
-XX:TieredStopAtLevel=4 int 8,80 ± 0,03 ns/op - - -
-XX:TieredStopAtLevel=4 float 8,81 ± 0,09 ns/op - - -
-XX:TieredStopAtLevel=4 long 8,84 ± 0,38 ns/op - - -
-XX:TieredStopAtLevel=4 double 8,80 ± 0,04 ns/op - - -
-XX:TieredStopAtLevel=4 Object 9,44 ± 1,18 ns/op - - -
Test took 1090 seconds.