-
Bug
-
Resolution: Fixed
-
P3
-
19
-
b29
Issue | Fix Version | Assignee | Priority | Status | Resolution | Resolved In Build |
---|---|---|---|---|---|---|
JDK-8289417 | 20 | Igor Veresov | P3 | Resolved | Fixed | b04 |
JDK-8291356 | 19.0.2 | Igor Veresov | P3 | Resolved | Fixed | b01 |
JDK-8291202 | 19.0.1 | Igor Veresov | P3 | Resolved | Fixed | b04 |
Seeing this now with JDK 19:
```
$ ~/trunks/jdk/build/linux-x86_64-server-release/images/jdk/bin/java -jar tests-chapter-2a/target/jcstress.jar -m sanity -af NONE -sc false -t copy --jvmArgs "-XX:TieredStopAtLevel=1"
..... [TIMEOUT] o.o.j.t.copy.manual.arrays.large.volatiles.ShortTest
Scheduling class:
actor1: package group free, core group free
actor2: package group free, core group free
CPU allocation: unspecified
Compilation: unified across all actors
JVM args: [-XX:TieredStopAtLevel=1]
Fork: #1
Messages:
Timeout waiting for tasks to complete: 30013 ms
Have stale threads, forcing VM to exit for proper cleanup.
```
The tests eventually succeed, if we bump the timeouts to much larger values (requires jcstress recompilation, unfortunately).
JDK 17 does these tests in 5 seconds. JDK 19 does these tests in 3.5 minutes, that is more than 40x slower.
```
$ ~/trunks/jdk/build/linux-x86_64-server-release/images/jdk/bin/java -jar tests-chapter-2a/target/jcstress.jar -m sanity -af NONE -sc false -t copy --jvmArgs "-XX:TieredStopAtLevel=1"
..... [TIMEOUT] o.o.j.t.copy.manual.arrays.large.volatiles.ShortTest
Scheduling class:
actor1: package group free, core group free
actor2: package group free, core group free
CPU allocation: unspecified
Compilation: unified across all actors
JVM args: [-XX:TieredStopAtLevel=1]
Fork: #1
Messages:
Timeout waiting for tasks to complete: 30013 ms
Have stale threads, forcing VM to exit for proper cleanup.
```
The tests eventually succeed, if we bump the timeouts to much larger values (requires jcstress recompilation, unfortunately).
JDK 17 does these tests in 5 seconds. JDK 19 does these tests in 3.5 minutes, that is more than 40x slower.
- backported by
-
JDK-8289417 Very slow C1 arraycopy jcstress tests after JDK-8279886
-
- Resolved
-
-
JDK-8291202 Very slow C1 arraycopy jcstress tests after JDK-8279886
-
- Resolved
-
-
JDK-8291356 Very slow C1 arraycopy jcstress tests after JDK-8279886
-
- Resolved
-
- duplicates
-
JDK-8290423 [TIMEOUT] o.o.j.t.init.fields.volatiles.LongTest
-
- Closed
-
- relates to
-
JDK-8279886 C1: Turn off SelectivePhiFunctions in presence of irreducible loops
-
- Resolved
-
(2 links to)