December 2022 - hotspot-jfr-dev

RFR: 8299031: JFR: Clean up jdk.management.jfr
by Erik Gahlin 20 Dec '22

20 Dec '22

Could I have a review of change that remove an unused method and add pattern matching for instanceof Thanks Erik ------------- Commit messages: - Initial Changes: https://git.openjdk.org/jdk/pull/11724/files Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=11724&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8299031 Stats: 23 lines in 4 files changed: 0 ins; 12 del; 11 mod Patch: https://git.openjdk.org/jdk/pull/11724.diff Fetch: git fetch https://git.openjdk.org/jdk pull/11724/head:pull/11724 PR: https://git.openjdk.org/jdk/pull/11724

2 2

[jdk20] RFR: 8298784: JFR: Test chunk integrity
by Erik Gahlin 20 Dec '22

20 Dec '22

Could I have a review of a test that checks that constants (thread, stack trace, classes and string) are properly handled when rotating. The test currently fails, so I have added @ignore, but it can be removed once the underlying product issues have been resolved. Thanks Erik ------------- Commit messages: - Fix whitespace - Fix index - Initial Changes: https://git.openjdk.org/jdk20/pull/43/files Webrev: https://webrevs.openjdk.org/?repo=jdk20&pr=43&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8298784 Stats: 316 lines in 1 file changed: 316 ins; 0 del; 0 mod Patch: https://git.openjdk.org/jdk20/pull/43.diff Fetch: git fetch https://git.openjdk.org/jdk20 pull/43/head:pull/43 PR: https://git.openjdk.org/jdk20/pull/43

2 2

[jdk20] RFR: 8287699: jdk/jfr/api/consumer/TestRecordingFileWrite.java fails with exception: java.lang.Exception: Found event that should not be there.
by Markus Grönlund 16 Dec '22

16 Dec '22

Greetings, the problem here addressed will close a gap in the constant pool data for threads, which up to now has been written on thread start only, but not on thread end. This leads to a race where a thread can write an event in a new chunk and terminate before the JFR recorder thread has had a chance to iterate it. By also writing constant pool data for threads on thread end, we close this gap. Testing: jdk_jfr, stress testing Thanks Markus ------------- Commit messages: - 8287699 Changes: https://git.openjdk.org/jdk20/pull/41/files Webrev: https://webrevs.openjdk.org/?repo=jdk20&pr=41&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8287699 Stats: 4 lines in 2 files changed: 3 ins; 1 del; 0 mod Patch: https://git.openjdk.org/jdk20/pull/41.diff Fetch: git fetch https://git.openjdk.org/jdk20 pull/41/head:pull/41 PR: https://git.openjdk.org/jdk20/pull/41

2 2

RFR: 8287699: jdk/jfr/api/consumer/TestRecordingFileWrite.java fails with exception: java.lang.Exception: Found event that should not be there.
by Markus Grönlund 15 Dec '22

15 Dec '22

Greetings, the problem here addressed will to close a gap in the constant pool data for threads, which up to now has been written on thread start only, but not on thread end. This leads to a race where a thread can write an event in a new chunk and terminate, before the JFR recorder thread has had a chance to iterate it. By also writing constant pool data for threads on thread end, we close this gap. Testing: jdk_jfr, stress testing Thanks Markus ------------- Commit messages: - 8287699 Changes: https://git.openjdk.org/jdk/pull/11692/files Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=11692&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8287699 Stats: 4 lines in 2 files changed: 3 ins; 1 del; 0 mod Patch: https://git.openjdk.org/jdk/pull/11692.diff Fetch: git fetch https://git.openjdk.org/jdk pull/11692/head:pull/11692 PR: https://git.openjdk.org/jdk/pull/11692

2 2

[jdk20] RFR: 8298649: JFR: RemoteRecordingStream support for checkpoint event sizes beyond u4
by Erik Gahlin 14 Dec '22

14 Dec '22

Could I have a review of a follow up fix to [JDK-8298129](https://bugs.openjdk.org/browse/JDK-8298129) to allow check point events with a size larger than u4 to be transferred over JMX. Testing: jdk/jdk/jfr An overflow happens at line 265 if v and payLoadSize is not long values: payLoadSize += (v << sizeShift); The field should be renamed to payloadSize, but I will do it separately to keep this change easy to review. Thanks Erik ------------- Commit messages: - Make v long - Initial Changes: https://git.openjdk.org/jdk20/pull/30/files Webrev: https://webrevs.openjdk.org/?repo=jdk20&pr=30&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8298649 Stats: 3 lines in 1 file changed: 0 ins; 0 del; 3 mod Patch: https://git.openjdk.org/jdk20/pull/30.diff Fetch: git fetch https://git.openjdk.org/jdk20 pull/30/head:pull/30 PR: https://git.openjdk.org/jdk20/pull/30

2 2

RFR: 8292265: Add old gen used field at G1HeapSummary JFR event.
by Sangheon Kim 09 Dec '22

09 Dec '22

Hi all, Can I have reviews for this change which adds 'old gen used' field at existing G1HeapSummary event? Testing: tier1 ~ tier3 Thanks, Sangheon ------------- Commit messages: - 8292265: Add old gen used field at G1HeapSummary JFR event Changes: https://git.openjdk.org/jdk/pull/11303/files Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=11303&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8292265 Stats: 10 lines in 4 files changed: 6 ins; 0 del; 4 mod Patch: https://git.openjdk.org/jdk/pull/11303.diff Fetch: git fetch https://git.openjdk.org/jdk pull/11303/head:pull/11303 PR: https://git.openjdk.org/jdk/pull/11303

3 8

[jdk20] RFR: 8298455: JFR: Add logging to TestClose.java
by Erik Gahlin 09 Dec '22

09 Dec '22

Hi, Could I have a review of a test fix. The test jdk/jfr/api/consumer/recordingstream/TestClose.java times out intermittently, but the thread dump looks fine. One possibility is that the onEvent action is never registered. It would be good to verify that the registration mechanism works properly. This can be done by logging with -Xlog:jfr+streaming+system=trace Testing: running test locally Thanks Erik ------------- Commit messages: - Initial Changes: https://git.openjdk.org/jdk20/pull/7/files Webrev: https://webrevs.openjdk.org/?repo=jdk20&pr=7&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8298455 Stats: 1 line in 1 file changed: 0 ins; 0 del; 1 mod Patch: https://git.openjdk.org/jdk20/pull/7.diff Fetch: git fetch https://git.openjdk.org/jdk20 pull/7/head:pull/7 PR: https://git.openjdk.org/jdk20/pull/7

2 2

RFR: JDK-8298255: JFR provide information about dynamization of number of compiler threads
by Matthias Baesken 08 Dec '22

08 Dec '22

The JVM supports both dynamic GC threads, and dynamic Compiler threads. While the GCConfiguration JFR event (see https://sap.github.io/SapMachine/jfrevents/#gcconfiguration ) seems to have at least some info about dynamization of GC threads ( field usesDynamicGCThreads), the JIT CompilerConfiguration event seems to miss this info (about JIT compiler threads dynamization), this should be added; currently the impression is given that the number is static but this is often not the case. As an extension, the dynamization process of adding / removing threads could be added as well to the JFR events but I am not sure if this is desired so I just started with a very simple patch. ------------- Commit messages: - JDK-8298255 Changes: https://git.openjdk.org/jdk/pull/11563/files Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=11563&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8298255 Stats: 4 lines in 3 files changed: 3 ins; 0 del; 1 mod Patch: https://git.openjdk.org/jdk/pull/11563.diff Fetch: git fetch https://git.openjdk.org/jdk pull/11563/head:pull/11563 PR: https://git.openjdk.org/jdk/pull/11563

4 13

RFR: 8287699: jdk/jfr/api/consumer/TestRecordingFileWrite.java fails with exception: java.lang.Exception: Found event that should not be there.
by Markus Grönlund 08 Dec '22

08 Dec '22

Greetings, The missing thread information is a result of a race that was introduced in [JDK-8277131](https://bugs.openjdk.org/browse/JDK-8277131), a race that could result in lost thread metadata information. [JDK-8289692](https://bugs.openjdk.org/browse/JDK-8289692) closed the race a bit but was not a complete fix. The proper fix, removing the race entirely, was made as part of [JDK-8293864](https://bugs.openjdk.org/browse/JDK-8293864). This change only removes the excluded test from ProblemList.txt. Testing: jdk_jfr Thanks Markus ------------- Commit messages: - 8287699 Changes: https://git.openjdk.org/jdk/pull/11588/files Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=11588&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8287699 Stats: 1 line in 1 file changed: 0 ins; 1 del; 0 mod Patch: https://git.openjdk.org/jdk/pull/11588.diff Fetch: git fetch https://git.openjdk.org/jdk pull/11588/head:pull/11588 PR: https://git.openjdk.org/jdk/pull/11588

2 2

RFR: 8298129: Let checkpoint event sizes grow beyond u4 limit
by Markus Grönlund 08 Dec '22

08 Dec '22

Greetings, This change pre-calculates the event size for checkpoint / constant pool events (event id 1) before serialization to disk and circumvents the need to reserve a limited checkpoint event size up front, to be later filled in, using a padded write of the actual size. It lets us represent small checkpoint events more effectively, but more importantly, it allows for larger checkpoint / constant pool events because the event size pre-reservation no longer restricts them. The parser side will continue to read checkpoint event sizes as varints, but will expose the checkpoint event size as a long instead of an int. Testing: jdk_jfr, stress testing Thanks Markus ------------- Commit messages: - remove min_intx - 8298129 Changes: https://git.openjdk.org/jdk/pull/11584/files Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=11584&range=00 Issue: https://bugs.openjdk.org/browse/JDK-8298129 Stats: 139 lines in 6 files changed: 84 ins; 17 del; 38 mod Patch: https://git.openjdk.org/jdk/pull/11584.diff Fetch: git fetch https://git.openjdk.org/jdk pull/11584/head:pull/11584 PR: https://git.openjdk.org/jdk/pull/11584

2 2