Bug ID: JDK-8253064 monitor list simplifications and getting rid of TSM

JDK-8253064 : monitor list simplifications and getting rid of TSM

Type: Enhancement
Component: hotspot
Sub-Component: runtime
Affected Version: 16

Priority: P4
Status: Resolved
Resolution: Fixed

Submitted: 2020-09-11
Updated: 2022-03-28
Resolved: 2020-11-11

Versions (Unresolved/Resolved/Fixed)

The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.

JDK 16
16 b24Fixed

Related Reports

Duplicate :	JDK-8236273 - clarify use and role of clear() and Recycle() in ObjectMonitor
Duplicate :	JDK-8238371 - ObjectSynchronizer::om_alloc() does not have to be public
Duplicate :	JDK-7021979 - rapid sustained monitor circulation causes asymptotic increase in # of extant monitors
Duplicate :	JDK-8134151 - Tune object monitor provisioning
Duplicate :	JDK-8238370 - ObjectSynchronizer::om_release() could be simplified
Relates :	JDK-8263864 - ~17% regression for DaCapo h2 after JDK-8253064
Relates :	JDK-8281450 - Remove unnecessary operator new and delete from ObjectMonitor
Relates :	JDK-8256237 - Zero: non-PCH build fails after JDK-8253064
Relates :	JDK-8257975 - Regression ~ 2.5% in SwingMark.Table after JDK-8253064
Relates :	JDK-8259008 - ArithmeticException was thrown at "Monitor Cache Dump" on HSDB
Relates :	JDK-8259349 - -XX:AvgMonitorsPerThreadEstimate=1 does not work right
Relates :	JDK-8256303 - revisit ObjectMonitor padding between fields
Relates :	JDK-8247281 - migrate ObjectMonitor::_object to OopStorage
Relates :	JDK-8256301 - ObjectMonitor::is_busy() should return bool
Relates :	JDK-8256306 - ObjectMonitor::_contentions field should not be 'jint'
Relates :	JDK-8256307 - cleanup AvgMonitorsPerThreadEstimate and _in_use_list_ceiling types
Relates :	JDK-8256241 - replace MonitorDeflation_lock with a semaphore
Relates :	JDK-8283674 - Pad ObjectMonitor allocation size to cache line size
Relates :	JDK-8257521 - runtime/logging/MonitorInflationTest.java crashed in MonitorList::unlink_deflated
Relates :	JDK-8273107 - RunThese24H times out with "java.lang.management.ThreadInfo.getLockName()" is null
Relates :	JDK-8254335 - logging/logStream.hpp includes memory/resourceArea.hpp but doesn't need it
Relates :	JDK-8256239 - monitorDeflationThread.cpp missing include of universe.hpp
Relates :	JDK-8225631 - Consider replacing muxAcquire/Release with PlatformMonitor
Relates :	JDK-8226416 - MonitorUsedDeflationThreshold can cause repeated async deflation requests
Relates :	JDK-8256304 - should MonitorUsedDeflationThreshold be experimental or diagnostic
Relates :	JDK-8277180 - Intrinsify recursive ObjectMonitor locking for C2 x64 and A64
Relates :	JDK-8281043 - Intrinsify recursive ObjectMonitor locking for PPC64
Relates :	JDK-8256302 - releasing oopStorage when deflating allows for faster deleting

Sub Tasks

JDK-8254029 :

ObjectMonitor cleanup/minor bug-fix changes extracted from JDK-8253064 - Resolved

Description

Erik O. has some ideas for:

- simplifying the ObjectMonitor lists
- getting rid of ObjectMonitor Type Stable Memory (TSM)

He developed those changes as part of his work on:

    JDK-8247281 migrate ObjectMonitor::_object to OopStorage

and I separated that work into parts. This bug covers part3
of Erik O's patch.

Comments

Baseline testing with Kitchensink8H reached the following population counts with release, fastdebug and slowdebug bits: $ grep 'async global_population=' build/linux-x86_64-normal-server-release/test-support/jtreg_closed_test_hotspot_jtreg_applications_kitchensink_Kitchensink8H_java/applications/kitchensink/monitorinflation.log | tail -1 [29397.284s][info][monitorinflation] async global_population=26543, global_in_use_count=2, global_free_count=25376, global_wait_count=8 $ grep 'async global_population=' build/linux-x86_64-normal-server-fastdebug/test-support/jtreg_closed_test_hotspot_jtreg_applications_kitchensink_Kitchensink8H_java/applications/kitchensink/monitorinflation.log | tail -1 [29084.494s][info][monitorinflation] async global_population=26543, global_in_use_count=0, global_free_count=25058, global_wait_count=6 $ grep 'async global_population=' build/linux-x86_64-normal-server-slowdebug/test-support/jtreg_closed_test_hotspot_jtreg_applications_kitchensink_Kitchensink8H_java/applications/kitchensink/monitorinflation.log | tail -1 [29332.358s][info][monitorinflation] async global_population=16510, global_in_use_count=0, global_free_count=16169, global_wait_count=2 $ grep ' max=' build/linux-x86_64-normal-server-release/test-support/jtreg_closed_test_hotspot_jtreg_applications_kitchensink_Kitchensink8H_java/applications/kitchensink/monitorinflation.log | tail -1 [29127.289s][info][monitorinflation] count=4, max=51261 $ grep ' max=' build/linux-x86_64-normal-server-fastdebug/test-support/jtreg_closed_test_hotspot_jtreg_applications_kitchensink_Kitchensink8H_java/applications/kitchensink/monitorinflation.log | tail -1 [29127.899s][info][monitorinflation] count=4, max=51627 $ grep ' max=' build/linux-x86_64-normal-server-slowdebug/test-support/jtreg_closed_test_hotspot_jtreg_applications_kitchensink_Kitchensink8H_java/applications/kitchensink/monitorinflation.log | tail -1 [29313.502s][info][monitorinflation] count=4, max=51369 ........................... jdk-16+23 .. 8253064 ........................... ------------- .. ------------ release bits: ....... 26543 ....... 51261 fastdebug bits: ... 26543 ....... 51627 slowdebug bits: ..16510 ........ 51369

12-11-2020

Changeset: 2e19026d Author: Daniel D. Daugherty <dcubed@openjdk.org> Date: 2020-11-11 16:20:11 +0000 URL: https://github.com/openjdk/jdk/commit/2e19026d

11-11-2020