回复：EA builds with changes for JDK 24

[何品(虎鸣)]

Thanks for the sharing, but in the referenced https://openjdk.org/jeps/8337395 <https://openjdk.org/jeps/8337395 > & https://bugs.openjdk.org/browse/JDK-8337395 <https://bugs.openjdk.org/browse/JDK-8337395 >
The Classloading do cause deadlock in our case, where both a Virtual thread and Platform thread try to loading a class and the virtual thread is currently have no chance to run.
And there is a blog at: https://craft.faire.com/java-virtual-threads-increasing-search-performance-while-avoiding-deadlocks-f12fa296d521 <https://craft.faire.com/java-virtual-threads-increasing-search-performance-while-avoiding-deadlocks-f12fa296d521 > too
Hope these cases been addressed soon.
------------------------------------------------------------------
发件人：Alan Bateman <alan.bateman at oracle.com>
发送时间：2024年9月4日(星期三) 16:34
收件人："loom-dev"<loom-dev at openjdk.org>
主　题：EA builds with changes for JDK 24
The builds at the Project Loom Early-Access Builds page [1] have been 
refreshed.
These builds contain most of the changes that this project would like to 
propose for JDK main line and JDK 24.
The builds include the changes to eliminate the pinning issues with 
object monitors. Thank you to those that tested with the previous builds 
from a few months ago. It would be great to get some fresh testing 
before the changes are proposed for main line. The changes are 
significant and a JEP [2] has been submitted to put this update on the 
technical roadmap.
Structured concurrency. The proposal is for StructuredTaskScope [3] to 
continue to preview. The API has been updated, the most significant 
change is the move from implementing policy in a sub-class to 
implementing it an object that you specify to the static open method. 
There will be a JEP. For this API, the greatest need is feedback from 
real-world usage, probably cases that fan-out to remote services or 
databases and experiences where there is handling of outcome and exceptions.
Scoped Values. Andrew Haley already sent a note [4] about proposing 
ScopedValue [5] continue to preview with a small update. As before, the 
need here is feedback from real-world usage. This could be insight into 
how many scoped values are bound or insight on which methods are used in 
performance critical code.
-Alan
[1] https://jdk.java.net/loom/ <https://jdk.java.net/loom/ >
[2] https://openjdk.org/jeps/8337395 <https://openjdk.org/jeps/8337395 >
[3] 
https://download.java.net/java/early_access/loom/docs/api/java.base/java/util/concurrent/StructuredTaskScope.html <https://download.java.net/java/early_access/loom/docs/api/java.base/java/util/concurrent/StructuredTaskScope.html >
[4] https://mail.openjdk.org/pipermail/loom-dev/2024-August/007133.html <https://mail.openjdk.org/pipermail/loom-dev/2024-August/007133.html >
[5] 
https://download.java.net/java/early_access/loom/docs/api/java.base/java/lang/ScopedValue.html <https://download.java.net/java/early_access/loom/docs/api/java.base/java/lang/ScopedValue.html >
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/loom-dev/attachments/20240905/008aaa78/attachment-0001.htm>