Issue related to backport JDK-7092821

[Shi, Tianmin]

Hi Everyone,

I haven’t get any response yet for my previous email. so I would like to raise this issue again and also attached a small code sample to repo the issue.
This sample code works fine in JDK8 but print out this line in JDK13 due to the race condition "CertificateFactory.Fixed is NULL"

Thank you,



From: "Shi, Tianmin" <tianshi at amazon.com>
Date: Tuesday, September 17, 2019 at 12:46 PM
To: "security-dev at openjdk.java.net" <security-dev at openjdk.java.net>
Subject: Issue related to backport JDK-7092821


Hi Everyone,
  This is Tianmin Shi from the Amazon Corretto team. I am new to JDK development, please let me know if I missed anything.
  I am working on backport JDK-7092821<https://bugs.openjdk.java.net/browse/JDK-7092821> to 8u and 11u. however I think there is a potential race condition introduced by the change, which could cause method getService in class java.security.Provider to return incorrect results. Which is due to the read of legacyMap at line 1260 is not synchronized with clear and repopulate of the same map’s entries inside ensureLegacyParsed(), the fix is a simpler one liner in the attached webrev. Can we push this change to tip so I can backport it with JDK-7092821<https://bugs.openjdk.java.net/browse/JDK-7092821> together?

Thank you




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/security-dev/attachments/20190923/b3b78e7f/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Main.java
Type: application/octet-stream
Size: 1919 bytes
Desc: Main.java
URL: <https://mail.openjdk.org/pipermail/security-dev/attachments/20190923/b3b78e7f/Main.java>