Review request: JDK-8004928 TEST_BUG: Reduce dependence of CoreLib tests from the AWT subsystem (ver. 2) (original) (raw)

Alexey Utkin alexey.utkin at oracle.com
Thu Dec 13 11:48:47 UTC 2012

Here is new version fix: http://cr.openjdk.java.net/~uta/openjdk-webrevs/JDK-8004928/webrev.02

On 12.12.2012 22:59, Mandy Chung wrote:

On 12/12/12 7:41 AM, Alexey Utkin wrote:

Bug description: https://jbs.oracle.com/bugs/browse/JDK-8004928

Summary: test/java/io/Serializable/resolveProxyClass/NonPublicInterface.java test/java/lang/reflect/Proxy/ClassRestrictions.java The set of non-public interfaces was changed to avoid AWT dependencies. These 2 tests look for the first non-public system interface. Instead of keeping the nonPublicInterfaces list, it may be good to simplify this to use "java.util.zip.ZipConstants" which seems to be a stable one that will unlikely be removed. I would suggest to add a check that the class is non-public (Modifier.isPublic) to catch any future change to this class. Done.

Swing-stored constants were changed to locally-defined.

test/java/util/Collections/EmptyIterator.java I believe your change captures what it's intended to test (typed enumeration and rawtype enumeration). Minor comment - it might be better to move the new static variables as local variable before calling testEmptyEnumeration to be consistent with convention used in this test. There is no need to keep them static anyway. I would also take out the comment about the substitution since this test doesn't seem to have any relationship with javax.swing. Done.

test/java/util/logging/LoggingDeadlock4.java Test case was simplified to avoid AWT class loading. Negative test result was tested on early JDK7 build.

Looks good. Nit: L46 - good to break it into 2 lines. It's good that you have verified with and without the fix. Done.

On 12.12.2012 22:40, Alan Bateman wrote:

For java/io/Serializable/resolveProxyClass/NonPublicInterface.java and java/lang/reflect/Proxy/ClassRestrictions.java then it would be nice if the types used were in compact1 [1], that would avoid needing to exclude those tests. I also see the test uses sun.tools.agent.StepConstants which I don't think exists but perhaps that is intentional. The list was reduces to "java.util.zip.ZipConstants" (from compact1 profile) with "non-public" status verification.

test/java/util/Collections/EmptyIterator.java, minor nit but I think we prefer "public static" over "static public". It doesn't of course need to be public anyway. The variables were made local-final in accordance with Mandy's recommendation.

You probably saw Dan's comment about changing test/java/util/logging/LoggingDeadlock4.java, I trust you'll double check this test with an older version of the JDK that doesn't have the fix. My only comment is that line 46 is too wide. Done.

Thanks for review! -uta

More information about the core-libs-dev mailing list