Sunbug 6934356: Vector.writeObject() synchronization risks serialization deadlock (original) (raw)

Alan Bateman Alan.Bateman at oracle.com
Mon Jan 10 11:56:14 UTC 2011

Neil Richards wrote:

Based on Alan Bateman's feedback for my fix for a similar problem seen in java.util.Hashtable (http://mail.openjdk.java.net/pipermail/core-libs-dev/2011-January/005601.html), I've revised the changeset once again to use diamond operators in the testcases, and to their main methods throw Exceptions directly (rather than catching them in the cause of a RuntimeException).

Also based on Dalibor Topic's suggestion (http://mail.openjdk.java.net/pipermail/core-libs-dev/2010-December/005533.html), I'm attaching the updated changeset as a webrev zip file, for ease of review. Please review this modified changeset, and let me know if anything further is required for its acceptance. Looks okay to me. One minor comment is that you might want to update writeObject's description where it currently reads "It just calls the default writeObject method". A minor nit in the tests is that the date range should be "2010, 2011," rather than "2010-2011".

-Alan.

More information about the core-libs-dev mailing list