RFR: 8284930: Re-examine FilterInputStream mark/reset [v2]

Joe Darcy darcy at openjdk.java.net
Wed Apr 20 20:05:25 UTC 2022


On Wed, 20 Apr 2022 18:40:13 GMT, Alan Bateman <alanb at openjdk.org> wrote:

> Thanks for the update. I can't think how anything could depend on this but unfortunately it's been like that since early JDK releases. We will need a release note, I'm sure Joe will give his opinion on whether a CSR should be created.

Hmm. It is true that the presence or absence of the "synchronized" modifier is not part of a method's contract. However, the multi-thread safety, or not, of a class is part of its behavioral contract.

Without a lot of investigation, my current inclination is to run a CSR for this change for the (small) behavioral compatibility hazard.

-------------

PR: https://git.openjdk.java.net/jdk/pull/8309


More information about the core-libs-dev mailing list