2009/4/9 Alan Bateman <alan.bate...@sun.com>: > Andrew John Hughes wrote: >> >> : >>> >>> icedtea-bytebuffer-compact.patch - this looks to be a partial fix to >>> 6593946. If so, 6593946 is already in jdk6/jdk6/jdk >>> (http://hg.openjdk.java.net/jdk6/jdk6/jdk/rev/3e7bfcdcf96f). >>> >>> >> >> IcedTea6 is still based on the last build drop of jdk6, not the >> Mercurial repository. So this 6 week old patch won't be in the 25th >> of November b14 tarball. >> Interestingly, it also still exists in IcedTea7 as there were cases >> covered by the patch which didn't make it into the commit. >> I've attached that version. It was originally added by Keith Seitz >> while he was working on OpenJDK at Red Hat (according to the >> ChangeLog). >> > > Once the fix gets into a build (b16?) then I don't think you'll need the > IcedTea6 patch.
Spot on, we need a new build drop (read tarball) to pick up the recent changes. There hasn't been one since Nov. 2008. Also, it does seem that the IcedTea6 patch is only a partial > fix in that I don't see the changes for the view buffers. Yeah, this is an old patch so I don't know all the origins, but it may have been pushed upstream and revised in the process, or even developed separately at Sun. One has clearMark(), one has discardMark IIRC. I looked at the > patch you attached in the last mail (is this the IcedTea patch to jdk7?). Yes, it's the leftovers after I took out the bits in upstream OpenJDK7 (last build against b50). > It > seems to be a no-op, or at least just replaces setting of mark to -1 with > calls to discardMark that does the same thing. Nothing wrong with it, just > inconsistent with the existing code. > Yeah I agree, it's not a must have. I meant to actually raise this at the time, but never got round to it obviously. Using discardMark is perhaps a little cleaner/maintainable, but has the downside of adding an extra method dispatch. > -Alan. > -- Andrew :-) Free Java Software Engineer Red Hat, Inc. (http://www.redhat.com) Support Free Java! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net PGP Key: 94EFD9D8 (http://subkeys.pgp.net) Fingerprint: F8EF F1EA 401E 2E60 15FA 7927 142C 2591 94EF D9D8
