Not that my voice has any weight, but I applaud and welcome a 3.0
release.  :)

-Toby.

On Sun, Dec 25, 2011 at 4:01 PM, Matthew Dillon
<dil...@apollo.backplane.com> wrote:
>    Hello everyone!  First, I apologize for the aborted 2.12 release.  We
>    got as far as rolling it but I decided to make a real push to try to
>    fix the occassional random seg-fault bug that we were still seeing
>    on 64-bit at the time.
>
>    The seg-fault issue has now been resolved, I posted an exhaustive
>    synopsis to the kernel@ list just a moment ago.  Basically it appears
>    to be an AMD cpu bug and not a DragonFly bug.  We don't have final
>    confirmation that it isn't a DragonFly bug because it is so sensitive
>    to %rip and %rsp values that reproducing the environment to test it
>    on other OSs (even FreeBSD) is difficult, but I'm 99% certain it's an
>    AMD bug.  Add a single NOP instruction to the end of one routine in
>    the gcc-4.4 codebase appears to work around the bug.
>
>    So moving on to rolling an official release...
>
>    (1) Through past experience we will NOT do a release during the
>        holidays!  So everyone please enjoy Christmas and New Years!
>
>    (2) I would like to call the release 3.0.  Why?  Because while
>        spending the last ~1-2 months tracking down the cpu bug a whole lot
>        of other work has gone into the kernel including major network
>        protocol stack work and major SMP work.  My contribution to the SMP
>        work was to completely rewrite the 64-bit pmap, VM object handling
>        code, and VM fault handling code, as well as some other stuff.
>
>        This has resulted in a phenominal improvement in concurrency and
>        in particular concurrent compilations or anything that takes a lot
>        of page faults.  SMP contention was completely removed from the
>        page fault path and for most VM related operations, and almost
>        completely removed from the exec*() path.
>
>        Other related work has continued to improve mysql and postgresql
>        numbers as well.
>
>    (3) Release date is as-yet undecided.  It will probably be mid-February
>        to end-February in order to synchronize with the pkgsrc 2011-Q4
>        release and give things time to settle.  The release meisters will
>        be discussing it on IRC.
>
>        I will say that there are NO serious showstoppers this time.  I'd
>        like us to take our time and make this the best release we've ever
>        done!
>
>                                                -Matt
>

Reply via email to