Review request for OPENJDK6-35: backport of JDK-6650759 to openjdk6

[Andrew Hughes]

----- Original Message -----
> I've thought about this at length, and I've decided to leave the patch
> in.  While I shouldn't have approved it after it was rejected the
> first time around, it is in IcedTea 6, and almost everybody who uses
> OpenJDK 6 gets it via IcedTea, so it makes little difference to any of
> our users whether this patch is in OpenJDK 6 or not.

This is my thinking too.  The only benefit I can see of having it in IcedTea 1.x
and not in OpenJDK 6 is that those who want a version without the patch then
have the option of using OpenJDK 6. But it's a rather spurious argument.

> 
> I'm still rather curious about why this patch was accepted into
> IcedTea 6 but rejected by OpenJDK, but I don't suppose it really
> matters.

It was added to IcedTea first.

The reason we didn't pull it out again is that not having it is known
to cause packages to fail to build. These same packages will build on the
proprietary JDK 6 from Oracle and on 7+, so it makes OpenJDK 6 look
inferior. I can't recall seeing any issues as a result of the patch
either, but I'd prefer to resolve such issues if possible rather than
reverting the patch. It may also be worth noting that Red Hat ships
binaries based on IcedTea and OpenJDK 6 that have passed the TCK,
so the changes introduced by the patch do not cause the compatibility tests
for 6 to fail.

In the long run, it's better that users upgrade to 7+ where possible,
rather than working with the hybrid 6-7 compiler in OpenJDK 6.

> 
> Andrew.
> 
> 
> On 01/07/14 11:48, Andrew Haley wrote:
> > On 07/01/2014 11:15 AM, Nikolay Gorshkov wrote:
> >>  From another side, we did extensive testing of the build of
> >> {OpenJDK 6 + 6650759 backport} using all JDK 6 tests available
> >> to us, both general and javac specific. We didn't find any
> >> single failure caused by the backport. So, in practice we
> >> don't see any indication of regressions or incompatibilities
> >> caused by the backport, and such, honestly, we see no reason
> >> for backing out the backport.
> >>
> >> Please, let us know if you are aware of a test suite that
> >> shows regressions or any kind of problems introduced by
> >> this changeset.
> > 
> > I'm not at all sure that this is a sound argument.  If there was
> > a test that showed regressions, you surely would have been told
> > about it already.
> > 
> > I don't know that this patch is bad; I don't know that is is good.
> > I do know that it was rejected.
> > 
> > On the other hand, it was accepted for IcedTea.  I don't know why
> > it was accepted; maybe that was before the patch was rejected
> > upstream.
> > 
> > Andrew Hughes, do you know?
> > 
> > Andrew.
> > 
> 
> 

-- 
Andrew :)

Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

PGP Key: 248BDC07 (https://keys.indymedia.org/)
Fingerprint = EC5A 1F5E C0AD 1D15 8F1F  8F91 3B96 A578 248B DC07