[aarch64-port-dev ] hg: aarch64-port/jdk8u/hotspot: 8157306: Random infrequent null pointer exceptions in javac

[Andrew Hughes]

----- Original Message -----
> On 10/08/16 17:02, Andrew Hughes wrote:
> > We are. So far I've stuck to just importing the current upstream release
> > from 8u (currently 8u102 waiting for review [0]). We could widen that to
> > cherry-pick certain fixes from 8u that we believe are critical enough to
> > not wait until they are included in the upstream release.
> 
> The problem is the way that Hg is working for us.  In an ideal world
> we'd import all this stuff immediately to the trunk and keep a
> separate non-divergent release branch.

I don't think that would solve the problem in something like the Zero example,
because it's the release branch that has to be patched at some point - whether
in the aarch64/jdk8u tree or a distro package - in order to be buildable on
certain architectures.

In other words, I don't see the need for the additional work of following 8u
trunk. After each CPU (latest: u101), we have the best part of three months
to test the latest PSU (latest: u102) before the next CPU (next: u111) is
built on top of it. [0]

> 
> > The Red Hat packages currently include the following fixes from future
> > releases:
> 
> I'd rather not import anything than to diverge as much as that.  It's too
> much to cope with.  I think I can justify the critical AArch64 crasher fix
> in the aarch64-port tree; but the decision not to #ifdef AARCH64 was fairly
> marginal.

Ok, then we're on the same page.

> 
> Andrew.
> 

[0] http://www.oracle.com/technetwork/java/javase/cpu-psu-explained-2331472.html
-- 
Andrew :)

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

PGP Key: ed25519/35964222 (hkp://keys.gnupg.net)
Fingerprint = 5132 579D D154 0ED2 3E04  C5A0 CFDA 0F9B 3596 4222