RFR (M): 8024265: Enable new build on AIX (top level part)

[Erik Joelsson]

Magnus left for the day, but I'm ok with you pushing this to the stage 
area.

/Erik

On 2013-09-12 14:29, Volker Simonis wrote:
> Hi Magnus,
>
> thanks for doing "JDK-8024665 Move open changes for JDK-8020411 to
> closed source"!
>
> Can you now please give Vladimir the GO signal (from a
> build-perspective) to integrate my changes (from
> http://cr.openjdk.java.net/~simonis/webrevs/8024265.v4/) into the
> ppc-aix-port/stage repository?
>
> Actually my changes revert 8020411 as well. This means that when
> 8024665 will flow into our staging repository, we would have to
> manually resolve platform.m4 to our version which we checked in, but
> that should be OK.
>
> I'd really appreciate if we could push my change now, otherwise we
> would have to wait another couple of weeks until 8024665 goes from
> build -> main -> stage.
>
> Thank you and best regards,
> Volker
>
>
> On Thu, Sep 12, 2013 at 10:11 AM, Magnus Ihse Bursie
> <magnus.ihse.bursie at oracle.com> wrote:
>> On 2013-09-11 18:45, Volker Simonis wrote:
>>> Argh! It conflicts with 8020411
>>> (http://cr.openjdk.java.net/~erikj/8020411/webrev.root.01/) from your
>>> last jdk8->stage synchronisation.
>>>
>>> @Magnus, Erik : it seems that '8020411' needed a similar 'feature' to
>>> my actual change but did it without abstracting over the name of the
>>> "compilers target bits" flag name vs. its actual value. Unfortunately
>>> all the users of the change 8020411 are in the closed sources. I'd
>>> really like to stay with my solution (because I think that's the most
>>> general one) and resolve the merge conflicts with 8020411 by
>>> eliminating TARGET_BITS_FLAG which can now be replaced by
>>> "${COMPILER_TARGET_BITS_FLAG}${OPENJDK_TARGET_CPU_BITS}.
>>
>> I see. When looking more closely in the fix for JDK-8020411, it turned out
>> that it could do with some improvements:
>> a) since it really only applies to the closed sources, it could (and should)
>> be done only in the closed sources.
>> b) it was actually incorrect, since it removed the ADDED_*FLAGS variables,
>> which are needed for later use to check for incorrect additions to *FLAGS.
>>
>> I am sorry your patch has become caught in this messiness. :( Nevertheless,
>> I think the best way forward is that I create a new patch, that  reverts the
>> JDK-8020411 in the open sources and re-implements them in the closed
>> sources. I agree that your solution is better, and what we should have in
>> the open sources. You shouldn't have to care about the TARGET_BITS_FLAG, it
>> is an internal hack.
>>
>> I'll start working on it immediately.
>>
>> /Magnus