hg: build-infra/jdk9/hotspot: Final fixes for static builds.

[Magnus Ihse Bursie]

On 2015-12-15 16:29, Bob Vandette wrote:
> Are you doing static Mac builds to ensure that your changes are complete?
>
> Gary thought that there were some regressions in jdk9/dev for static builds caused
> by the verona integration.

Don't take the "final" in my comment too literary. :-)

We're not regularly testing static builds, but I have noted on my todo 
list that we should

Actually, come to think of it, a static build definition really should 
be in the jib profile list. Ping Erik :-)

For this occasion, I've been running the compare script manually for 
static build, comparing the libjvm.a from the old build system with the 
one from the new. There are minor differences, but we see these on all 
libraries and all platforms. When we're getting closer to a final 
integration, we'll have to come back to these minor differences and 
determine if we should try to hunt them down, or live with them. Most 
likely they are just artifacts of e.g. different order of the options on 
the command lines etc.

The only "major" difference is the libjvm.symbols, as I wrote elsewhere.

/Magnus

>
> Bob.
>
>
>> On Dec 15, 2015, at 8:10 AM, magnus.ihse.bursie at oracle.com wrote:
>>
>> Changeset: 6c1e87d92d47
>> Author:    ihse
>> Date:      2015-12-15 14:06 +0100
>> URL:       http://hg.openjdk.java.net/build-infra/jdk9/hotspot/rev/6c1e87d92d47
>>
>> Final fixes for static builds.
>>
>> ! makefiles/Common.gmk
>> ! makefiles/JvmPlatforms.gmk
>>