Merge ppc port into jdk7u-dev

[Andrew Hughes]

----- Original Message -----
> Hi Andrew :),
> 
> sorry, I don't understand.
> 
> > $ hg pull -r 9ee6abf28de9 ../../ppc7/hotspot
> This will give you most of the changes we did without openJdk bugid.
> Just the ones Andrew Ha. doesn't want at all as I understand.
> You will have to do some meaningful merges, without being able
> to build, as, e.g., Better Byte Behaviour will be missing.
> 

One merge and the result can be compared against the ppc-aix-port repo
at the same point. We wouldn't push that; we'd merge into an interim
repository then just apply the patch to 7u, which is what you were
proposing to do with the whole thing, I believe.

> If you form the diffs you have after the merge into a new
> changeset and submit that to  jdk7u/jdk7u, you still can't
> pull the remaining changes for the port.  The ones without bugid
> still will come along.

Yes, I know; you can't do that either way if you don't want those
changesets.

> If you do all of them individually on top this really is a major effort.
> You would have
>   1.) to find all changes in ppc-aix-port and not in jdk7u/jdk7u

hg in tells you this.

>   2.) export all of them
>   3.) select a useful order of them
>   4.) edit the patches so they reflect the changes done in merges

IME, this is a 5-10 minute job with most of the work actually being
checking the final result is the same as the original tree.

>   5.) submit them to jdk7u/jdk7u
> 
> If you have all these you can make a patch queue and do a test build,
> and then move the fixes required to the proper patches.
> 
> Hmmm... sounds like quite some work.
> 
> How long will the CPU keep you busy?
> 

The 7 work is done. It's waiting on successful testing before submission.

I'll try and find some time to look at this, and then you can critique
the result rather than what you think I'm saying. However, I don't see
the sudden rush for this. We're not having another release until October.

> And ... probably 6f0b8f13db4d, "New files for template interpreter." should
> be used.

Why so? It seems like something that was missing from the preceding changeset,
but I'll check.

> 
> Best regards,
>   Goetz.
> 
> 
> 


-- 
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