[aarch64-port-dev ] Bring AArch64 into jdk8u mainline, pros/cons?

[Patrick Zhang OS]

Hi Andrew,
Thanks for the replies.
Could I undertand this as: there was a partially done patch for merging aarch64 into jdk8u main, however we don't have a fixed timeline yet due to priroites? or do we have a rough schdule for it now?
About the jdk8u open source projects, I meant not only those in openjdk.java.net but also at external repos, e.g. github.com  

Regards
Patrick

-----Original Message-----
From: Andrew Haley <aph at redhat.com> 
Sent: Tuesday, March 26, 2019 5:21 PM
To: Patrick Zhang OS <patrick at os.amperecomputing.com>; aarch64-port-dev at openjdk.java.net
Subject: Re: Bring AArch64 into jdk8u mainline, pros/cons?

Hi,

On 3/26/19 5:12 AM, Patrick Zhang OS wrote:

> I am not sure whether there had been similar discussion before. It 
> seems to me bringing aarch64 into jdk8u mainline can be a positive.
> 
> The sustaining efficiency of patches/features among jdk/jdk, jdk8u 
> main, and aarch64-port/jdk8u-shenandoah can be improved, whilst the 
> process could be well aligned with those of other jdk updates, i.e., 
> 11u, 12u, etc. While we would have to take care of patches (ported to 
> aarch64-port/jdk8u-shenandoah) more strictly, allow fewer 
> customizations, than before, to ensure a low rate of potential 
> side-effects to main and other ports.

Sure, it'd save some work, especially at CPU time, but it's not a huge amount of work. It's on the list of things to consider, but not at the top of the list.

> Recently I personally reviewed and tested some jdk8u open source 
> projects, derived from OpenJDK 8u mainline, it is a pity that 
> initially aarch64 was not in (someone did the porting later on).

"Someone" was initially Andrew Dinn and me. The patch wasn't ready at JDK 8 time. It's not a pity, it just is.

--
Andrew Haley
Java Platform Lead Engineer
Red Hat UK Ltd. <https://www.redhat.com>
EAC8 43EB D3EF DB98 CC77 2FAD A5CD 6035 332F A671