Backports to jdk11u and jdk17u
Lindenmaier, Goetz
goetz.lindenmaier at sap.com
Tue Jan 17 15:06:35 UTC 2023
Hi,
As pointed out in the mail you cited below, not breaking
existing code is the major concern of such a backport in
the view of the maintainers.
Shared changes should be kept to a minimum.
Especially in 11, only enabling shared changes are
eligible, not changes that aim to improve the performance.
It would be great to see a complete list of
changes that need to be backported. These should
make up a functioning port. A project repo can serve
the purpose to collect these changes and prove
they are a complete set.
We can then look at this list of changes and give a
basic ok for these to be integrated in 17u/11u. Then
you can run the normal backport process for them.
Best regards,
Goetz.
> -----Original Message-----
> From: jdk-updates-dev <jdk-updates-dev-retn at openjdk.org> On Behalf Of
> yangfei at iscas.ac.cn
> Sent: Tuesday, January 17, 2023 9:23 AM
> To: Ludovic Henry <ludovic at rivosinc.com>
> Cc: 章翔 <zhangxiang at iscas.ac.cn>; riscv-port-dev at openjdk.org;
> yadonn.wang at huawei.com; jdk-updates-dev at openjdk.java.net
> Subject: Re: Re: Re: Backports to jdk11u and jdk17u
>
> [Some people who received this message don't often get email from
> yangfei at iscas.ac.cn. Learn why this is important at
> https://aka.ms/LearnAboutSenderIdentification ]
>
> Hi,
>
> -----Original Messages-----
> From:"Ludovic Henry" <ludovic at rivosinc.com>
> Sent Time:2023-01-16 23:36:33 (Monday)
> To: "章翔" <zhangxiang at iscas.ac.cn>
> Cc: yangfei at iscas.ac.cn, riscv-port-dev at openjdk.org,
> yadonn.wang at huawei.com, jdk-updates-dev at openjdk.java.net
> Subject: Re: Re: Backports to jdk11u and jdk17u
>
> >> However there is one big difference between general jdk17u and
> riscv secific one: while jdk17u accepts only bug fixes and minor perf
> improvements, the riscv-port-jdk17u will have to accept larger patches: as
> risc-v port is far from being perfect so far, performance-wise is worse than
> arm64 port.
> > I expect to initiate the backport in a dedicated openjdk/riscv-port-
> jdk17u repository for exactly that. Until the riscv port is considered very
> stable and performant, we likely want to keep it in a separate repository.
> When stable, we can work with the relevant maintainers of jdk8u, jdk11u
> and jdk17u for possible integrations.
>
> That sounds reasonable to me as we are able to find more time to look at the
> jvm performance issues on RISCV now.
>
> > > In fact, there were some discussions before on the list about the
> backporting work [1].
> >
> > I missed this one, thanks for pointing it out.
> >
> > > I'm Xiang Zhang from the Institute of Software, Chinese Academy
> of Sciences. I'm backporting jdk8u for RISC-V support. Here is my
> development repository:
> >> https://github.com/zhangxiang-plct/jdk8u
> > If there is a riscv-port-jdk8u project, would you be interested to
> contribute there directly?
> >
> > In the meantime, I'll contact skara-dev to figure out what it would take
> to create the riscv-port-jdk11u and riscv-port-jdk17u repositories.
>
> Besides the repositories needed, I also wonder the correct workflow for
> future reviewing & merging of backporting PRs on those repos.
> I would like to hear comments/suggestions from the jdk update maintainers.
>
> Thanks,
> Fei </zhangxiang at iscas.ac.cn></ludovic at rivosinc.com>
More information about the jdk-updates-dev
mailing list