Porting Fedora for the LoongArch architecture.

[孙海勇]

Hi everyone,

I am Sun Haiyong, from China. I want to port Fedora for the LoongArch 
architecture.
LoongArch is a RISC ISA released by Loongson Technology Corporation Limited,
and has supported a series of (Binutils, GCC, Linux, Glibc, LLVM, QEMU, 
etc.)
core open source projects.

Currently, there are many linux distributions that can run on LoongArch 
machines,
they are OpenEuler, OpenAnolis, UOS, Kylin.

I am good at cross-compiling operating systems and often build Linux systems
using something like LFS or CLFS.

I have built Linux distributions using rpm package management from scratch
several times since 2015 (some systems are not publicly available):

1 Fedora 21, 28, 32 based on MIPS64EL architecture;
2 CentOS 7 based on MIPS64EL architecture;
3 CentOS 7 based on Power8 architecture;
4 CentOS 8.3 based on LoongArch architecture;
5 OpenEuler 2109 based on LoongArch architecture.

And I have published a book on porting Fedora systems to new architectures.

I want to add LoongArch to the official Fedora support architecture, and 
I've
been doing so for some time, here's some of what I've done so far：

To verify the feasibility of building a LoongArch architecture branch for
Fedora, I have used the software version from the rawhide git repository,
and have now compiled a large number of base packages and built a temporary
repository that can be accessed at https://mirrors.wsyu.edu.cn/fedora/

I have compiled and generated more than 45,000 installable rpm files (of
course there are a lot of perl, Python, rust and texlive files), and the
number is still expanding, the scope of the package is enough to build a
LiveCD system, for which I have built LXDE, MATE, WorkStation ( Gnome3) of
the LiveCD and the installation of the ISO, you can get in the following
address: https://github.com/fedora-remix-loongarch/releases-info

Of course, there are still a lot of problems with LoongArch's Fedora system,
for example, some software is not yet fully supported by the upstream
community, but I believe the power of the community can gradually improve
them, so I am sending out an email here to get more people to support this
new LoongArch architecture.

I have recruited some developers who are interested in this and they are：

Wu Xiaotian
Chen Huacai
Shi Pujin
Si Yanteng
Chen Feiyang

Of course, there are many other users who are interested in Fedora systems.

I'm currently a newbie in the Fedora community, so I need help from 
community
developers, and would like someone to guide me on what to do next, such as
what would be a better time to submit necessary patches to packages in the
Fedora repository, how to develop in a collaborative manner, what other
systems to be used for management, etc. In short, any information would be
useful. Could I get help here? :)

Again, thanks for reading this email.

Best Regards
Sun Haiyong

[Neal Gompa]

On Tue, Jan 10, 2023 at 10:13 AM 孙海勇  wrote:
> Hi everyone,
>
> I am Sun Haiyong, from China. I want to port Fedora for the LoongArch
> architecture.
> LoongArch is a RISC ISA released by Loongson Technology Corporation Limited,
> and has supported a series of (Binutils, GCC, Linux, Glibc, LLVM, QEMU,
> etc.)
> core open source projects.
>
> Currently, there are many linux distributions that can run on LoongArch
> machines,
> they are OpenEuler, OpenAnolis, UOS, Kylin.
>
> I am good at cross-compiling operating systems and often build Linux systems
> using something like LFS or CLFS.
>
> I have built Linux distributions using rpm package management from scratch
> several times since 2015 (some systems are not publicly available):
>
> 1 Fedora 21, 28, 32 based on MIPS64EL architecture;
> 2 CentOS 7 based on MIPS64EL architecture;
> 3 CentOS 7 based on Power8 architecture;
> 4 CentOS 8.3 based on LoongArch architecture;
> 5 OpenEuler 2109 based on LoongArch architecture.
>
> And I have published a book on porting Fedora systems to new architectures.
>
> I want to add LoongArch to the official Fedora support architecture, and
> I've
> been doing so for some time, here's some of what I've done so far：
>
> To verify the feasibility of building a LoongArch architecture branch for
> Fedora, I have used the software version from the rawhide git repository,
> and have now compiled a large number of base packages and built a temporary
> repository that can be accessed at https://mirrors.wsyu.edu.cn/fedora/
>
> I have compiled and generated more than 45,000 installable rpm files (of
> course there are a lot of perl, Python, rust and texlive files), and the
> number is still expanding, the scope of the package is enough to build a
> LiveCD system, for which I have built LXDE, MATE, WorkStation ( Gnome3) of
> the LiveCD and the installation of the ISO, you can get in the following
> address: https://github.com/fedora-remix-loongarch/releases-info
>
> Of course, there are still a lot of problems with LoongArch's Fedora system,
> for example, some software is not yet fully supported by the upstream
> community, but I believe the power of the community can gradually improve
> them, so I am sending out an email here to get more people to support this
> new LoongArch architecture.
>
> I have recruited some developers who are interested in this and they are：
>
> Wu Xiaotian
> Chen Huacai
> Shi Pujin
> Si Yanteng
> Chen Feiyang
>
> Of course, there are many other users who are interested in Fedora systems.
>
> I'm currently a newbie in the Fedora community, so I need help from
> community
> developers, and would like someone to guide me on what to do next, such as
> what would be a better time to submit necessary patches to packages in the
> Fedora repository, how to develop in a collaborative manner, what other
> systems to be used for management, etc. In short, any information would be
> useful. Could I get help here? :)
>
> Again, thanks for reading this email.

I think a starting point is to talk to Fedora Infrastructure and
Release Engineering about this.

Also, is there any server-class hardware that can be racked for us to
power builds? We don't typically do emulated builds because they're
incredibly slow.

That said, Fedora RISC-V is probably a good blueprint for how to get
started bootstrapping LoongArch in Fedora. Those folks are here on
this list and they can reply with some details (I don't remember
exactly how they did it).

--
真実はいつも一つ！/ Always, there's only one truth!

[孙海勇]

在 2023/1/10 23:54, Neal Gompa 写道:
> On Tue, Jan 10, 2023 at 10:13 AM 孙海勇  wrote:
> > Hi everyone,
> >
> > I am Sun Haiyong, from China. I want to port Fedora for the LoongArch
> > architecture.
> > LoongArch is a RISC ISA released by Loongson Technology Corporation Limited,
> > and has supported a series of (Binutils, GCC, Linux, Glibc, LLVM, QEMU,
> > etc.)
> > core open source projects.
> >
> > Currently, there are many linux distributions that can run on LoongArch
> > machines,
> > they are OpenEuler, OpenAnolis, UOS, Kylin.
> >
> > I am good at cross-compiling operating systems and often build Linux systems
> > using something like LFS or CLFS.
> >
> > I have built Linux distributions using rpm package management from scratch
> > several times since 2015 (some systems are not publicly available):
> >
> > 1 Fedora 21, 28, 32 based on MIPS64EL architecture;
> > 2 CentOS 7 based on MIPS64EL architecture;
> > 3 CentOS 7 based on Power8 architecture;
> > 4 CentOS 8.3 based on LoongArch architecture;
> > 5 OpenEuler 2109 based on LoongArch architecture.
> >
> > And I have published a book on porting Fedora systems to new architectures.
> >
> > I want to add LoongArch to the official Fedora support architecture, and
> > I've
> > been doing so for some time, here's some of what I've done so far：
> >
> > To verify the feasibility of building a LoongArch architecture branch for
> > Fedora, I have used the software version from the rawhide git repository,
> > and have now compiled a large number of base packages and built a temporary
> > repository that can be accessed at https://mirrors.wsyu.edu.cn/fedora/
> >
> > I have compiled and generated more than 45,000 installable rpm files (of
> > course there are a lot of perl, Python, rust and texlive files), and the
> > number is still expanding, the scope of the package is enough to build a
> > LiveCD system, for which I have built LXDE, MATE, WorkStation ( Gnome3) of
> > the LiveCD and the installation of the ISO, you can get in the following
> > address: https://github.com/fedora-remix-loongarch/releases-info
> >
> > Of course, there are still a lot of problems with LoongArch's Fedora system,
> > for example, some software is not yet fully supported by the upstream
> > community, but I believe the power of the community can gradually improve
> > them, so I am sending out an email here to get more people to support this
> > new LoongArch architecture.
> >
> > I have recruited some developers who are interested in this and they are：
> >
> > Wu Xiaotian
> > Chen Huacai
> > Shi Pujin
> > Si Yanteng
> > Chen Feiyang
> >
> > Of course, there are many other users who are interested in Fedora systems.
> >
> > I'm currently a newbie in the Fedora community, so I need help from
> > community
> > developers, and would like someone to guide me on what to do next, such as
> > what would be a better time to submit necessary patches to packages in the
> > Fedora repository, how to develop in a collaborative manner, what other
> > systems to be used for management, etc. In short, any information would be
> > useful. Could I get help here? :)
> >
> > Again, thanks for reading this email.
>
> I think a starting point is to talk to Fedora Infrastructure and
> Release Engineering about this.
>
> Also, is there any server-class hardware that can be racked for us to
> power builds? We don't typically do emulated builds because they're
> incredibly slow.
>
> That said, Fedora RISC-V is probably a good blueprint for how to get
> started bootstrapping LoongArch in Fedora. Those folks are here on
> this list and they can reply with some details (I don't remember
> exactly how they did it).

Thanks! We have completed the bootstrap steps, and we want to learn more 
about how to further interface with Fedora's build system.

Regards,

Haiyong
> --
> 真実はいつも一つ！/ Always, there's only one truth!
> _______________________________________________
> devel mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
> Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives: https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue

[Matthew Miller]

On Tue, Jan 10, 2023 at 11:12:22PM +0800, 孙海勇 wrote:
> I want to add LoongArch to the official Fedora support architecture,

This is really cool -- welcome, and I'd love to help make sure you succeed!
> I'm currently a newbie in the Fedora community, so I need help from
> community developers, and would like someone to guide me on what to do
> next, such as what would be a better time to submit necessary patches to
> packages in the Fedora repository, how to develop in a collaborative
> manner, what other systems to be used for management, etc. In short, any
> information would be useful. Could I get help here? :)

This message is a good start! There is a little bit of information here
https://fedoraproject.org/wiki/Architectures, but it's not really complete.
I'm afraid a lot of the knowledge is really held in a few people's heads.
Hopefully we can get you connected with the right people!

[Peter Robinson]

Hi Folks,
> On Tue, Jan 10, 2023 at 11:12:22PM +0800, 孙海勇 wrote:
> > I want to add LoongArch to the official Fedora support architecture,
>
> This is really cool -- welcome, and I'd love to help make sure you succeed!
> > I'm currently a newbie in the Fedora community, so I need help from
> > community developers, and would like someone to guide me on what to do
> > next, such as what would be a better time to submit necessary patches to
> > packages in the Fedora repository, how to develop in a collaborative
> > manner, what other systems to be used for management, etc. In short, any
> > information would be useful. Could I get help here? :)
>
> This message is a good start! There is a little bit of information here
> https://fedoraproject.org/wiki/Architectures, but it's not really complete.
> I'm afraid a lot of the knowledge is really held in a few people's heads.
> Hopefully we can get you connected with the right people!

There's generally about a 5 stage process from initial bootstrap of an
arch to mainline Fedora support, it roughly lays out as follows:
1) Low level toolchain bring up
2) A small Linux kernel+userspace
3) Using 2) to build a reduced size/dep standard Fedora rpm userspace
of a random Fedora release
4) Standalone koji instance shadowing the primary Fedora instance
5) Proposal to merge/import the architecture packages into the main
Fedora koji instance

From your previous message [1] I gather you're basically now at 3) but
I'm not sure if you're yet at 4 and doing a shadow build operation of
rawhide.

As part of stages 3 and 4 you'll likely have a bunch of hacks, patches
and changes to both upstream project releases to add support for the
architecture, or possibly as simple as bumping autoconf macros to
detect the architecture that are required to get things to build or
run. This will include things like spec changes that need to get
merged into Fedora package git repos. Long before we get to proposing
5) this needs to be done and the shadow instance of koji should
ultimately be building unchanged packages.

Also before becoming a main Fedora architecture the infrastructure
team will need to be able to source enterprise grade hardware that can
be racked in a datacentre and remotely managed. There's likely some
other requirements the infrastructure team has here.

We have in the past not progressed to stage 5 on a number of
archtectures, and aarch64 was delayed for some time from this, due to
lack of readily available hardware for people to be able to purchase
for testing, development and just general hacking purposes. I have
read some articles about restrictions on getting Loongson chips [2]
but I am not sure if that's all variants or just a subset of the
Loongson architecture. I have in the past actually looked at getting a
Loongson device or two but ultimately gave up because it was far from
straight forward, this may end up being a problem.

Peter

[1] https://lists.fedoraproject.org/archives/list/[e-mail]....
[2] https://www.theregister.com/2022/12/15/china_loongson_chip_export_ban/

[孙海勇]

在 2023/1/11 14:24, Peter Robinson 写道:
> 主题:
> Re: Porting Fedora for the LoongArch architecture.
> From:
> Peter Robinson 
> 日期:
> 2023/1/11 14:24
>
> 收件人:
> Development discussions related to Fedora 
> 抄送:
> "[e-mail]" , 
> "[e-mail]" , 
> "[e-mail]" , Xiaotian Wu 
> , [e-mail], [e-mail]
>
> Hi Folks,
> > On Tue, Jan 10, 2023 at 11:12:22PM +0800, 孙海勇 wrote:
> > > I want to add LoongArch to the official Fedora support architecture,
> >
> > This is really cool -- welcome, and I'd love to help make sure you succeed!
> > > I'm currently a newbie in the Fedora community, so I need help from
> > > community developers, and would like someone to guide me on what to do
> > > next, such as what would be a better time to submit necessary patches to
> > > packages in the Fedora repository, how to develop in a collaborative
> > > manner, what other systems to be used for management, etc. In short, any
> > > information would be useful. Could I get help here? 😄
> >
> > This message is a good start! There is a little bit of information here
> > https://fedoraproject.org/wiki/Architectures, but it's not really complete.
> > I'm afraid a lot of the knowledge is really held in a few people's heads.
> > Hopefully we can get you connected with the right people!
>
> There's generally about a 5 stage process from initial bootstrap of an
> arch to mainline Fedora support, it roughly lays out as follows:
> Low level toolchain bring up
> A small Linux kernel+userspace
> Using 2) to build a reduced size/dep standard Fedora rpm userspace
>
> of a random Fedora release
> 4) Standalone koji instance shadowing the primary Fedora instance
> 5) Proposal to merge/import the architecture packages into the main
> Fedora koji instance
>
> From your previous message [1] I gather you're basically now at 3) but
> I'm not sure if you're yet at 4 and doing a shadow build operation of
> rawhide.

Yes, as you said, we are going to start on step 4.
> As part of stages 3 and 4 you'll likely have a bunch of hacks, patches
> and changes to both upstream project releases to add support for the
> architecture, or possibly as simple as bumping autoconf macros to
> detect the architecture that are required to get things to build or
> run. This will include things like spec changes that need to get
> merged into Fedora package git repos. Long before we get to proposing
> 5) this needs to be done and the shadow instance of koji should
> ultimately be building unchanged packages.
>
> Also before becoming a main Fedora architecture the infrastructure
> team will need to be able to source enterprise grade hardware that can
> be racked in a datacentre and remotely managed. There's likely some
> other requirements the infrastructure team has here.
>
> We have in the past not progressed to stage 5 on a number of
> archtectures, and aarch64 was delayed for some time from this, due to
> lack of readily available hardware for people to be able to purchase
> for testing, development and just general hacking purposes. I have
> read some articles about restrictions on getting Loongson chips [2]
> but I am not sure if that's all variants or just a subset of the
> Loongson architecture. I have in the past actually looked at getting a
> Loongson device or two but ultimately gave up because it was far from
> straight forward, this may end up being a problem.

Thanks！everything is moving in a good direction, so if you have time, 
stay tuned to LoongArch.

Regards,

Haiyong
> Peter
>
> [1]https://lists.fedoraproject.org/archives/list/[e-mail]....
> [2]https://www.theregister.com/2022/12/15/china_loongson_chip_export_ban/
> _______________________________________________
> devel mailing list [e-mail]
> To unsubscribe send an email [e-mail]
> Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives:https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam, report it:https://pagure.io/fedora-infrastructure/new_issue

[Przemek Klosowski]

On 1/11/23 01:24, Peter Robinson wrote:
> Also before becoming a main Fedora architecture the infrastructure
> team will need to be able to source enterprise grade hardware that can
> be racked in a datacentre and remotely managed. There's likely some
> other requirements the infrastructure team has here.

The Registrer claims that China is banning exports of Loongson hardware

https://www.theregister.com/2022/12/15/china_loongson_chip_export_ban/

I have no additional information, just reporting what I saw on the Internet.

[孙海勇]

在 2023/1/11 0:20, Matthew Miller 写道:
> On Tue, Jan 10, 2023 at 11:12:22PM +0800, 孙海勇 wrote:
> > I want to add LoongArch to the official Fedora support architecture,
>
> This is really cool -- welcome, and I'd love to help make sure you succeed!

Thank you very much！
> > I'm currently a newbie in the Fedora community, so I need help from
> > community developers, and would like someone to guide me on what to do
> > next, such as what would be a better time to submit necessary patches to
> > packages in the Fedora repository, how to develop in a collaborative
> > manner, what other systems to be used for management, etc. In short, any
> > information would be useful. Could I get help here? :)
>
> This message is a good start! There is a little bit of information here
> https://fedoraproject.org/wiki/Architectures, but it's not really complete.
> I'm afraid a lot of the knowledge is really held in a few people's heads.
> Hopefully we can get you connected with the right people!

Thanks for the link! Thank you!

Regards,

Haiyong

[Robin Lee]

On Tue, Jan 10, 2023 at 11:13 PM 孙海勇  wrote:
> Hi everyone,
>
> I am Sun Haiyong, from China. I want to port Fedora for the LoongArch
> architecture.
> LoongArch is a RISC ISA released by Loongson Technology Corporation Limited,
> and has supported a series of (Binutils, GCC, Linux, Glibc, LLVM, QEMU,
> etc.)
> core open source projects.
>
> Currently, there are many linux distributions that can run on LoongArch
> machines,
> they are OpenEuler, OpenAnolis, UOS, Kylin.
>
> I am good at cross-compiling operating systems and often build Linux systems
> using something like LFS or CLFS.
>
> I have built Linux distributions using rpm package management from scratch
> several times since 2015 (some systems are not publicly available):
>
> 1 Fedora 21, 28, 32 based on MIPS64EL architecture;
> 2 CentOS 7 based on MIPS64EL architecture;
> 3 CentOS 7 based on Power8 architecture;
> 4 CentOS 8.3 based on LoongArch architecture;
> 5 OpenEuler 2109 based on LoongArch architecture.
>
> And I have published a book on porting Fedora systems to new architectures.
>
> I want to add LoongArch to the official Fedora support architecture, and
> I've
> been doing so for some time, here's some of what I've done so far：
>
> To verify the feasibility of building a LoongArch architecture branch for
> Fedora, I have used the software version from the rawhide git repository,
> and have now compiled a large number of base packages and built a temporary
> repository that can be accessed at https://mirrors.wsyu.edu.cn/fedora/
>
> I have compiled and generated more than 45,000 installable rpm files (of
> course there are a lot of perl, Python, rust and texlive files), and the
> number is still expanding, the scope of the package is enough to build a
> LiveCD system, for which I have built LXDE, MATE, WorkStation ( Gnome3) of
> the LiveCD and the installation of the ISO, you can get in the following
> address: https://github.com/fedora-remix-loongarch/releases-info
>
> Of course, there are still a lot of problems with LoongArch's Fedora system,
> for example, some software is not yet fully supported by the upstream
> community, but I believe the power of the community can gradually improve
> them, so I am sending out an email here to get more people to support this
> new LoongArch architecture.
>
> I have recruited some developers who are interested in this and they are：
>
> Wu Xiaotian
> Chen Huacai
> Shi Pujin
> Si Yanteng
> Chen Feiyang
>
> Of course, there are many other users who are interested in Fedora systems.
>
> I'm currently a newbie in the Fedora community, so I need help from
> community
> developers, and would like someone to guide me on what to do next, such as
> what would be a better time to submit necessary patches to packages in the
> Fedora repository, how to develop in a collaborative manner, what other
> systems to be used for management, etc. In short, any information would be
> useful. Could I get help here? :)
>
> Again, thanks for reading this email.

Hi, 海勇 and other LoongArch contributors!

I am a long-term Chinese Fedora packager. I also got interested in
LoongArch recently
and started to create some PRs[1][2] related to LoongArch.

Please feel free to reach out to me personally. I hope I can help you
from the Fedora side.

[1] https://src.fedoraproject.org/rpms/cross-binutils/pull-request/2
[2] https://src.fedoraproject.org/rpms/cross-gcc/pull-request/3
-robin
> Best Regards
> Sun Haiyong
> _______________________________________________
> devel mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
> Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives: https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue

[孙海勇]

在 2023/1/11 0:50, Robin Lee 写道:
> On Tue, Jan 10, 2023 at 11:13 PM 孙海勇  wrote:
> > Hi everyone,
> >
> > I am Sun Haiyong, from China. I want to port Fedora for the LoongArch
> > architecture.
> > LoongArch is a RISC ISA released by Loongson Technology Corporation Limited,
> > and has supported a series of (Binutils, GCC, Linux, Glibc, LLVM, QEMU,
> > etc.)
> > core open source projects.
> >
> > Currently, there are many linux distributions that can run on LoongArch
> > machines,
> > they are OpenEuler, OpenAnolis, UOS, Kylin.
> >
> > I am good at cross-compiling operating systems and often build Linux systems
> > using something like LFS or CLFS.
> >
> > I have built Linux distributions using rpm package management from scratch
> > several times since 2015 (some systems are not publicly available):
> >
> > 1 Fedora 21, 28, 32 based on MIPS64EL architecture;
> > 2 CentOS 7 based on MIPS64EL architecture;
> > 3 CentOS 7 based on Power8 architecture;
> > 4 CentOS 8.3 based on LoongArch architecture;
> > 5 OpenEuler 2109 based on LoongArch architecture.
> >
> > And I have published a book on porting Fedora systems to new architectures.
> >
> > I want to add LoongArch to the official Fedora support architecture, and
> > I've
> > been doing so for some time, here's some of what I've done so far：
> >
> > To verify the feasibility of building a LoongArch architecture branch for
> > Fedora, I have used the software version from the rawhide git repository,
> > and have now compiled a large number of base packages and built a temporary
> > repository that can be accessed at https://mirrors.wsyu.edu.cn/fedora/
> >
> > I have compiled and generated more than 45,000 installable rpm files (of
> > course there are a lot of perl, Python, rust and texlive files), and the
> > number is still expanding, the scope of the package is enough to build a
> > LiveCD system, for which I have built LXDE, MATE, WorkStation ( Gnome3) of
> > the LiveCD and the installation of the ISO, you can get in the following
> > address: https://github.com/fedora-remix-loongarch/releases-info
> >
> > Of course, there are still a lot of problems with LoongArch's Fedora system,
> > for example, some software is not yet fully supported by the upstream
> > community, but I believe the power of the community can gradually improve
> > them, so I am sending out an email here to get more people to support this
> > new LoongArch architecture.
> >
> > I have recruited some developers who are interested in this and they are：
> >
> > Wu Xiaotian
> > Chen Huacai
> > Shi Pujin
> > Si Yanteng
> > Chen Feiyang
> >
> > Of course, there are many other users who are interested in Fedora systems.
> >
> > I'm currently a newbie in the Fedora community, so I need help from
> > community
> > developers, and would like someone to guide me on what to do next, such as
> > what would be a better time to submit necessary patches to packages in the
> > Fedora repository, how to develop in a collaborative manner, what other
> > systems to be used for management, etc. In short, any information would be
> > useful. Could I get help here? :)
> >
> > Again, thanks for reading this email.
>
> Hi, 海勇 and other LoongArch contributors!
>
> I am a long-term Chinese Fedora packager. I also got interested in
> LoongArch recently
> and started to create some PRs[1][2] related to LoongArch.
>
> Please feel free to reach out to me personally. I hope I can help you
> from the Fedora side.
>
> [1] https://src.fedoraproject.org/rpms/cross-binutils/pull-request/2
> [2] https://src.fedoraproject.org/rpms/cross-gcc/pull-request/3
> -robin

Thanks！if you don't mind, we can get it together.

Regards,

Haiyong
> > Best Regards
> > Sun Haiyong
> > _______________________________________________
> > devel mailing list -- [e-mail]
> > To unsubscribe send an email to [e-mail]
> > Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> > List Archives: https://lists.fedoraproject.org/archives/list/[e-mail]
> > Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
>
> devel mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
> Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives: https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue

[Richard W.M. Jones]

Hi,

Porting Fedora to a new architecture is quite a challenge but a lot of
fun.  It sounds as if you've made a lot of progress already.  A few
thoughts from the Fedora/RISC-V effort ...

(1) Cross-compiling RPMs (eg. from x86-64 to LoongArch) isn't really
useful.  Almost any significantly complex RPM must be compiled on the
same architecture as it targets.  For example it may run tests or
tools that it builds as part of building the RPM.

For Fedora/RISC-V the problem was that we didn't have a viable Linux
environment to even run RPM on (in 2016), so I came up with a pretty
nasty bootstrapping hack, which cross-compiled enough of an
environment to be able to run rpmbuild.  For historical interest
that's here: https://github.com/rwmjones/fedora-riscv-bootstrap

However you won't need to use this.  There already exists a major
distro on LoongAch (ie. Debian).  You could use Debian to build a base
set of RPMs, and then once you have enough, install them into a
buildroot and continue using the Fedora you've built.

(I guess from your email that you've already done something like this.)

(2) noarch RPMs don't need to be compiled!  You can just copy them
from x86-64.  (At least for bootstrapping purposes, you'll want to be
able to compile them eventually.)  This is a nice time-saving tip to
remember.

(3) For RISC-V we didn't start with Koji (used our own hacky build
system), but did eventually set one up:
http://fedora.riscv.rocks/koji/

You'll also want to set up a Koji instance eventually.

(4) As another reply mentioned, to get LoongArch as a primary Fedora
architecture, an essential requirement is 19" rackable server-class
hardware.  It needs to be fully manageable through a BMC.  Fedora will
probably need a few such servers to be donated.

This is the main reason why RISC-V isn't a primary Fedora architecture
yet, although progress is happening.

(5) PLEASE PLEASE PLEASE get your patches upstream!  If they are
patches to the upstream software, send them to the upstream
maintainers.  If they are patches to the RPM builds, add them to
Fedora.  (People on this list can help with both these tasks.)

Getting stuff upstream benefits the whole community beyond Fedora, and
makes everyone's life easier.

Rich.

[Stephen Smoogen]

On Tue, 10 Jan 2023 at 17:37, Richard W.M. Jones  wrote:
> Hi,
>
> ... clipped as I want to focus on this part.
> (4) As another reply mentioned, to get LoongArch as a primary Fedora
> architecture, an essential requirement is 19" rackable server-class
> hardware.  It needs to be fully manageable through a BMC.  Fedora will
> probably need a few such servers to be donated.

This is usually the major hanging point for bringing up it as a 'main' OS.
I am going to outline in depth what it takes from an operational side to
get any new 'deliverable' into Fedora currently. I am not doing so to say
'this can't be done' as much as to explain why it can't usually be done at
the speed most people seem to think can be done.

The complete Fedora koji build system has a lot of assumptions of being in
the same datacenter with just one set of hardware remote and 'breaking'
regularly due to that (mounts are being done over NFS over ssh and
disconnect regularly.. connections between the main DC and other places may
time out etc.). The build system is generally locked together so that if
one architecture has slowness/problems it affects all builds on the other
arches. Adding more architectures to this tends to add a non-linear
complexity in places. [The opposite of having multiple koji's adds a
different level of complexity and requires more dedicated people time which
is in even shorter supply.]

The main DC is in the continental United States and in a 'lights' out
facility (aka there are no hands in the facility regularly). This means
that the hardware needs to be able to be managed remotely and be redundant
so that it can 'work' in a degraded shape for weeks until replacements can
be reached. There is no 'lab' for fixing equipment so the parts generally
need to have a dedicated tech to fix. [Practically this means the hardware
in the facility needs to be rated and insurable for this sort of
data-center. Certain hardware is rated only for 'labs' or places where
someone can unplug quickly if it smokes.. this is not that kind of place.]

The next issue is space to put more equipment into this facility. Pretty
much every build architecture takes at least 1 full sized rack for the
number of systems needed to keep up with builds. That requires additional
planning as the space used is shared between multiple groups and may not be
possible to add in quickly (or at all). Other solutions are possible but
would require additional work and time. There are also the requirements for
additional NFS storage, power, etc etc that each deliverable requires.
Those all need to be added, budgeted for and purchased. [What has happened
several times to slow things down is that it wasn't
ordered/budgeted/purchased/put-in-place.. and well 6-12 months got lost
making it happen.]

That all sounds like a lot of 'no' wrapped up in a 'it depends', but it can
be done. In general it can be a 2-5 year process from when someone says
'Hey wouldn't it be great to build X' to getting the 'factory' built and
working with the rest of the build system. During that time, usually a
secondary external build system may be built elsewhere using shadow-koji or
some other build tools to keep up and work out a lot of the bugs in the
many parts of the build system (bodhi, koji, pungi, odcs, osbs, git,
bugzilla, messaging system, pdc, mbs, oci, signing-infra, qa, koschei, and
various things which are containers that affected/are affected by any
build). Doing that allows for the eventual infusion of the arch into the
main Fedora to take weeks versus months/year.
> This is the main reason why RISC-V isn't a primary Fedora architecture
> yet, although progress is happening.

[孙海勇]

在 2023/1/11 21:15, Stephen Smoogen 写道:
> 主题:
> Re: Porting Fedora for the LoongArch architecture.
> From:
> Stephen Smoogen 
> 日期:
> 2023/1/11 21:15
>
> 收件人:
> Development discussions related to Fedora 
> 抄送:
> "[e-mail]" , 
> "[e-mail]" , 
> "[e-mail]" , Xiaotian Wu 
> , [e-mail], [e-mail]
>
> On Tue, 10 Jan 2023 at 17:37, Richard W.M. Jones  
> wrote:
> Hi,
>
> ... clipped as I want to focus on this part.
> (4) As another reply mentioned, to get LoongArch as a primary Fedora
> architecture, an essential requirement is 19" rackable server-class
> hardware.  It needs to be fully manageable through a BMC. Fedora will
> probably need a few such servers to be donated.
>
>
> This is usually the major hanging point for bringing up it as a 'main' 
> OS. I am going to outline in depth what it takes from an operational 
> side to get any new 'deliverable' into Fedora currently. I am not 
> doing so to say 'this can't be done' as much as to explain why it 
> can't usually be done at the speed most people seem to think can be done.
>
> The complete Fedora koji build system has a lot of assumptions of 
> being in the same datacenter with just one set of hardware remote and 
> 'breaking' regularly due to that (mounts are being done over NFS over 
> ssh and disconnect regularly.. connections between the main DC and 
> other places may time out etc.). The build system is generally locked 
> together so that if one architecture has slowness/problems it affects 
> all builds on the other arches. Adding more architectures to this 
> tends to add a non-linear complexity in places. [The opposite of 
> having multiple koji's adds a different level of complexity and 
> requires more dedicated people time which is in even shorter supply.]
>
> The main DC is in the continental United States and in a 'lights' out 
> facility (aka there are no hands in the facility regularly). This 
> means that the hardware needs to be able to be managed remotely and be 
> redundant so that it can 'work' in a degraded shape for weeks until 
> replacements can be reached. There is no 'lab' for fixing equipment so 
> the parts generally need to have a dedicated tech to fix. [Practically 
> this means the hardware in the facility needs to be rated and 
> insurable for this sort of data-center. Certain hardware is rated only 
> for 'labs' or places where someone can unplug quickly if it smokes.. 
> this is not that kind of place.]
>
> The next issue is space to put more equipment into this facility. 
> Pretty much every build architecture takes at least 1 full sized rack 
> for the number of systems needed to keep up with builds. That requires 
> additional planning as the space used is shared between multiple 
> groups and may not be possible to add in quickly (or at all). Other 
> solutions are possible but would require additional work and time. 
> There are also the requirements for additional NFS storage, power, etc 
> etc that each deliverable requires. Those all need to be added, 
> budgeted for and purchased. [What has happened several times to slow 
> things down is that it wasn't 
> ordered/budgeted/purchased/put-in-place.. and well 6-12 months got 
> lost making it happen.]
>
> That all sounds like a lot of 'no' wrapped up in a 'it depends', but 
> it can be done. In general it can be a 2-5 year process from when 
> someone says 'Hey wouldn't it be great to build X' to getting the 
> 'factory' built and working with the rest of the build system. During 
> that time, usually a secondary external build system may be built 
> elsewhere using shadow-koji or some other build tools to keep up and 
> work out a lot of the bugs in the many parts of the build system 
> (bodhi, koji, pungi, odcs, osbs, git, bugzilla, messaging system, pdc, 
> mbs, oci, signing-infra, qa, koschei, and various things which are 
> containers that affected/are affected by any build). Doing that allows 
> for the eventual infusion of the arch into the main Fedora to take 
> weeks versus months/year.
> This is the main reason why RISC-V isn't a primary Fedora architecture
> yet, although progress is happening.

Thank you so much!

The information you provided is very useful and this information can 
save us a lot of trouble.

I'm going to digest this information slowly next.

Regards,

Haiyong
> -- 
> Stephen Smoogen, Red Hat Automotive
> Let us be kind to one another, for most of us are fighting a hard 
> battle. -- Ian MacClaren
>
> devel mailing list [e-mail]
> To unsubscribe send an email [e-mail]
> Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives:https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam, report it:https://pagure.io/fedora-infrastructure/new_issue

[孙海勇]

Hi

在 2023/1/11 6:37, Richard W.M. Jones 写道:
> Hi,
>
> Porting Fedora to a new architecture is quite a challenge but a lot of
> fun.  It sounds as if you've made a lot of progress already.  A few
> thoughts from the Fedora/RISC-V effort ...
>
> (1) Cross-compiling RPMs (eg. from x86-64 to LoongArch) isn't really
> useful.  Almost any significantly complex RPM must be compiled on the
> same architecture as it targets.  For example it may run tests or
> tools that it builds as part of building the RPM.
>
> For Fedora/RISC-V the problem was that we didn't have a viable Linux
> environment to even run RPM on (in 2016), so I came up with a pretty
> nasty bootstrapping hack, which cross-compiled enough of an
> environment to be able to run rpmbuild.  For historical interest
> that's here: https://github.com/rwmjones/fedora-riscv-bootstrap
>
> However you won't need to use this.  There already exists a major
> distro on LoongAch (ie. Debian).  You could use Debian to build a base
> set of RPMs, and then once you have enough, install them into a
> buildroot and continue using the Fedora you've built.
>
> (I guess from your email that you've already done something like this.)

Yes, I use cross-compilation to build the bootstrap environment, and use 
it to create an rpm group that is sufficient to build the rest of the 
system self-sufficiently, and then use this package group to build the 
whole system.
> (2) noarch RPMs don't need to be compiled!  You can just copy them
> from x86-64.  (At least for bootstrapping purposes, you'll want to be
> able to compile them eventually.)  This is a nice time-saving tip to
> remember.
>
> (3) For RISC-V we didn't start with Koji (used our own hacky build
> system), but did eventually set one up:
> http://fedora.riscv.rocks/koji/
>
> You'll also want to set up a Koji instance eventually.

We're already setting up koji.
> (4) As another reply mentioned, to get LoongArch as a primary Fedora
> architecture, an essential requirement is 19" rackable server-class
> hardware.  It needs to be fully manageable through a BMC.  Fedora will
> probably need a few such servers to be donated.
>
> This is the main reason why RISC-V isn't a primary Fedora architecture
> yet, although progress is happening.

This is a difficult task, but I will try to do it.
> (5) PLEASE PLEASE PLEASE get your patches upstream!  If they are
> patches to the upstream software, send them to the upstream
> maintainers.  If they are patches to the RPM builds, add them to
> Fedora.  (People on this list can help with both these tasks.)

Ok, I will do these as soon as possible.
> Getting stuff upstream benefits the whole community beyond Fedora, and
> makes everyone's life easier.

Thank you very much！

Regards,

Haiyong
> Rich.

[Zbigniew Jędrzejewski-Szmek]

On Tue, Jan 10, 2023 at 11:12:22PM +0800, 孙海勇 wrote:
> Hi everyone,
>
> I am Sun Haiyong, from China. I want to port Fedora for the LoongArch
> architecture.

Cool!

What is the status of qemu support?

Zbyszek

[Robin Lee]

On Wed, Jan 11, 2023 at 4:34 PM Zbigniew Jędrzejewski-Szmek
 wrote:
> On Tue, Jan 10, 2023 at 11:12:22PM +0800, 孙海勇 wrote:
> > Hi everyone,
> >
> > I am Sun Haiyong, from China. I want to port Fedora for the LoongArch
> > architecture.
>
> Cool!
>
> What is the status of qemu support?

QEMU support is fully upstreamed. Qemu[1] in F38/Rawhide already built
loongarch64 sub-packages.
EDK2 support is also upstreamed but Fedora does not catch up. We can
download edk2-loongarch64-code.fd from here[2]

[1] https://koji.fedoraproject.org/koji/buildinfo?buildID=2102503
[2] https://github.com/loongson/Firmware/blob/b5cbb26048ab3000e76df9ebae72a2a3ec...
-robin
> Zbyszek
> _______________________________________________
> devel mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
> Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives: https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue

[Gerd Hoffmann]

On Wed, Jan 11, 2023 at 05:26:33PM +0800, Robin Lee wrote:
> EDK2 support is also upstreamed but Fedora does not catch up.

There is no cross compiler (yet) in Fedora so I can hardly
build edk2 firmware binaries ...

I expect that will change when the gcc-13 update lands in
rawhide.

take care,
  Gerd

[Dan Horák]

On Wed, 11 Jan 2023 11:45:59 +0100
Gerd Hoffmann  wrote:
> On Wed, Jan 11, 2023 at 05:26:33PM +0800, Robin Lee wrote:
> > EDK2 support is also upstreamed but Fedora does not catch up.
>
> There is no cross compiler (yet) in Fedora so I can hardly
> build edk2 firmware binaries ...
>
> I expect that will change when the gcc-13 update lands in
> rawhide.

right now we have a FTBFS issue in cross-binutils after their rebase to
2.39. Once solved we can continue with cross-gcc and the LoongArch
PRs, which are pretty straightforward.

Dan

[Jonathan Schleifer]

Is there still any work being done on this? I'm interested in running 
Fedora on LoongArch and helping with the porting efforts. I do have 
hardware as well (it's available on AliExpress, even outside of China).