Skip to content

Add some upcoming distro RIDs#34088

Merged
bartonjs merged 1 commit intodotnet:masterfrom
omajid:more-rids-march-2020
May 11, 2020
Merged

Add some upcoming distro RIDs#34088
bartonjs merged 1 commit intodotnet:masterfrom
omajid:more-rids-march-2020

Conversation

@omajid
Copy link
Member

@omajid omajid commented Mar 25, 2020

This adds some RIDs that we know are current/upcoming:

  • Fedora 33 is in development, and the version after this is most likely going to be Fedora 34. Adding this RID now, so we can get this RID in all current releases when it comes time to build on Fedora 34

  • RHEL has generally committed to doing a major release every 3 years. With RHEL 8 released in 2019, RHEL 9 is under development. There's even a plan on how to create that from Fedora here: https://fedoraproject.org/wiki/Changes/ELN_Buildroot_and_Compose. Also called out here

  • If there's a RHEL 9, there will most likely be an equivalent CentOS 9. Lets add that too.

  • CentOS 7 has an arm64 (aka aarch64) port. As does CentOS 8.

One question: I noticed that this creates an RID graph with an entry for rhel.7-arm64. This RID doesn't already exist in the graph. Is that a bad thing? Should I split out the CentOS entries? Should I add an RID for rhel.7-arm64, which isn't really "supported", but was a product at one point?

@danmoseley
Copy link
Member

I think (?) this is part of what @wfurt took over from @bartonjs.

@omajid
Copy link
Member Author

omajid commented Apr 2, 2020

Anyone willing to review this?

@danmoseley danmoseley requested a review from bartonjs April 2, 2020 18:26
@danmoseley
Copy link
Member

Maybe @bartonjs if @wfurt isn't available

@bartonjs
Copy link
Member

bartonjs commented Apr 2, 2020

Historically, we've avoided adding things into the RID graph until there's some evidence that the next version is already named (think of it as "everyone was surprised at Windows 10 being 10", but also SUSE's rebrand/reset). I think I'm willing to handwave and say that someone from RedHat pushing for things RedHat has a lot of influence over (if not outright control over) is acceptable.

This currently doesn't build, do to reporting a dependency on a node that doesn't exist. I don't have strong opinions on whether we should add a vacuous entry for rhel7-arm64 or split the definition of CentOS inheritance. @ericstj?

@omajid
Copy link
Member Author

omajid commented Apr 2, 2020

Historically, we've avoided adding things into the RID graph until there's some evidence that the next version is already named (think of it as "everyone was surprised at Windows 10 being 10", but also SUSE's rebrand/reset). I think I'm willing to handwave and say that someone from RedHat pushing for things RedHat has a lot of influence over (if not outright control over) is acceptable.

If it makes you more comfortable, I can remove this part of the patch and just deal with centos.8-arm64 + fedora.34-x64 for now. We can perhaps revisit rhel.9-x64 in a few months/year?

@wfurt
Copy link
Member

wfurt commented Apr 3, 2020

If it makes you more comfortable, I can remove this part of the patch and just deal with centos.8-arm64 + fedora.34-x64 for now. We can perhaps revisit rhel.9-x64 in a few months/year?

With 32 & 33 Fedora still in future, adding 34 seems quite far given @bartonjs guidance.
For RHEL, 9 seems more reasonable. Are there any developer preview builds or branch somebody can checkout and build as 9.x @omajid?

'rhel7-arm64' seems fine since isn't really "supported", but was a product at one point.
It feels like simplicity e.g. forking at RHEL would be be better that tracking support contracts.
I would also defer this one to @ericstj

@ericstj
Copy link
Member

ericstj commented Apr 3, 2020

It feels like simplicity e.g. forking at RHEL would be be better that tracking support contracts.
I would also defer this one to @ericstj

Just follow consistent patterns in this file for rhel and centos. See my comment.

@omajid
Copy link
Member Author

omajid commented Apr 3, 2020

As an additional piece of context: please note that Fedora (32 and later) includes .NET Core SDK 3.1 as a standard distro package. You can install dotnet-sdk-3.1 in Fedora just like you can install clang or gcc or curl. As part of that, the SDK needs to know the RID.

With 32 & 33 Fedora still in future, adding 34 seems quite far given @bartonjs guidance.

In practice, they are not far at all. According to the schedule, 32 is "done" development. 33 is in development. At some point in the future, that will be forked to Fedora 34. When that fork is created, the SDK build in Fedora already needs to know about that RID. Or the SDK in Fedora 34 will stop working.

I can definitely wait a few months, but we need to make this change (in isolation in Fedora, or here in the main repositories) to make this RID available before Fedora 34 starts development and the fedora.34-x64 RID is set up on the system.

For RHEL, 9 seems more reasonable. Are there any developer preview builds or branch somebody can checkout and build as 9.x @omajid?

Nope, there's nothing there that I know of. As I linked to earlier, right now there's only discussions going on in the Fedora community on how to create something that might eventually be called RHEL 9.

FWIW, I think RHEL 9 is much further off than Fedora 34.

This adds some RIDs that we know are current/upcoming:

- Fedora 33 is in development, and the version after this is most likely
  going to be Fedora 34. Adding this RID now, so we can get this RID in
  all current releases when it comes time to build on Fedora 34

- RHEL has generally committed to doing a major release every 3 years.
  With RHEL 8 released in 2019, RHEL 9 is under development. There's
  even a plan on how to create that from Fedora here:
  https://fedoraproject.org/wiki/Changes/ELN_Buildroot_and_Compose

- If there's a RHEL 9, there will most likely be an equivalent CentOS 9.
  Lets add that too.

- CentOS 7 has an aarch64 port. As does CentOS 8.
@omajid omajid force-pushed the more-rids-march-2020 branch from 85476ae to 01fb572 Compare May 8, 2020 16:01
@omajid
Copy link
Member Author

omajid commented May 8, 2020

@ericstj Thanks, fixed.

Copy link
Member

@ericstj ericstj left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This looks good to me. @wfurt?

Copy link
Member

@wfurt wfurt left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM.

@bartonjs bartonjs merged commit 19a0db4 into dotnet:master May 11, 2020
omajid added a commit to omajid/dotnet-corefx that referenced this pull request Aug 20, 2020
Fedora 34 is currently in development. Building .NET Core 3.1 there (via
source-build) fails because the runtime id `fedora.34` is unknown:

    error NETSDK1083: The specified RuntimeIdentifier 'fedora.34-x64' is not recognized.

This is a partial backport of
dotnet/runtime#34088
Anipik pushed a commit to dotnet/corefx that referenced this pull request Sep 10, 2020
Fedora 34 is currently in development. Building .NET Core 3.1 there (via
source-build) fails because the runtime id `fedora.34` is unknown:

    error NETSDK1083: The specified RuntimeIdentifier 'fedora.34-x64' is not recognized.

This is a partial backport of
dotnet/runtime#34088
@ghost ghost locked as resolved and limited conversation to collaborators Dec 10, 2020
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

6 participants