Update the Fedora 37 GNU Toolchain to glibc 2.36, binutils 2.38.
The set of core GNU Toolchain packages for Fedora 37 are as follows:
The glibc 2.36 change will be tracked in this top-level GNU Toolchain system-wide update.
The binutils 2.38 change will be tracked in this top-level GNU Toolchain system-wide update.
Owners, do not implement this work until the FESCo vote has explicitly ended. The Fedora Program Manager will create a tracking bug in Bugzilla for this Change, which is your indication to proceed. See the FESCo ticket policy and the Changes policy for more information.
+1
I'm confused why we're not going to binutils 2.39, which releases soon, but sure... +1
Switching to binutils 2.39 just before the Fedora release entails too much risk. Using binutils 2.38 which was introduced into Rawhide just after F36 branched is the current operating procedure for the toolchain team.
There are ongoing conversations within the Fedora Toolchain team to determine how we might prepare for such transitions in the future.
Frederic Berat recently commented on a tool that he's developing to help us make such upgrades possible in a timely fashion: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/message/7LAJUKG4PKX7FKLTCKPSGAQQB2H6P7CN/
The short story is that we would need to start testing binutils ahead of the release by mass prebuilding parts of Fedora and tracking failure rates and other issues. This would give us the required risk reduction to switch to a newer binutils. Keep in mind that mistakes in the assembler or static linker are part of the binary artifact and to clean them up requires tracking affected binaries and rebuilding them. This is much different from just rebuilding a shared library or other modular binary artifact.
I know that the 6 month delay in a released binutils can be problematic (specifically for RISC-V), but it is where we are today with testing and tooling.
I hope that answers your questions.
It does, but I'd definitely like to see more proactivity here to improve this situation. The freshness of the GNU Toolchain is extremely important to me and the broader community, so I would really like to see that prioritized and supported. Lagging behind as we do makes it hard for us to do our best as we build out the future of the Linux ecosystem.
Thanks for the feedback. We definitely want to get new features to our users quickly. Speeding this up comes with a unique set of challenges. All I can say is that we are working on it :-) I have a conflict with the Fedora Hatch in Rochester, otherwise I'd attend, I'm only 3.5h from Datto's offices. Perhaps we can arrange another Fedora meeting in the future and I'd be happy to present on the GNU Toolchain, and the Fedora work we do. We're off topic for this fesco issue, but I wanted to answer your question as a fesco council member commenting on this proposal.
It's been a week and I see +6, so this is approved.
APPROVED (+6, 0, 0)
Metadata Update from @kevin: - Issue tagged with: pending announcement
Announced.
Metadata Update from @kevin: - Issue close_status updated to: Accepted - Issue status updated to: Closed (was: Open)
Log in to comment on this ticket.