-
-
Notifications
You must be signed in to change notification settings - Fork 15.5k
linuxPackages: 4.19 -> 5.4 #78713
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
linuxPackages: 4.19 -> 5.4 #78713
Conversation
It's a longterm version that has been out for quite some time: 5.4.15 and 5.5 are current. I've been using it, so far I'm not aware of any issues with it. Feature freeze for the next NixOS release is in two weeks, so now seems to be high time to decide the default kernel version. https://discourse.nixos.org/t/nixos-20-03-feature-freeze/5655
I suppose we should be good even if we stick with 4.19 for 20.03. Details: EOL for 4.19 on the web is only Dec 2020, but it seems likely to get extended (I think I have seen that somewhere) and that 20.03 is planned to EOL before that anyway. |
This pull request has been mentioned on NixOS Discourse. There might be relevant details there: https://discourse.nixos.org/t/nixos-20-03-feature-freeze/5655/6 |
I haven't had issues with the 5.4 series so far either, so I'd be happy if we could pick that, especially since it provides support for Intels AX200 wifi cards, present in most newer laptops. |
I've been using 5.4 as well, and haven't experienced any issues. |
I'll add that I've had no problems with the 5.4 kernel on my laptop. In fact I've been using |
I'm running 5.4.x in my homelab (w/ ZFS) since the end of December without any issues. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
My AMD-based laptop only works with 5.0+ kernels and 5.4 finally fixed some lockups that could otherwise only mitigated with some weird kernel parameters.
Also have been running latest on different hardware and environment for quite some time. 👍
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I was going to do the same, ensure we stick with the "current latest LTS at the moment of release" we've been doing.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Just wanted to point out 5.4 behaves weird when using VRFs (just run the test from #78476 with linuxPackages_5_4
). Apart from that, the kernel seems perfectly fine, using it on several setups, so overall 👍
If this is the same thing that @WilliButz showed me in Munich then you should probably file an upstream kernel issue about it. Besides that I am also in favor of upgrading to this kernel as it has been working flawlessly on my systems for a while now. |
Yeah, I remeber that debugging session and I'm fairly certain that this is related. Let's see qhen I have sufficient time for investigation :) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I also think we should do this 👍
Ahh I just remembered, there is this bug and I don't think fixes have been backported for it |
5.4 seems to have random freezes with an Intel (i915) driver. My system seems to be affected https://bugzilla.redhat.com/show_bug.cgi?id=1780800 |
Yeah, my aforementioned issue I believe is just one of the few mentioned. |
(Something I wanted to write way before people chimed in about issues.) Anyone's pet issue with the latest kernel is irrelevant to the question of changing the default, I believe. The previous LTS, 4.19, is still available. There will always be issues for a subset of users with different kernel versions. Unless there were egregious regressions or bugs in the newer kernel, waiting until 20.09 is only punting the issue 6 months. There will not be a new LTS to move to in the next 6 months. 4.19 is slated for a December 2020 EOL. Furthermore, sticking with an older release really hampers adoption on new devices, as new hardware is released, especially on laptops, it often becomes the case we need to point users towards the "new kernel" variant of the ISO. Though, considering that intel iGPU issue, which I'm pretty sure I faced with 5.2 or 5.3 while test running it, it would be good to list this as an "errata" for the release, or something. It is bound to become a new common grievance. |
I have issues with |
I agree fully and ref #69687, punting to another release is a practice that will achieve nothing. I'm pretty sure this particular issue is affecting a lot of people so it's really inevitable people to be aware of it, it just shouldn't be a discussion we're having to why we shouldn't upgrade the default. But as you mentioned, something like an errata or anything is needed. |
Another guy with iGPU issues with 5.4. Managed to regain a suitable system by disabling Doesn't look like it'll be fixed in 5.4 currently though. Here's my sources, |
This pull request has been mentioned on NixOS Discourse. There might be relevant details there: https://discourse.nixos.org/t/go-no-go-meeting-nixos-20-03-markhor/6495/19 |
@worldofpeace should we mention the i915 issues in the release notes? |
Motivation for this change
It's a longterm version that has been out for quite some time:
5.4.15 and 5.5 are current. I've been using it, so far I'm not aware
of any issues with it.
Feature freeze for the next NixOS release is in two weeks,
so now seems to be high time to decide the default kernel version.
https://discourse.nixos.org/t/nixos-20-03-feature-freeze/5655
Things done
sandbox
innix.conf
on non-NixOS linux)nix-shell -p nixpkgs-review --run "nixpkgs-review wip"
Tested execution of all binary files (usually in./result/bin/
)nix path-info -S
before and after)