| Git Clone URL: | https://aur.archlinux.org/linux-mainline.git (read-only, click to copy) |
|---|---|
| Package Base: | linux-mainline |
| Description: | Headers and scripts for building modules for the Linux Mainline kernel |
| Upstream URL: | https://kernel.org/ |
| Licenses: | GPL-2.0-only |
| Submitter: | miffe |
| Maintainer: | miffe |
| Last Packager: | miffe |
| Votes: | 278 |
| Popularity: | 2.51 |
| First Submitted: | 2011-07-22 06:50 (UTC) |
| Last Updated: | 2024-11-18 03:24 (UTC) |
Just a forewarning about the upcoming kernel 5.8-rc1, there's a critical problem with ext4 that needs to be patched before running it or you may experience data corruption.
The patch is at: https://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4.git/commit/?h=dev&id=811985365378df01386c3cfb7ff716e74ca376d5
This is already documented in many other places, but take care that you will need to apply this patch if you build this kernel using GCC 10 (which is already in core), or you will get an unbootable kernel:
EDIT: This patch is already included in -rc6 so it is unnecessary to include it.
With the 5.7 release, the (currently present) CONFIG_STAGING_EXFAT_FS is gone, replaced by a wholly separate implementation by samsung in CONFIG_EXFAT_FS. This patch removes the staging exfat selections (they're gone anyway) and selects CONFIG_EXFAT_FS as a module: https://brpaste.xyz/NpG3Lg
@miffe you may add makedepend to my new package python-sphinx-2 to fix build of docs
Same people (including me) can't get intel wifi chipsets to work with linux 5.6. Just a warning, hopefully someone will find a fix soon.
Edit: applying this patch should work: https://patchwork.ozlabs.org/patch/1263553/ (see this thread https://www.reddit.com/r/linux/comments/frfbz5/the_56_kernel_has_been_released/flwti4p/)
Edit 2: added said patch to the PKGBUILD, and wifi now works.
Now that 5.6 is being tested, this would be a good opportunity to include support for wireguard with CONFIG_WIREGUARD=y.
Repository here
As an update, it appears the issue was with the default makepkg.conf flags. I changed -mtune=generic to -march=native (might as well, right?) and everything compiled just fine, after ~10 total failed attempts (over the last few days). No idea why that's the case, but it looks like -mtune=generic might be triggering something untowards; posting this in case it helps someone else.
Weird, that's the same version as me. Maybe check the kernel log if it's been OOM killed. Other than that i've no idea.
Pinned Comments
miffe commented on 2020-01-30 21:44 (UTC)
Repository here