Comments
jforbes
It was not, read the ticket about that. There has been active discussion for quite some time. https://github.com/coreos/fedora-coreos-tracker/issues/2087 And based on the feedback in that ticket, I am about to build 6.18.4 without a fix as well. I expect 6.18.5 next week will have the fix for that issue.
The rawhide kernel has the fix, and the fedora-6.17 branch of kernel-ark has the fix. Now that 6.17.7 has been pushed to stable, I expect most of the users who immediately rebuild a new kernel have done so. We should be good to push this update from the kernel side.
The rawhide kernel has the fix, and the fedora-6.17 branch of kernel-ark has the fix. Now that 6.17.7 has been pushed to stable, I expect most of the users who immediately rebuild a new kernel have done so. We should be good to push this update from the kernel side.
The rawhide kernel has the fix, and the fedora-6.17 branch of kernel-ark has the fix. Now that 6.17.7 has been pushed to stable, I expect most of the users who immediately rebuild a new kernel have done so. We should be good to push this update from the kernel side.
This breaks kernel builds
This breaks kernel builds.
@sunwire this has nothing to do with the kernel updates. It seems that rust was updated on Friday in rawhide, and broke kernel compilation there. It has now been filed as an update in all stable Fedora releases. Comment and karma on those bugs, not the kernel ones.
It seems this version of rust breaks kernel compilation (across all releases, including rawhide)
--> rust/build_error.rs:3:1
|
3 | //! Build-time error.
| ^
|
= help: the `-Zsanitizer` flag modifies the ABI so Rust crates compiled with different values of this flag cannot be used together safely
= note: unset `-Zsanitizer` in this crate is incompatible with `-Zsanitizer=kernel-address` in dependency `core`
= help: set `-Zsanitizer=kernel-address` in this crate or unset `-Zsanitizer` in `core`
= help: if you are sure this will not cause problems, you may use `-Cunsafe-allow-abi-mismatch=sanitizer` to silence this error
error: mixing `-Zsanitizer` will cause an ABI mismatch in crate `build_error`
That is a rawhide userspace problem and has nothing to do with sending this kernel as an update to Fedora 42
Sounds like you tried to install the F42 build for test week on an F43 package? perf-6.17.3-300.fc42.x86_64 has nothing to do with this update
@maketopsite can you try booting without rhgb and quiet to see if you get some idea of the issue?
I might also suggest trying the 6.16.5 update in updates-testing. This update is 2 kernels ago, and there have been many fixes since. And I am not sure that -1 karma on a kernel that was stable and has already been obsoleted makes much difference, though that is certainly the thing to do on a kernel that is in testing when the regression is new.
@nixuser if those 2 are in fact related, it means the bug was backported to 6.15.x stable kernels at some point as well, and is already in a release version of Fedora.
For the initramfs size, I haven't done a compare, but I would likely suggest the more recent linux-firmware. A firmware update doesn't rebuild the initramfs, but when you install a new kernel the initramfs is built with the current firmwares on the system. Lately we have had a nasty trend of firmwares growing massively. You can use lsinitrd on various ones and compare.
@sunnyyang I expect that is some fall out from the datacenter move. It is supposed to update itself.
@sunwire your version of dracut is broken. The 107-1 update was unpushed because of this. Revert to the previous version.
This update has been unpushed.
This update breaks the build of the kernel uki. Tested the same kernel build before and after this update with no other change.
@allmorninglong Thanks for testing that. Looks like it is a different issue (and that is a good thing because it is more likely to get a fix quickly.)
@neonica57 Then that would not be a regression from the current stable kernel of 6.18.4, and the negative karma is incorrect.