4:58 a.m.
From: David Ward on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042
NOTE: Truncated patchset due to missing @redhat.com email
address on your GitLab profile at https://gitlab.com/-/profile.
Once that is fixed, close and reopen the merge request to
retrigger sending the emails.
These changes remove unneeded handling for debug kernels, and adjust the
GRUB classes that are configured.
Signed-off-by: David Ward <david.ward(a)ll.mit.edu>
5:02 a.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: David Ward on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5613...
Yes I agree with both points, thank you! I updated the commit.
8:50 a.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Don Zickus on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5616...
@prarit - who are some good kernel/grub developers to review this?
9:19 a.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Prarit Bhargava on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5616...
@dzickusrh I've added @lszubowi.
9:41 a.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Javier Martinez Canillas on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5616...
Thanks @dpward for the merge-request. But the BLS snippet shipped by the
kernel package turned out being not flexible enough and since Fedora 33
the BLS are generated by `kernel-install` using its `grub2` plugin
(`/usr/lib/kernel/install.d/20-grub.install`).
So please propose the changes to `20-grub.install` in the
[Fedora](https://src.fedoraproject.org/rpms/grub2) and [CentOS
Stream](https://gitlab.com/redhat/centos-stream/rpms/grub2) dist-git.
Probably the `generate_bls_conf.sh` script should just be removed. I
believe that wasn't done in case something else would be using it, but I
don't think that's the case.
3:34 p.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: David Ward on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5663...
@martinezjavier Thanks for the info! I changed this merge request so
that it removes the BLS snippet instead. Does this work?
7:11 p.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: David Ward on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5692...
(The testing pipeline failure was unrelated to the changes in this merge
request; it succeeded after a rebase.)
12:34 p.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Don Zickus on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5701...
@martinezjavier How should this work? If we merge @dpward changes
without updating Fedora, things should still work correctly, right?
Currently CKI only tests by building a srpm, which obviously doesn't
test this change. I wouldn't mind through a 'include in release' label
on this to run it through the whole cki pipeline to ensure a new kernel
can still install and boot correctly. Unfortunately upstream still has
a compile problem, so I feel a little stuck.
@jmflinuxtx do we have any patch we can throw on here with 'include in
release' for now to get by our compile problems? We can drop the MR
later with a proper fix.
1:07 p.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Justin Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5701...
@dzickusrh Not that I am aware of, the patches I have seen thrown around
upstream are to pahole, not the kernel. This is still being actively
discussed, with the last message in the thread 10 hours ago.
1:10 p.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Justin Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5701...
I added an include in releases to this patch. Note, while the build
fails, it is only aarch64 that is failing. Koji is still producing build
and rpms for all other arch, you just have to install them from koji as
they will not end up in a compose until aarch64 is fixed.
1:11 p.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Javier Martinez Canillas on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5701...
@dzickusrh yes, it should be safe to merge that for Fedora since as
mentioned in my previous comment the BLS snippets generated on build
time are not used anymore since F33.
1:18 p.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Don Zickus on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5701...
@jmflinuxtx Hmm, ok. cki is still seeing s390x failures too for the
same problem, but meh. CKI also reads FedMessageBus events from Koji.
But if that build is failing then it won't do anything. I guess I can
manually install to verify..
Thanks!
1:30 p.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Justin Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5701...
From the upstream thread, it seems that building for different targets
can cause problems, for instance, building ppc64 fails, but building
ppc64le on a ppc64 host works. Cross compiling can either bypass or
introduce the issue. S390x native works in koji. I am assuming the
pahole fix will address all of these corner cases.
7:48 a.m.
New subject: [OS-BUILD PATCHv2 0/0] Improve generation of boot loader
specification
files
From: Justin Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1042#note_5736...
We have been building with this for a couple of days now, and I haven't
seen any issues on test systems.
1080
days inactive
1095
days old
kernel@lists.fedoraproject.org
13 comments
5 participants
participants (5)
-
David Ward (via Email Bridge) -
Don Zickus (via Email Bridge)
-
Javier Martinez Canillas (via Email Bridge)
-
Justin Forbes (via Email Bridge)
-
Prarit Bhargava (via Email Bridge)