Bug 1706177

Summary:	[REGRESSION] grub2-mkconfig no longer sees custom built kernels in /boot
Product:	[Fedora] Fedora	Reporter:	Artem S. Tashkinov <aros>
Component:	grub2	Assignee:	Peter Jones <pjones>
Status:	CLOSED DUPLICATE	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	urgent	Docs Contact:
Priority:	unspecified
Version:	30	CC:	fmartine, lkundrak, pjones
Target Milestone:	---
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2019-05-03 19:14:28 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Artem S. Tashkinov 2019-05-03 17:50:34 UTC

Description of problem:

I have a custom kernel _without_ initrd installed under /boot/vmlinuz-5.0-ic64.x86_64

Version-Release number of selected component (if applicable): grub2-tools-2.02-78.fc30.x86_64

How reproducible: always

Steps to Reproduce: grub2-mkconfig

Actual results: my kernel is omitted from the output

Expected results: the kernel being included

Additional info: Fedora 29 did NOT have this bug, i.e. everything worked fine.

Comment 1 Artem S. Tashkinov 2019-05-03 17:56:29 UTC

# grub2-mkconfig > /dev/null
Generating grub configuration file ...
device-mapper: reload ioctl on osprober-linux-sdb5  failed: Device or resource busy
Command failed.
done

I had the same error message with grub2-mkconfig under Fedora 29 but it did not affect its behavior.

My partition layout:

sdb1: EFI System Partition: /boot/efi
sdb5: / (root)
sdb6: /boot

Comment 2 Artem S. Tashkinov 2019-05-03 19:09:37 UTC

strace -fF grub2-mkconfig shows that something stat()'s my customs kernels but they don't show up:

11164 stat("/boot/vmlinuz-4.20-ic64.x86_64", {st_mode=S_IFREG|0644, st_size=4133536, ...}) = 0
11164 rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
11164 rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
11164 stat("/boot/vmlinuz-5.0-ic64.x86_64", {st_mode=S_IFREG|0644, st_size=4133536, ...}) = 0
11164 rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
11164 rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
11164 stat("/vmlinuz-*", 0x7ffd55d2e4a0) = -1 ENOENT (No such file or directory)
11164 rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
11164 rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
11164 stat("/boot/kernel-*", 0x7ffd55d2e4a0) = -1 ENOENT (No such file or directory)
11164 rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
11164 rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
11164 openat(AT_FDCWD, "/boot/ostree/", O_RDONLY|O_NONBLOCK|O_CLOEXEC|O_DIRECTORY) = -1 ENOENT (No such file or directory)
11164 stat("/boot/ostree/*/vmlinuz-*", 0x7ffd55d2e610) = -1 ENOENT (No such file or directory)

I tried creating dummy initrd-5.0-ic64.x86_64 and config-5.0-ic64.x86_64 files but that didn't help either.

Comment 3 Artem S. Tashkinov 2019-05-03 19:14:28 UTC

This comment saved my poor ***:

https://bugzilla.redhat.com/show_bug.cgi?id=1698227#c2

You guys seriously f***ed up.

Why on Earth does a F29 to F30 upgrade changes /etc/default/grub to a non-working state?

*** This bug has been marked as a duplicate of bug 1652806 ***