Bug 31581 - microcode new security issues CVE-2022-21216 CVE-2022-33196 CVE-2022-33972 CVE-2022-38090
Summary: microcode new security issues CVE-2022-21216 CVE-2022-33196 CVE-2022-33972 CV...
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: Security (show other bugs)
Version: 8
Hardware: All Linux
Priority: Normal normal
Target Milestone: ---
Assignee: QA Team
QA Contact: Sec team
URL:
Whiteboard: MGA8-64-OK, MGA8-32-OK
Keywords: advisory, validated_update
Depends on:
Blocks:
 
Reported: 2023-02-21 17:04 CET by David Walser
Modified: 2023-03-11 20:02 CET (History)
6 users (show)

See Also:
Source RPM: microcode-0.20220809-1.mga8.nonfree.src.rpm
CVE:
Status comment: Fixed upstream in 20230214


Attachments
last lines of rdsosreport.txt failed booting kernel kernel-desktop-6.1.6-1.mga8-1-1.mga8.x86_64 with LVM on LUKS (58.95 KB, image/jpeg)
2023-03-05 18:23 CET, Morgan Leijström
Details
Dracut output executed for kernel 1.6.1 (5.53 KB, text/plain)
2023-03-05 19:06 CET, Morgan Leijström
Details

David Walser 2023-02-21 17:04:37 CET

Status comment: (none) => Fixed upstream in 20230214

Comment 1 David Walser 2023-02-27 16:26:51 CET
Ubuntu has issued an advisory for this today (February 27):
https://ubuntu.com/security/notices/USN-5886-1

Summary: microcode new security issues CVE-2022-21216 CVE-2022-33196 CVE-2022-38090 => microcode new security issues CVE-2022-21216 CVE-2022-33196 CVE-2022-33972 CVE-2022-38090

Comment 2 Thomas Backlund 2023-03-04 18:30:12 CET
Already built on 2023-02-16, but I forgot to open a bug...


(S)RPM:
microcode-0.20230214-1.mga8.nonfree

Assignee: tmb => qa-bugs

Comment 3 Len Lawrence 2023-03-05 09:41:19 CET
Mageia8, x86_64, linus kernel

Intel Core i9-7900X

Updated and rebooted.
$ dmesg | grep microcode
[    0.000000] microcode: microcode updated early to revision 0x2006e05, date = 2022-03-08
[    0.717801] microcode: sig=0x50654, pf=0x4, revision=0x2006e05
[    0.717887] microcode: Microcode Update Driver: v2.2.
[    4.705677] em28xx 1-7:1.0: 	microcode start address = 0x0004, boot configuration = 0x01

None of the issues should affect this hardware.

CC: (none) => tarazed25

Comment 4 Len Lawrence 2023-03-05 12:27:47 CET
Kernel: 5.15.88-desktop-1.mga8 x86_64
Mobo: Intel model: NUC12WSBi7
8-Core model: 12th Gen Intel Core i7-1260P
Intel Alder Lake-P Integrated Graphics driver: i915
Realtek RTL8153 Gigabit Ethernet Adapter type: USB driver: r8152

Updated and rebooted:
installing microcode-0.20230214-1.mga8.nonfree.noarch.rpm from //home/lcl/qa-testing/x86_64
starting installing packages
created transaction for installing on / (remove=0, install=0, upgrade=1)
dracut: systemd-initrd needs systemd in the initramfs
dracut: systemd-networkd needs systemd in the initramfs
dracut: dracut module 'ifcfg' depends on 'network', which can't be installed
dracut: dracut-systemd needs systemd-initrd in the initramfs
dracut: dracut-squash only supports systemd bases initramfs
dracut: dracut module 'ifcfg' depends on 'network', which can't be installed
dracut: dracut-squash only supports systemd bases initramfs

$ dmesg | grep microcode
[    1.153555] microcode: sig=0x906a3, pf=0x80, revision=0x421
[    1.154056] microcode: Microcode Update Driver: v2.2.

Don't know what to make of that.

$ sudo journalctl -b | grep microcode
Mar 05 09:22:54 yildun kernel: microcode: sig=0x906a3, pf=0x80, revision=0x421
Mar 05 09:22:54 yildun kernel: microcode: Microcode Update Driver: v2.2.
Mar 05 11:09:38 yildun [RPM][159728]: erase microcode-0.20220809-1.mga8.nonfree.noarch: success
Mar 05 11:09:46 yildun [RPM][159728]: install microcode-0.20230214-1.mga8.nonfree.noarch: success

Looks like it is relevant to this machine.
Comment 5 Morgan Leijström 2023-03-05 18:20:35 CET
Installed on my "svarten" workstation
Works with  5.15.88-desktop-1.mga8 x86_64, but not with our released backport kernel-desktop-6.1.6-1.mga8-1-1.mga8.x86_64 !
Tried now both kernels twice.
(running 5.15.88 and i keep using it for now...)

System is using encrypted partition, on which it use LVM

I dont understand how to from that shell mount a USB to dump the rdsoreport file so I listed it instead and attach a photo of the last lines.

Strange lane that third from bottom: it say it cancel the resume operation, I ordered it to reboot. By device not found i assume it mean the swap partition.

CC: (none) => fri

Comment 6 Morgan Leijström 2023-03-05 18:23:40 CET
Created attachment 13731 [details]
last lines of rdsosreport.txt failed booting kernel kernel-desktop-6.1.6-1.mga8-1-1.mga8.x86_64 with LVM on LUKS
Morgan Leijström 2023-03-05 18:28:31 CET

CC: (none) => tmb
Keywords: (none) => feedback

Comment 7 Thomas Backlund 2023-03-05 18:38:03 CET
not really a microcode bug, it does not affect how lvm works...

I guess something in your system prevents dracut from generating a working initrd...

what kernel were you running when you installed the microcode update ?

Keywords: feedback => (none)

Comment 8 Morgan Leijström 2023-03-05 18:43:10 CET
Kernel running when updating was 6.1.6-desktop-1.mga8
Comment 9 Thomas Backlund 2023-03-05 18:45:32 CET
if you boot back to 5.15.88, then recreate the 6.1.6 initrd, does it work then ?
Comment 10 Morgan Leijström 2023-03-05 18:50:20 CET
What command line to use for that?
Comment 11 Thomas Backlund 2023-03-05 18:52:58 CET
dracut -f /boot/initrd-6.1.6-desktop-1.mga8.img 6.1.6-desktop-1.mga8
Comment 12 Morgan Leijström 2023-03-05 19:06:39 CET
Created attachment 13732 [details]
Dracut output executed for kernel 1.6.1

Same boot failure after
Comment 13 Thomas Backlund 2023-03-05 19:15:11 CET
first of fix this:

/etc/dracut.conf.d/51-mageia-resume.conf:add_device+="/dev/vg-mag/lv_swap"

dracut: WARNING: <key>+=" <values> ": <values> should have surrounding white spaces!
dracut: WARNING: This will lead to unwanted side effects! Please fix the configuration file.
Comment 14 Morgan Leijström 2023-03-05 20:07:55 CET
First I fixed the missing spaces.

( It is a mystery they were missing, I have only used Mageia tools.
But on my mga9 test machine there are correctly spaces in a similar setup.
So I will not file a bug on that for now )

Then I again executed
dracut -f /boot/initrd-6.1.6-desktop-1.mga8.img 6.1.6-desktop-1.mga8
And the text messages were identical minus the missing spaces - warnings.

Unfortunately boot still fail with 6.1.6

I also installed 6.1.15 from backports testing
And 5.15.98 from updates testing.

All fail.  Except 5.15.88.
Comment 15 Thomas Backlund 2023-03-05 20:16:00 CET
well, 5.15.88 has not had its initrd updated, so that does not say much...

you can try  5.15.98 update to see if that one works or not..

and you can of course  try to uninstall microcode package, recreate  the 6.1.6 initrd and see if it boots then...

if that does not help, something elsa has been screwed up in your system...
Comment 16 Dave Hodgins 2023-03-05 22:25:15 CET
To work around the swap volume not being available change the fstab line
for the swap to have "defaults,nofail 0 0". Also add it for /home.
Then run the dracut -f command for each initrd. Double check that /boot is not
full.

That should allow booting, but will still leave figuring out why /home and the
swap are being inactivated. Once it boots, login as root on a tty and
run "lvchange -ay <logical volume>" for the swap and home volume groups and
reboot.

Pay attention to the messages during reboot. Is the kernel crashing just before
the reboot? If so, I think that may be why the volume groups are being
deactivated. It's been a long time since I used lvm. Well before systemd,
so am not familiar with how it handles activation.

CC: (none) => davidwhodgins

Comment 17 Morgan Leijström 2023-03-05 22:42:48 CET
Oh, I was under the impression initrd was supposed to be automatically updated for all kernels, but it only updates the initrd for the running kernel version?

Apparently that have saved me...

I will not downgrade initrd because of fear to wreck my only booting kernel alternative.

Yes something goes wrong...

As I wrote 5.15.98 also did not work
Now I tried old 5.15.79, also stops booting.

I am out of time for a while and need this machine working so will not mess with this any more for a while.

Thank you for the tips though :)
Comment 18 Dave Hodgins 2023-03-05 22:53:33 CET
What ever the lvm problems are, that is not related to the microcode update.
If the problems continue, please open a new bug report for that.
Comment 19 Herman Viaene 2023-03-06 10:10:06 CET
MGA8-64 kernel-server 5.15.88-server on Acer Aspire 5253
No installation issues
Rebooted after update, no problems seen.

CC: (none) => herman.viaene

Comment 20 Morgan Leijström 2023-03-08 19:08:26 CET
mga8-64 OK here;
Tested OK on a laptop Acer Aspire 7, with kernels
kernel-desktop-5.15.82-1.mga8-1-1.mga8
kernel-desktop-5.15.88-1.mga8-1-1.mga8
kernel-desktop-5.15.98-1.mga8-1-1.mga8
(the old already installed i issued dracut-f for manually)

And this was a similar system to the failing system, ext4 partitions in LVM on LUKS, and here no problem so definately system specific.
One detail that was the same was the missing spaces in 51-mageia-resume.conf, comment 13.

Hardware: laptop Acer Aspire7 A717-71G; Intel EFI, LVM on LUKS on NVMe, i5-7300HQ, using the intel GPU, wifi=QCA6174
Thomas Backlund 2023-03-11 18:44:11 CET

Whiteboard: (none) => MGA8-64-OK, MGA8-32-OK
Keywords: (none) => advisory, validated_update
CC: (none) => sysadmin-bugs

Comment 21 Mageia Robot 2023-03-11 20:02:00 CET
An update for this issue has been pushed to the Mageia Updates repository.

https://advisories.mageia.org/MGASA-2023-0085.html

Status: NEW => RESOLVED
Resolution: (none) => FIXED


Note You need to log in before you can comment on or make changes to this bug.