openSUSE maintainers received notification of a supply chain attack against the “xz” compression tool and “liblzma5” library.
Background
Security Researcher Andres Freund reported to Debian that the xz / liblzma library had been backdoored.
This backdoor was introduced in the upstream github xz project with release 5.6.0 in February 2024.
Our rolling release distribution openSUSE Tumbleweed and openSUSE MicroOS included this version between March 7th and March 28th.
SUSE Linux Enterprise and Leap are built in isolation from openSUSE. Code, functionality and characteristics of Tumbleweed are not automatically introduced in SUSE Linux Enterprise and/or Leap. It has been established that the malicious file introduced into Tumbleweed is not present in SUSE Linux Enterprise and/or Leap.
Impact
Current research indicates that the backdoor is active in the SSH Daemon, allowing malicious actors to access systems where SSH is exposed to the internet.
As of March 29th reverse engineering of the backdoor is still ongoing.
Mitigations
openSUSE Maintainers have rolled back the version of xz on Tumbleweed on March 28th and have released a new Tumbleweed snapshot (20240328 or later) that was built from a safe backup.
The reversed version is versioned 5.6.1.revertto5.4 and can be queried with rpm -q liblzma5.
User recommendation
For our openSUSE Tumbleweed users where SSH is exposed to the internet we recommend installing fresh, as it’s unknown if the backdoor has been exploited. Due to the sophisticated nature of the backdoor an on-system detection of a breach is likely not possible. Also rotation of any credentials that could have been fetched from the system is highly recommended. Otherwise, simply update to openSUSE Tumbleweed 20240328 or later and reboot the system.
More Information about openSUSE:
Damn this is really bad. Good that they fixed it.
I wouldn’t be surprised if older versions have a backdoor too as they were a maintainer for 2 year so who knows what they added.
Damn… damn!!
Someone is going back over their contributions, right?
Right?
Right, here’s an enlightening synopsis by Evan Boehs:
https://boehs.org/node/everything-i-know-about-the-xz-backdoor
It’s good that it didn’t get into Leap or Enterprise. Servers for businesses shouldn’t generally be using Tumbleweed.
It sounds like to be really vulnerable a machine has to expose SSH to the internet. So if that’s correct then most typical home computers should be safe. Based on that assumption I’m not rushing to wipe and reinstall.
What scary is the maintainer that insert the backdoor has been main maintainer for
xzfor the last two years. Who know if they have other backdoors inserted in the last 2 years? Investigation is still ongoing so I expect more juicy revelations in the next few days.doesn’t pushing to github (and probably a selfhosted equivalent) require ssh to do without entering your password every single time?
It’s quite interesting to see that the developer who has embedded the backdoor “simplified” the SECURITY.md file just 4 days ago, by deleting information about how to report a security vulnerability.
The reason it’s really interesting is that if I understand it right, the backdoor was just discovered less than a day ago.https://github.com/tukaani-project/xz/commit/af071ef7702debef4f1d324616a0137a5001c14c
I think they only started pushing for updated version on Fedora and Debian a few days ago.
I assume SSH is not exposed to the internet by default on openSUSE? I have not used SSH on my install so should I be safe if I just update?
On desktop systems, the SSH server is disabled by default. It should say “disabled” in
systemctl status sshd, then you’re good.On server systems, the SSH server is enabled, but you would usually know that it accepts incoming connections from the internet.
So, a home-server would need to be exposed in your modem/router. And if you’re hosting on a VPS, well, then it’s obviously going to be reachable.If you didn’t explicitly open the SSH port during the install, you are okay.
So this is what happens when package maintainers fail to find the problematic bits during package updates. I’ll be honest, after seeing how Linux package management is done (automation and semi-automation galore) and by whom (people who often don’t know the programming language of the source and who don’t have much time either), I am more surprised that it took this long.
In this case the upstream maintainer heavily obfuscated the code to be able to compromise ssh. Package maintainers aren’t responsible for vetting for that.
The original email talks about a line that is in the release tar balls but not the repository itself that actually arms the exploit. This seems like something a maintainer should be able to verify.
Not saying that they should have immediately seen that that is an exploit, the exploit is obfuscated very well. But this should be a big red flag right?
As a Homebrew maintainer, what is there to red flag about a project providing tarballs of their source?
We would have to red flag pretty much every project that uses autoconf (since those usually provide a tarball where the user doesn’t have to run
autoreconf)
Yes, ~it took a security researcher to find this out~. Package manager maintainers failed, and they were whom the average Linux advocate supposed will catch the back door, whenever they boasted about the safe-ness of Linux.
deleted by creator
OP wrote security researcher reported to Debian, so I’m mildly confused to read your reply.
deleted by creator
Right. His Mastodon bio also says he’s a postgres dev at MS. Something went wrong by the time this post was done.
It would also be interesting to know if any OS was in feature freeze recently that is not supposed to get updates anymore. I was thinking about Android, OpenWRT or any of those where people update update their devices more rarely
OpenWRT does not use liblzma or systemd so i think that one is pretty safe. I would also be surprised if Android included OpenSSH server binaries in that way.
Hi there! Looks like you linked to a Lemmy community using a URL instead of its name, which doesn’t work well for people on different instances. Try fixing it like this: [email protected]
