1y ago

Backdoor found in widely used Linux utility breaks encrypted SSH connections | Ars Technica

Backdoor found in widely used Linux utility breaks encrypted SSH connections

TL;DR there was a backdoor found in the XZ program. All major distros have been updated but it is recommended that you do a fresh install on systems that are exposed to the internet and that had the bad version of the program. Only upstream distros were affected.

42 comments

stable release of Arch Linux is also affected. That distribution, however, isn't used in production systems.
Don't tell me how to live my life, Ars Technica.
- "stable" release of Arch?
  
  They mean a variant you use in a stable, like to run an automatic feeder for horses. According to Ars Technica, however, you are not to use it in your production stable.
  
  yes, like my marriage
- I am not deep enough in it, but from the arch-announce mailinglist:
  From the upstream report [1]:
  openssh does not directly use liblzma. However debian and several other distributions patch openssh to support systemd notification, and libsystemd does depend on lzma.
  Arch does not directly link openssh to liblzma, and thus this attack vector is not possible. You can confirm this by issuing the following command:
  
  $(command -v sshd)
  However, out of an abundance of caution, we advise users to remove the malicious code from their system by upgrading either way. This is because other yet-to-be discovered methods to exploit the backdoor could exist.
  https://www.openwall.com/lists/oss-security/2024/03/29/4
- You should not run Arch in production. Boom, I said it
  
  Well I don't see any cops.
- Ars Technica sounds like a weirdo to me these days. Loves to attack big techs (although understandable), now adds this to their description of Arch.
This was basically a lucky catch. Sadly makes you wonder how many backdoors like that have not been found (yet). Never the less the distro model of not feeding in upstream binaries directly is an important part of the multi-barrier security.
- One still could hide something in source code. I think we need to just be more security aware in general. Having source code isn't useful if someone deliberately put a security hole in it
Catching this now is pretty huge, because it mainly targets distro build systems. Had this gone undetected, we'd be in shiznit creek couple of years down the line.
For those on Android running Termux, it is also affected. Just checked my version of xz-utils and it was 5.6.1. Running "pkg upgrade" will roll back to version 5.4.5 (tagged as "5.6.1+really5.4.5" for both liblzma and xz-utils packages).
- Makes you wonder why Termux ships the latest stuff. It might be smart to allow more time for critical problems to get caught.
  
  Probably for the exact same reason this backdoor was introduced. Users complain about slow feature rollouts so (unpaid) devs (maintaining software in their spare time out of the kindness of their hearts) cut corners. In some situations that looks like bringing on a second maintainer without thorough vetting, in others it looks like importing upstream packages without thorough vetting.
  Don't blame the Termux devs here, blame the community that keeps pushing them to move faster.
- So if I am seeing this I am good?
  
  Yeah, that's what mine "upgraded" to. All that update does is rollback to 5.4.5.
This is the best summary I could come up with:
Researchers have found a malicious backdoor in a compression tool that made its way into widely used Linux distributions, including those from Red Hat and Debian.
An update the following day included a malicious install script that injected itself into functions used by sshd, the binary file that makes SSH work.
So-called GIT code available in repositories aren’t affected, although they do contain second-stage artifacts allowing the injection during the build time.
In the event the obfuscated code introduced on February 23 is present, the artifacts in the GIT version allow the backdoor to operate.
“This could break build scripts and test pipelines that expect specific output from Valgrind in order to pass,” the person warned, from an account that was created the same day.
The malicious versions, researchers said, intentionally interfere with authentication performed by SSH, a commonly used protocol for connecting remotely to systems.
The original article contains 810 words, the summary contains 146 words. Saved 82%. I'm a bot and I'm open source!
Just makes you wonder what else (if anything) is backdoored. I am seriously 🤏 this close to just switching all of my boxes over to OpenBSD.
The last time someone over there was approached about backdooring a related piece of software (which they refused), the OpenBSD devs manually screened the entire codebase, just in case something got in.
Really, the only things I'd miss would be Minecraft, KDE, and Mullvad Browser; and of course I'd have to buy a couple more WiFi dongles (or learn how to port drivers from Linux).
- I honestly think BSD has the potential to be worse due lack of people. I think the best option is to not be paranoid as a user. If someone needs to be paranoid it is the maintainers.
  
  I think the best option is to not be paranoid as a user.
  Yeah, just never be a dissident, or a whistleblower, or an activist, or a member of a vulnerable marginalized group. Remember, if you obey there's no reason to fear being spied on.
  I really don't think you understand how serious this kind of backdoor is. It puts certain people in real danger.
- BSD is the way to go for ultimate security imo, but the lack of software support makes it difficult for even the average techie.
Meanwhile non SystemD systems like NetBSD FreeBSD OpenBSD are safer.
- What does this have to do with systemd? Aren't they safer in this situation because they aren't using the beta xz release?
  My systems running Debian stable with systemd also aren't affected...
  
  This particular backdoor affects sshd on systems that use libsystemd for logging.
  your Debian system is probably not affected because Debian stable doesn't update packages very quickly. You're probably on an older release of the backdoored package.
  
  The reason openssh links liblzma in the first place is to enable a systemd feature, so naturally "systemd bad, it's proximity to a security issue is yet more proof that a pile of shell scripts in a trenchcoat is a superior init system" etc
- You can have a nefarious developer working for a nation state infiltrate the supply chain for ANY OS.
  You don’t know.
  
  Agreed
- That is not really true. If anything they would be easier to infiltrate
- What's Wayland support for the BSDs like?
  
  https://docs.freebsd.org/en/books/handbook/wayland/

42 comments