Hey, I’ve been hearing a LOT about the xz backdoor. Crazy story, but rather than reading 10 different articles about it from 3 days ago when the story was quite new, does anybody know a high quality write-up that has all the juicy details and facts? I really like in-depth guides that cover every aspect of the story.
Thanks in advance guys!
That’s way more steps than I thought
I wonder if something so complex makes it easier to potentially detect than something slightly simpler.
Btw what does the .o file do, is that the one running inside ssh through system d?
I am by far an expert, but imo it seems it needed to be complex to achieve its goals, and to get that level of change in without notice required breaking it down and sneaking pieces in over time.
Happy to be corrected of course if this is a wrong interpretation
You’re correct. Right now many experts are scrambling all those small pieces together as to how this could happen in the first place, as a lot of it was public too: the social pressure on maintainers, random software changes that now seem suspicious, and the absence of a real identity of the perpetrators. Every expert who’s onto this seems to be a real person, with a real identity and a real face to the name.
Prior to the first commits, there must have been months if not years of planning too.
But just the fact that some of the code those perpetrators wrote took +0.5 seconds more for something that would normally only take 0.2-0.3 seconds is what gave them away.
0.5 seconds of CPU time vs. years of planning.
It’s an intriguing story.
They tried so hard, they got so far, but in the end, it didn’t even matter.
I wonder exactly why ssh was taking so long more. Perhaps the bits that scan ssh logs with a regex to extract IP address and username?
Whatever it is, that particular bit should be easy to deactivate since somehow a full fledged binary file with executable code was being bundled. I can imagine it only being active under a toggle that would make it harder to detect, such as a specific time of day.
That’s my take as well
Not everything is known yet, so you won’t find a comprehensive summary. The latest news I’ve seen: https://www.openwall.com/lists/oss-security/2024/03/30/36
I’m watching some folks reverse engineer the xz backdoor, sharing some preliminary analysis with permission. The hooked RSA_public_decrypt verifies a signature on the server’s host key by a fixed Ed448 key, and then passes a payload to system(). It’s RCE, not auth bypass, and gated/unreplayable.
All the technical details that are widely known (and some that aren’t yet) seem to be there, including the original report from Andres Freund. For rumours about who might be behind it and high-level speculation about what it all means, you’d have to look elsewhere.
thx :)
https://boehs.org/node/everything-i-know-about-the-xz-backdoor gives a good overview with links to further reads.
awesome, thanks!
Sam Jones’s FAQ is by far the best single source, links to other solid sources for more in-depth technical details and also lightly debunks a few things.
The main thing sources online disagree on are which distros are affected. That’s because it’s not a simple yes/no and some distros are taking a nuanced approach in their public communication, while others have chosen the sledgehammer in an attempt to get people to upgrade their systems but keep/kept the nuance in the back room where the audience understood not everything was known yet. Some distros are underselling how vulnerable they were, others are overselling it.
I read a few articles. I think Andres Freund’s announcement gave me the best context for the exploit itself. https://www.openwall.com/lists/oss-security/2024/03/29/4
The most helpful source I saw on which systems are affected was this Lemmy post, https://beehaw.org/post/12813772
Best coverage I’ve seen so far has been on Ars Technica.
Yeah, their original coverage was good, but they also had a really good follow up today.
thanks!
It’s still ongoing and still being reverse engineered, so I would expect the good writeups to come in a couple days.