I admittedly stopped reading halfway through but I feel like these newest vulnerabilities being discovered are probably just fucking government back doors the manufacturers have been forced to include.
I can’t comment on the general trend, but this specific one seems a bit too circumstantial to be of use for a serious spying effort. You’d have to have the spyware running parallel to the apps usong passwords you want to steal in a specific way.
The risk exists, which is bad enough for stochastic reasons (eventually, someone will get lucky and manage to grab something sensitive, and since the potential damage from that is incalculable, the impact axis alone drives this into firm "you need to get that fix out asap), but probably irrelevant in terms of consistency, which would be what you’d need to actually monitor anyone.
If you manage to grab enough info to crack some financial access data, you can steal money. If you can take over some legit online account or obtain some email-password combo, you can sell it. But if you want to monitor what people are doing in otherwise private systems, you need some way to either check on demand or log their actions and periodically send them to your server.
It would be far more reliable to have injection backdoors to allow you access by virtue of forcing a credential check to come up valid than to hope for the lucky grab of credentials the user might change at an arbitrary moment in time.
I think it also means software running can access other software’s memory which is probably bad but personally I’m not keen for that performance hit on my desktop
Seems very similar to Zenbleed in terms of using certain register optimisation and speculative execution to get crippling security exploits. Thus far I haven’t read too much into the detail of the attack but This article on Zenbleed, written by the attack’s author, describes how the attack in detail and how he came to find it using fuzzing techniques - in this case two sets of instructions that should have had the same result, but they didn’t.
If it’s anything like the industry that I work in, the CEO would have been informed of the short comings of the design numerous times and given a response along the lines of “does it make our CPUs faster and more powerful though?”.
The CEO won’t be pissed of at his chip designer, they’ll be pissed because they’ve been caught out.
…All Ryzen and Epyc processors were found to have a very similar bug not too long ago, it’s actually addressed in the article. You might want to read it…
Here is a good write-up of Zenbleed for the Ryzen 2 and up vulnerability. It uses similar register optimisation and speculative execution to get the same effect.
Every article is a copy paste of the same bullshit talking about the vulnerability and pointing to the stupid cryptic list of processors that requires you to jump through hoops to read it. You can’t just search for your processor in a database I mean fuck that would take them at least an a couple hours of their precious time to set up and they have only had a year. How do you fix it? Why with a microcode update of course!!..from where you ask? Well don’t worry just look at the cryptic list it will tell you if you need a microcode update!!
Fuck every article about this shit. Anyone wanna bust an Eli5 on how to fix this problem for people? (I was assuming it’s a BIOS update but the articles have only confused me further)
You can’t just search for your processor in a database I mean fuck that would take them at least an a couple hours of their precious time to set up and they have only had a year. How do you fix it?
I figured out how to do it fairly quickly but it would be a hell of a lot easier if people could just type in “11700K” in a box on a web page or something and it could just tell them. Or they could have added a little bit of code to their CPU ID utility that says “yupp your processor is effected by the flaw”. I am mostly annoyed at all this not for me but for all the people who would read those pages and the contents would seem like an insane foreign language to them all while articles are telling them it’s a major security flaw that would allow people to steal their encryption keys.
ELI5, or ELIAFYCSS (Explain like I’m a first year CS student): modern x86 CPUs have lots of optimized instructions for specific functionality. One of these is “vector instructions”, where the instruction is optimized for running the same function (e.g. matrix multiply add) on lots of data (e.g. 32 rows or 512 rows). These instructions were slowly added over time, so there are multiple “sets” of vector instructions like MMX, AVX, AVX-2, AVX-512, AMX…
While the names all sound different, the way how all these vector instructions work is similar: they store internal state in hidden registers that the programmer cannot access. So to the user (application programmer or compiler designer) it looks like a simple function that does what you need without having to micromanage registers. Neat, right?
Well, problem is somewhere along the lines someone found a bug: when using instructions from the AVX-2/AVX-512 sets, if you combine it with an incorrect ordering of branch instructions (aka JX, basically the if/else of assembly) you get to see what’s inside these hidden registers, including from different programs. Oops. So Charlie’s “Up, Up, Down, Down, Left, Right, Left, Right, B, B, A, A” using AVX/JX allows him to see what Alice’s “encrypt this zip file with this password” program is doing. Uh oh.
So, that sounds bad. But lets take a step back: how bad would this affect existing consumer devices (e.g. Non-Xeon, non-Epyc CPUs)?
Well good news: AVX-512 is not available on most Intel/AMD consumer CPUs until recently (13th gen/zen 4, and zen 4 isn’t affected). So 1) your CPU most likely doesn’t support it and 2) even if your CPU supports it most pre-compiled programs won’t use it because the program would crash on everyone else’s computer that doesn’t have AVX-512. AVX-512 is a non-issue unless you’re running Finite Element Analysis programs (LS-DYNA) for fun.
AVX-2 has a similar problem: while released in 2013, some low end CPUs (e.g. Intel Atom) didn’t have them for a long time (this year I think?). So most compiled programs wouldn’t compile with AVX-2 enabled. This means whatever game you are running now, you probably won’t see a performance drop after patching since your computer/program was never using the optimized vector instructions in the first place.
So, the affect on consumer devices is minimal. But what do you need to do to ensure that your PC is secure?
Three different ideas off the top of my head:
BIOS update. The CPU has a some low level firmware code called microcode which is included in the BIOS. The new patched version adds additional checks to ensure no data is leaked.
Update the microcode package in Linux. The microcode can also be loaded from the OS. If you have an up-to-date version of Intel-microcode here this would achieve the same as (1)
Re-compile everything without AVX-2/AVX-512. If you’re running something like Gentoo, you can simply tell GCC to not use AVX-2/AVX-512 regardless of whether your CPU supports it. As mentioned earlier the performance loss is probably going to be fine unless you’re doing some serious math (FEA/AI/etc) on your machine.
It’s the AVX-2/AVX-512 instructions that have issues. In most cases unless you’re running a server CPU (or extremely recent consumer CPU) you’ll be fine.
Scary for HPC/AI? Yes. For most people? Not really.
I would say you’ll be fine. Most games don’t compile with avx-2 anyways since it’ll crash if you run it on something that doesn’t have them (which is a lot of CPUs) and AVX-512 is straight up only available on Xeons, Epyc and zen 4. Nobody is going to use that for consumer software.
The only game I can think of using AVX is a Skyrim mod for realistic physics, where the author provided binaries for AVX-2/AVX-512. So it won’t affect most compiled applications much since you need to compile with it first (which almost nobody does).
Add comment