gchristensen changed the topic of #nixos-security to: Vulnerability Roundup Issues: https://github.com/NixOS/nixpkgs/issues?utf8=%E2%9C%93&q=is%3Aissue+is%3Aopen+Vulnerability+roundup + https://broken.sh
Synthetica has quit [Quit: Connection closed for inactivity]
copumpkin has quit [Ping timeout: 246 seconds]
justanotheruser has quit [Ping timeout: 255 seconds]
pie__ has quit [Ping timeout: 258 seconds]
pie___ has quit [Ping timeout: 258 seconds]
Synthetica has joined #nixos-security
pie_ has joined #nixos-security
pie__ has joined #nixos-security
pie_ has quit [Ping timeout: 245 seconds]
<pie__> in the news, though i suppose youve already seen: chosen prefix attack on sha1
<pie__> or something
<pie__> actually hold on, this looks like it predicates on already having a collision?
<pie__> " In particular, we have a chosen-prefix collision attack againstSHA-1with complexity between 266.9and 269.4(depending on assump-tions about the cost of finding near-collision blocks), while the best-known attack has complexity 277.1. This is within a small factor of thecomplexity of the classical collision attack onSHA-1(estimated as 264.7).This represents yet another warning that industries and users have
<pie__> tomove away from usingSHA-1as soon as possible"
<pie__> thats to be read as 2^66.9 and etc
<pie__> not sure what implications this had for git
<pie__> random google yields https://github.com/git/git/blob/master/Documentation/technical/hash-function-transition.txt but "Add SHA-256 support to Git protocol." is a non-goal in that document
<pie__> "This is valuable and the logical next step but it is out of scope for this initial design."
<infinisil> I'm surprised the paper doesn't mention git
<pie__> didnt know this "Why do collisions matter for Git’s security?
<pie__> If a Git fetch or push tries to send a colliding object to a repository that already contains the other half of the collision, the receiver can compare the bytes of each object, notice the problem, and reject the new object. Git has implemented this detection since its inception."
<gchristensen> git's store is content addressed, so why bother sending (or receiving) an object you already have
<pie__> well yeah
<pie__> so does that mean when you push you retrieve an list of remote hashes?
<pie__> or..hm
<pie__> i guess you wouldnt need to since you have a remote pointer and a local pointer and can just send whats in between
<pie__> well, i should look up how git syncing works if i want to know
<pie__> that mitigation seems to be specific to SHAttered though
erictapen has joined #nixos-security
<pie__> "The Git project is also developing a plan to transition away from SHA-1 to another, more secure hash algorithm, while minimizing the disruption to existing repository data. As that work matures, we plan to support it on GitHub." seems encouraging
<gchristensen> it is a bit complicated because all the old hashes will still exist
justanotheruser has joined #nixos-security
erictapen has quit [Ping timeout: 258 seconds]
erictapen has joined #nixos-security
erictapen has quit [Ping timeout: 244 seconds]
erictapen has joined #nixos-security
pie___ has joined #nixos-security
pie__ has quit [Ping timeout: 255 seconds]
pie__ has joined #nixos-security
pie___ has quit [Ping timeout: 246 seconds]
andi- has quit [Remote host closed the connection]
andi- has joined #nixos-security
pie__ has quit [Ping timeout: 258 seconds]
pie_ has joined #nixos-security
<gchristensen> cool cool cool
<pie_> oh geez is this another cpu thing
galaxie has left #nixos-security [#nixos-security]
<gchristensen> I thought everybody turned off hyperthreading by now
<pie_> ah, its from the usual crows
<andi-> You are still using computers? :D
<pie_> * crowd
<gchristensen> let's all just dump this nixos stuff and adopt templeos
<gchristensen> no worry about processor attacks when you can just read memory from any process at any time
<pie_> lol
<pie_> "We are the first to do post-processing of the leaked datawithin the transient domain to eliminate noise."
<pie_> (well, they did put that as the last item)
<pie_> i kind of meant that as "duh", but maybe its not so "duh"
<pie_> hm. too bad im preoccupied with other stuff i guess
<pie_> funny that I just put Computer Architecture: A Quantitative Approach on my desk again today
erictapen has quit [Ping timeout: 268 seconds]
erictapen has joined #nixos-security
erictapen has quit [Ping timeout: 246 seconds]
erictapen has joined #nixos-security
<pie_> huh this seems a pretty decent summary of multi factor authentication stuff https://blog.dcso.de/modern-authentication-services-more-than-passwords-plus-smart-card/
erictapen has quit [Ping timeout: 245 seconds]
<Foxboron> gchristensen: I'm considering taking up wood chopping.
<Foxboron> No worries about reading memory when there is none.
erictapen has joined #nixos-security
<gchristensen> still have rings of privilege though
<gchristensen> and hope those bugs don'tget past the outer rings
<Foxboron> Nothing to exploit if we just remove everything and the stem.
<Foxboron> and https://cpu.fail/ :)
<pie_> "how did you test it on all this hardware" https://mdsattacks.com/images/rack.jpg
erictapen has quit [Ping timeout: 246 seconds]
erictapen has joined #nixos-security
<pie_> gchristensen, do we have a nixos flag yet for fix_all_my_shit_i_dont_care_if_its_slow?
<pie_> because i havent been keeping up with mitigations
<pie_> how do i disable hyperthreading anyway
<pie_> unrelated: can we / do we blacklist known vulnerable kernel modules
<samueldr> there is at least `boot.initrd.luks.mitigateDMAAttacks` I know of
<samueldr> but here it's vulnerable _as designed_
<pie_> <gchristensen> I thought everybody turned off hyperthreading by now
Synthetica has quit [Quit: Connection closed for inactivity]
<samueldr> To protect users, Chrome OS 74 disables Hyper-Threading by default. For the majority of our users, whose workflows are primarily interactive, this mitigates the security risk of MDS without a noticeable loss of responsiveness. Chrome OS 75 will contain additional mitigations.
<samueldr> turns out google does :)
<samueldr> (for now)
<pie_> yeah so how do we :p
<pie_> or how do i i mean
<pie_> im kind of lazy / preoccupied to start googlign thart
<samueldr> AFAIUI the bios setting is the right one
<pie_> the blacklisted kernel modules thing was a throwback to the RDS thing yesterday
<pie_> <ggreer> The following packages will be upgraded:
<pie_> <ggreer> ... intel-microcode ...
<pie_> <ggreer> oh shit. what happened now?
<pie_> looks like debians got updates
<pie_> not surprising given the embargoed distro list i suppose
pie_ has quit [Ping timeout: 258 seconds]
<samueldr> and /sys/devices/system/cpu/smt/ for runtime things
<samueldr> too bad pie_ left
pie_ has joined #nixos-security
erictapen has quit [Ping timeout: 244 seconds]
pie_ has quit [Ping timeout: 258 seconds]