New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

Comments
Yes exactly, this output means it is not exploitable (since
algif_aeadmodule is not present).Yes this, and also python should be 3.10+ . It wont work on lower versions. In Rhel 8 family and Slackware python was +/- 3.9.25 thats why it was not working.
hey let me rawdog this exploit poc on my servers
https://github.com/tgies/copy-fail-c
no python needed, static binary that you can cross compile. uses some suid binary or change passwd to impersonate some other 4-digit user. scary.
Me the only user on all my VPS servers.
I am root on all my servers.
It does work, and it is available in go too (https://github.com/badsectorlabs/copyfail-go)
If regular users have no access to c dev. environment, they can just go golang.
My temporary "patch" is to follow the instructions at https://mondoo.com/vulnerability-intelligence/vulnerability/MONDOO-CVE-2026-31431, to disable algif_aead module.
So it is scary but not that scary. It's not "every linux kernel since 2007" it's more like "every kernel with some module that is 50/50 not even there and 80% of people do not need it or even know what it is" situation.
Running yabs and turn off the vps prevents hacking
Can't get hacked if OOM
Even though it's based on a non-root user escalating the privilege, we were quick at closing this just in case, and apply kernel patches where applicable.
I've built a free packages scanner so you can check how bad your server is with vulnerabilities: https://fivenines.io/tools/linux-package-vulnerability-scanner
Well, it's there on Debian, Ubuntu, RHEL, and RHEL clones by default.
Not in RHEL7 (and clones), and it's distribution from 2014.
One simple way to protect against this and similar vulnerabilities is just to disable module autoloading:
Unlike
kernel.modules_disabled = 1, this can be reverted by root and doesn't block manual module loading or unloading. All it does is prevent unprivileged processes from triggering module loading by requesting some obscure functionality from the kernel like a rare socket family.Updated ...
Am I the only one who think Ubuntu sucks in security nowadays?
Debian took a day to release a new kernel patching dirty frag, and still nothing from Ubuntu. Not in 24.04, nothing in 26.04. Waiting for a week now, that is not good.
New updates are available for debian.
It's usually the opposite. Ubuntu promptly releases fixes for even minor security issues while Debian only does emergency releases for major issues. Ubuntu also uses more compile-time hardening than Debian.
Nope, you are not.
And this 404 is also not good: https://ubuntu.com/security/CVE-2026-46333 (Sunday, 22:20 UTC)
A three week security cycle on kernel builds just does not help if you get a potential zero day every friday...