- New Linux kernel flaw CVE‑2026‑46300 “Fragnesia” allows local attackers to gain root
- Discovered by William Bowling of Zellic;
- PoC shows corruption of /usr/bin/su page cache to get root shell
Security researchers have discovered a new vulnerability in the Linux kernel which could allow malicious actors to run code with elevated privileges, exposing systems to risk of data theft, malware deployment, and even full device takeover.
The vulnerability is tracked as CVE-2026-46300, and was given a severity score of 7.8/10 (high). It’s nicknamed Fragnesia and is apparently in the same vulnerability class as Dirty Frag, another kernel bug that was disclosed recently.
While Dirty Frag chains multiple flaws, Fragnesia comes in the form of a logic bug in the Linux XFRM EST-in-TCP subsystem. By writing arbitrary bytes to the kernel page cache of read-only files, unprivileged local attackers can gain root privileges, thus compromising the entire system.
Patches and killswitches
The bug was discovered by William Bowling of Zellic, who also shared a proof-of-concept (PoC) that “achieves a memory-write primitive in the kernel that is used to corrupt the page cache memory of the /usr/bin/su binary to get a shell with root privileges.”
"Fragnesia is a member of the Dirty Frag vulnerability class. This is a separate bug in the ESP/XFRM from dirtyfrag which has received its own patch. However, it is in the same surface and the mitigation is the same as for dirtyfrag," Bowling said. "It abuses a logic bug in the Linux XFRM ESP-in-TCP subsystem to achieve arbitrary byte writes into the kernel page cache of read-only files, without requiring any race condition."
To mitigate the risk, Linux users should apply kernel updates for their distros without delay.
Linux kernel vulnerabilities are a hot topic these days. Prompted by both Dirty Frag and Copy Fail, two recently disclosed flaws, co-maintainer Sasha Levin proposed a new patch that would allow system administrators to temporarily disable a vulnerable kernel function.
That way, if security researchers discover malicious code in the future, users would be able to quickly instruct the kernel not to use it. The feature would not address underlying issues, but since the function would return an error, it could prevent the vulnerability from causing any serious harm before a proper patch is deployed.
The new feature is currently being reviewed by the Linux community and has not yet been officially introduced.
Via BleepingComputer
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.