SLAM Attack Gets Root Password Hash in 30 Seconds by Exploiting Hardware Security

Spectre is a class of speculative execution vulnerabilities in microprocessors that can allow threat actors unauthorized access to sensitive data.

Hackers exploit Spectre because it enables them to extract confidential information by manipulating the speculative execution capabilities of CPUs, bypassing traditional security measures.

Cybersecurity researchers at Systems and Network Security Group of VU Amsterdam recently identified SLAM attack that gives attackers access to the root password hash in 30 seconds by exploiting the hardware security.

Disclosing

[

#SLAM

](https://twitter.com/hashtag/SLAM?src=hash&ref_src=twsrc%5Etfw)

, aka how to combine Spectre and Intel LAM (& co.) to leak kernel memory on future CPUs (demo below). Thousands of exploitable "unmasked" (or pointer chasing) gadgets in the Linux kernel. Joint work by

[

@MatheHertogh

](https://twitter.com/MatheHertogh?ref_src=twsrc%5Etfw)

[

@SanWieb

](https://twitter.com/SanWieb?ref_src=twsrc%5Etfw)

[

@c_giuffrida

](https://twitter.com/c_giuffrida?ref_src=twsrc%5Etfw)

:

[

https://t.co/8sgL8t8eC5

](https://t.co/8sgL8t8eC5)

[

pic.twitter.com/7XFf8BBAN9

](https://t.co/7XFf8BBAN9)

— VUSec (@vu5ec)

[

December 5, 2023

](https://twitter.com/vu5ec/status/1732099516621521003?ref_src=twsrc%5Etfw)

SLAM Attack Gets Root Password

SLAM probes Spectre’s residual attack space on current and future CPUs with Intel LAM.

It bypasses new transient execution methods, like BHI or Inception, and exploits an overlooked class of Spectre disclosure gadgets, avoiding standard “masked” gadgets using secret data to index arrays.

Secret data to index arrays (Source – VUSec)

Uncommon code patterns limit standard Spectre gadgets in regular software, like the Linux kernel with believed non-exploitable gadgets.

While the SLAM concentrates on various code patterns, especially pointer-chasing snippets, “unmasked” devices that exploit confidential data as pointers are produced.

Besides this, the common code breeds widespread unmasked gadgets. However, the scanner developed by the researchers discovered tens of thousands in Linux, with hundreds ripe for exploitation.

SLAM uses unmasked gadgets for userland data leaks (ASCII kernel data). Not only that, but researchers even extract root password hash in under 30 seconds on the latest Ubuntu, emulating Intel LAM.

CPUs Impacted

SLAM targets future CPUs, exploiting upcoming linear address masking features like Intel’s LAM and AMD’s UAI.

Despite being designed for security, SLAM reveals that loosening canonicality checks can expose vulnerabilities, affecting even CPUs with weak checks.

Here below, we have mentioned all the CPUs that are impacted:-

• Existing AMD CPUs are vulnerable to CVE-2020-12965.

• Future Intel CPUs supporting LAM (both 4- and 5-level paging).

• Future AMD CPUs support UAI and 5-level paging.

• Future Arm CPUs are supporting TBI and 5-level paging.

SLAM innovates data leakage through hidden gadget channels, exploiting address masking and microarchitectural race conditions on AMD CPUs.

To get around SMAP mitigation, it switches cache covert channels with translation-based ones. For an exploitable covert channel, security experts use sliding methods and just-in-time reload buffers.

The post SLAM Attack Gets Root Password Hash in 30 Seconds by Exploiting Hardware Security appeared first on GBHackers on Security | #1 Globally Trusted Cyber Security News Platform.
posted by pod_feeder