Light

Episode #967 Quiz

GoFetch
Date: 2024-03-26 | Length: 1.75 hrs | Episode page at twit.tv

About this episode

GoFetch exploits Apple and Intel data memory-dependent prefetchers: cacheable pointer-like data triggers speculative dereferencing, creating a cache side channel that leaks cryptographic secrets even from constant-time implementations. The attack can extract RSA, Diffie-Hellman, Kyber, and Dilithium keys locally. Mitigations include disabling DMP on M3 and moving crypto to safer cores.

Your name and email are stored only in your browser local storage for convenience. They are not retained server-side.

NameEmail

Question 1: According to the episode, what did Apple add to M3 chips that makes the GoFetch issue avoidable on that generation?

Question 1

According to the episode, what did Apple add to M3 chips that makes the GoFetch issue avoidable on that generation?

A microcode update that randomizes pointer dereferences A hardware AES engine that isolates key material from memory A switch to disable DMP during cryptographic operations A new cache partitioning mode for secret-bearing processes

Question 2: What was the key technique the researchers used to turn DMP into a key-extraction side channel?

Question 2

What was the key technique the researchers used to turn DMP into a key-extraction side channel?

Forcing secret data into shared TLB entries for cache probing Crafting chosen inputs so intermediate values resemble pointers Exploiting speculative execution on mispredicted indirect jumps Measuring branch predictor state after secret-dependent branches

Question 3: Which of the following accurately summarizes the affected Apple chip generations as Steve described them?

Question 3

Which of the following accurately summarizes the affected Apple chip generations as Steve described them?

M1 and M2 are affected on efficiency cores, while M3 has no DMP at all Only M3 is affected, because older chips never included DMP M1 and M2 are affected on performance cores, while M3 can disable DMP All Apple silicon is equally affected on every core type

Question 4: Why did Steve say the headline claims about GoFetch were overstated?

Question 4

Why did Steve say the headline claims about GoFetch were overstated?

Because the issue was not new, not a chip flaw, and not unpatchable Because the researchers had not demonstrated any working proof of concept Because Apple had already publicly fixed every affected library Because the attack only worked against outdated, unpatched Intel processors

Question 5: What did Steve identify as the likely practical impact if the attack were used in the real world?

Question 5

What did Steve identify as the likely practical impact if the attack were used in the real world?

Only cloud-hosted workloads would be vulnerable to the leakage Remote exploitation would work through a malicious web page Local malware would be needed to target a victim machine A physical attacker would need to probe the chip with external hardware

Cancel

My Data//Security & Privacy//Clear Local Data

© 2026 Ken ArmstrongSecurity Now™ is a trademark of TWiT LLC. All trademarks and copyrights of the Security Now podcast are respected and reserved to their respective holders.
This is a hobby project made with love and admiration for what Steve Gibson and Leo Laporte have built over 20 years.