Conversation

Replying to
Big difference between a device with components that are strongly hardened, highly audited and have good ongoing security support vs. the complete opposite. Also, portraying it backwards by misinterpreting how DMA / IOMMUs work is just wrong.
1
Replying to
Devices you're talking about have entirely closed source hardware and firmware. If you choose components that are known to be insecure and also don't apply fixes to known security vulnerabilities, backdoors are a non-issue, because you have the front door wide open to attackers.
2
Replying to
So how does it help to use an SoC platform with far worse security and isolation, and then ship it in a way that's not set up in a proper way for production and is a massive step backwards from the mainstream status quo? I don't get it.
1
Replying to
I don't know enough about the isolation options of the SoC here yet so I can't speak to that. But as I stated three times, I agree with you. We need to either be able to have SoC features to distrust the radios, or have replaceable radios to swap to trustworthy ones later.
1
Replying to
The SoC has a ton of components itself, not just the massive complexity of the CPUs and GPU. It's the core of the security for a phone, and you can't just ship an over-the-air SoC update after the fact. Firmware / microcode allows flaws to be fixed with updates, hardware doesn't.
Replying to and
Hardware and firmware security really does matter a lot. For example, the GPU security model / implementation. A lot of GPU vendors don't do much to stop trivial exploitation leading to not just gaining access to all graphics but more powerful memory access than even the kernel.