Hector Martin

I pretty much entirely agree but I also thinkthinl it's useful having constant reminder that Thunderbolt and every other inane "external connection DMA bus" is Intel & friends being reckless with their customers' safety.

Like, there shouldn't be a key that gives you a backdoor to a laptop to begin with, so even if the method by which attacker obtains it is no big surprise, the existence of the key already is.

The key is in *your* already trusted device. It's not some kind of global master key. You need to break into an already trusted device. The right way to stop DMA attacks is with an IOMMU, which is complementary to all this. Proper IOMMU usage blocks all this crap anyway.

IOMMU, or just not having bus mastering. But yeah. Problem is none of their IOMMUs work. They don't fail closed.

Why is that? I haven't looked at how one actually programs an IOMMU. Is it a fundamental limitation or some sort of "we can't break existing devices relying on being able to read/write anywhere"?

You don't have to tell me about drivers fucking up IOMMU mappings. That's how we owned the PS4. The solution isn't forcing bounce buffers on everything, it's better software frameworks for getting this right without just dumbly copying data.

If final destination is userspace you can't. Other data may be in same page. Based on the LWN summary I posted it looks like they're doing bounce buffer mappings only for first/last partial pages...