Skip to content

Instantly share code, notes, and snippets.

AGX coherency, caching, and TLBs

These are just some notes on my current understanding of the subtleties of the AGX memory model and the TLB/caching issues I'm seeing.

Hypervisor shenanigans

TLBI instructions do not broadcast to the GPU from EL1 with stage 2 translation enabled. That's it. That's what the bug was.

GPU side