<qyliss>
Yeah, cloud-hypervisor is definitely using multiple memfds for guest memory
<qyliss>
Another thing I was just thinking was that maybe it would be good if the main part of guest memory wasn't fd-backed at all (and just an anonymous mmap), because that way you couldn't even accidentally share it with another process
<qyliss>
(the rust-vmm vm-memory crate, which cloud-hypervisor (and maybe also Firecracker) uses, does this by default, but I think cloud-hypervisor always passes it memfds anyway)
nicoo has joined #spectrum
ehmry has joined #spectrum
jb55 has quit [Ping timeout: 240 seconds]
jb55 has joined #spectrum
qyliss has quit [Quit: bye]
qyliss has joined #spectrum
jb55 has quit [Ping timeout: 240 seconds]
jb55 has joined #spectrum
nicoo has quit [Read error: Connection reset by peer]