On Mon, 18 Sept 2023 at 16:40, Christian Brauner <brauner@xxxxxxxxxx> wrote: > What we're talking about here is a nicely typed struct which returns two > paths @mnt_root and @mnt_point which can both be represented as u64 > pointers with length parameters like we do in other binary structs such > as bpf and clone3 and a few others. That is a compromise I can live > with. I'm really trying to find as much common ground here as we can. So to be clear about your proposal: .mnt_root and .mountpoint are initialized by the caller to buffers that the kernel can copy paths into? If there's an overflow (one of the buffers was too small) the syscall returns -EOVERFLOW? Thanks, Miklos
