I have experience writing kernel modules, but I have probably only encountered 5% of the kernel’s systems. I have done no work with VFS before.

I have an idea for a new filesystem. If I proved the new filesystem by implementing it using the fuse_lowlevel API, how difficult would it be to port into the kernel?

Is it too much to hope that there’s a kernel-side shim of the FUSE API, so that I could just drop the code in?

If not, how closely does the fuse_lowlevel API resemble what I need to implement in the kernel? At first glance, some of the important ones (e.g. listing directories) look to have very different semantics.