Hacker News Reader

Comment by moonbug | Hacker News Reader

Can't see why this needs to be in the kernel

It doesn't need to be. Samba works from userspace. This performs better because the fs primitives are in-kernel. It's also reportedly easier to maintain and easier to further optimize as part of the kernel. Everything's a tradeoff.

Same reason knfsd is in the kernel: performant file + locking access synchronized with the fs primitives.

The post indicates this is not to replace samba, but to improve the file sharing bits of it.

Isn't the main reason for knfsd being in-kernel that it depends on inode number to file lookups? Does samba have the same problem?

The kernel nfs server is IME more robust than userspace options like ganesha, but that seems to mostly be a function of it being widely used - a kernel samba server doesn't have that advantage?

No, Samba does not need to do inode to file lookups. There are no "open by inode" primitives in SMB1/2/3.

Dumb question: if this (or any other) kernel module crashes does the entire kernel?

Yup! And if an attacker hacks this module they can hijack the kernel. And it can probably be reached by anyone on the same network, or may even be exposed to the internet :|

You mean like the TCP/IP stack?

Good point, but SMB has earned quite a reputation in this area.

And if I could move that into user space without a hideous performance and/or compatibility penalty, I probably would.

> The target is to provide optimized performance, GPLv2 SMB server

High performance doesn't have to mean being in the kernel.

Reply on news.ycombinator.com