Hi, On (21/02/12 15:38), Stefan Metzmacher via Linux-cifsd-devel wrote: > I haven't found the exact place, but ksmbd.mountd starts the kernel-part. > > ksmbd.mountd also acts as some kind of upcall, for the server part, > that takes care of authentication and some basic DCERPC calls. > > I'm wondering why there are two separate ways to kill the running server, > 'killall ksmbd.mountd' for the userspace part and > 'ksmbd.control -s' (which is just a wrapper for > 'echo -n "hard" > /sys/class/ksmbd-control/kill_server') to shutdown the server part. > > As it's not useful to run any of these two components on its own, > so I'm wondering why there's no stronger relationship. > > As naive admin I'd assume that the kernel part would detect the exit of ksmbd.mountd > and shutdown itself. User-space daemon is just some sort of a database engine, kernel module queries it when it needs something; otherwise kernel module works without user-space part just fine and doesn't need it. The goal is that when user-space crashes or gets restarted after update (critical fix) the server doesn't panic and doesn't kill itself. E.g. - when you restart mysql you don't expect httpd to kill itself and to terminate all existing TCP connections. [..] > Can we require that the userspace tool matches the kernel version for a while? I think such a check exists (or at least it used to). Note that the only time when version mismatch matters is when sizeof() or layout of the structures that are used for RPC get updated, or new request commands or status codes added. IOW, RPC version mismatch is the critical thing. Otherwise there is no real reason (case by case) to forbid version mismatch. If user-space gets a memory leak or a NULL pointer dereference fix then there is no real reasons to force server restart or to force server module version update, because the fix is completely internal to the user-space daemon. -ss
