> Thanks for the info, but the devices encumbered with ioctls are the tricky > ones and even if they can be sorted out I'm sure there are some other traps > out there. Too bad there are no RFS gurus lurking here to offer their wisdom > on remoting devices.
I'm no RFS guru, thank deity, but I did RTFC once and "F" was apposite. ioctl was handled by having the client "know" exactly what each ioctl "looked like", i.e. it only worked for known cases. There was a humungous switch which knew the "right stuff" to copy in&out, how to byte swap and align any words, structs etc. After deciding what the "right stuff" was, it copied in the "right stuff" from user space, passed over the call to the server, the server then executed it and passed back the "right stuff" to the client, the client then copied out the "right stuff", then I started losing the will to live. Dave.
