Re: [PATCH 6/8] staging/mei: Header file contain the Userland API, (IOCTL and its struct)

[Alan Cox <alan@linux.intel.com>]

> Was this discussed in public?

Not much yet - but this is a staging driver so there is no defined ABI
and nothing is set in stone.

> > if someone has a good solution for that we are happy to hear it.
> 
> Did you rule out sysfs, configfs, netlink, etc.?

Pretty much. You need to create a handle, connect it to something and
perform transactions on it. Some of those must be serialized and they
have both input and output data, so they don't fit sysfs and friends.

It's effectively mostly a message passing system - the closest other
interface is the socket one but it doesn't really fit that and I don't
think Dave wants 'its got a connect sort of op so lets make it a socket
family' to be the device norm !

The identifier for the object is also long so doesn't work for a set of
devices by minor number either.

I think the more useful step is going to be some example user space as
that above all will show how it is intended to work

Alan