>In other words, just push all device mgmt whatsoever into userspace is to
>me broken, and I believe that if the original Unix had such an inerface an
>idea like DevFS would have popped up a long time ago.
>
>If we can avoid a hoaky HP-UX-ish solution to device creation (which seems
>to be the userland one), I like it. I don't see that it breaks some
>functionality, but if it does, you've tested it, and it's fundamentally
>impossible to do with DevFS, there are even other ways to do DevFS.
I'd imagine that a lot of the complaints about devfs would go away
if it had inherited-mount semantics; if you mounted it over an
existing /dev, it would pick up the permissions and ownerships of
the underlying filesystem, and *poof* any horrible kludgery you'd
have to do with setting permissions and picking apart tarballs would
go away just like that.
____
david parsons \bi/ But, alas, I don't have room in this margin to
\/ code this solution.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html