USB and DevFS (was Re: DEVFSv50 etc etc etc...)

Shawn Leas (sleas@ixion.honeywell.com)
Sun, 16 Aug 1998 19:33:20 -0500 (CDT)


On Sun, 16 Aug 1998, Theodore Y. Ts'o wrote:

> Date: Sun, 16 Aug 1998 21:32:03 +0200
> From: Inaky Perez Gonzalez <inaky@peloncho.fis.ucm.es>
>
> I'm considering that, but it looks like you'll have to have all
> your device nodes statically made [like nowadays] and don't play _at
> all_ with the plug order into the hub port's once you've got your
> system set up; in other words: crappy, it breaks USB's spirit.
>
> Nope; that's not the only way to do it. Instead, you could have a
> user-mode daemon (just like PCMCIA does) which dynamically creates the
> /dev files when each USB device is plugged in.
>
> I've suggested doing this for all devices nodes as an alternative to
> devfs, but the devfs folks seem to have ignored this proposal.

The devfsd could do just that, and because there has been an issue of
persistance brought to the forefront, this devfsd could have a standard
configuration file for permissions, ownership, etc.

This is a + for devfs, because its dynamic nature fits more with
USB/firewire concept in general, and the persistance of any USB device
node could depend on some sort of /etc/conf.devfs.

The same situation with a static /dev layout could also work with devfsd
doing the grunt work if I'm not mistaken...

It is a little irresponsible to say "the devfs folks seem to have ignored
this proposal", though. I thought this flameware ended. Leave it to
Theodore to bring it back.

-Shawn
<=========== America Held Hostage ===========>
Day 2034 for the poor and the middle class.
Day 2053 for the rich and the dead.
888 days remaining in the Raw Deal.
<============================================>

-
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