I think it might be reasonable to include the bus number and address (I presume that's what you mean by 003/038, tho I'm not certain).
Hrm... is the 003/038 topology information?
If someone can figure out where it comes from, I have no objection to
adding it to the /proc file for usb-storage.
Matt
On Fri, Oct 11, 2002 at 11:11:09AM +1300, Michael Hamilton wrote:
> Looks like I'll give up on the serial number being usefull. As
> mentioned in another recent post I now treat devices generically.
>
> I would like a simple way to take the DEVICE reported to
> /etc/hotplug/usb/usb-storage (eg /proc/bus/usb/003/038) and figure out
> which scsi device it maps to.
>
> The /proc/scsi/usb-storage-n/m only details the GUID. The GUID
> depends on the serialnumber, so that's no good. Couldn't the two be
> linked by something definitly unique, perhaps the 003/038 could appear
> in /proc/scsi/usb-storage-n/m; or maybe there could be counter; or a
> timestamp. At the moment, because I can't be sure which device is
> the one the user just plugged in, I have to resort to mounting
> all attached devices on the bus - which may include one I've just
> umounted but not unplugged.
>
> On Tue, 08 Oct 2002 08:24, Matthew Dharm wrote:
> > On Mon, Oct 07, 2002 at 11:26:35AM -0700, Randy.Dunlap wrote:
> > > On Mon, 7 Oct 2002, Matthew Dharm wrote:
> > > | On Tue, Oct 08, 2002 at 01:11:01AM +1300, Michael Hamilton wrote:
> > > | > 1. What part of the serial remains constant for the same device,
> > > | > if any?
> > > |
> > > | The whole thing should. The compliance test that the USB-IF (including
> > > | myself) are working on requires a constant serial number.
> > >
> > > Matt, do you read that as part of any spec (requirement)?
> >
> > Not exactly. But the entire Mass Storage DWG was pretty supprised when I
> > pointed out devices that had this broken behavior.
> >
> > The spec requires a 'unique' serial number, but does not specify what the
> > domain for that uniqueness is. The intent was to be unique over all
> > devices with that VID:PID. A few (a _very_ few) manufacturers chose to
> > interpret this as unique for a given USB bus.
> >
> > The changing serial number is because the bus address is incorporated into
> > the serial number.
> >
> > Of course, that means that it's possible to have two units on two different
> > USB busses with the same bus address, and thus the same serial number.
> >
> > Matt
--
Matthew Dharm Home: [EMAIL PROTECTED]
Maintainer, Linux USB Mass Storage Driver
C: They kicked your ass, didn't they?
S: They were cheating!
-- The Chief and Stef
User Friendly, 11/19/1997
msg05020/pgp00000.pgp
Description: PGP signature
