Re: [PATCH] USB speedtouch: better proc info

[Oliver Neukum <oliver@neukum.name>]

> Hi Oliver, thanks for your comments.  While I agree with you in
> principle, I disagree in practice.  The driver exports the following
> information in /proc/net/atm/speedtch:
>
> (1) name and location of the USB device
> (2) MAC address (serial number)
> (3) AAL5 transmission statistics
> (4) Line status
> (5) Modem status
>
> (1) is needed in order to work out which modem corresponds to
> which ATM device.  This should be dealt with using sysfs, however
> the ATM layer has not yet been ported to sysfs.  Until it is, this
> seems like the best way to export this information.

For the time being I agree.

> (2) and (3) are redundant - they are published by the ATM layer
> in other proc files.  I thought about removing them, but decided
> against it because (a) it can be convenient having everything in
> one proc file, and (b) it is backwards compatible with the 2.4
> out-of-kernel driver.  They could go.

As long as you have a proc file at all, you may as well print them, IMHO.

> You suggested (in a private mail) using netif_carrier_on/off to
> export (4).  The ATM layer already has a method for reporting this,
> and I use it: set the ATM_PHY_SIG_FOUND/LOST bits in
> atm_dev->signal.  The problem is that the ATM layer doesn't do
> anything with this info (like export it to user space).  So I think
> it is fair enough to export it in the proc file while waiting for the
> ATM layer to be fixed.

Yes, but I think that you should notify the network layer too.
I see no reason any network driver shouldn't report this directly.
There's nothing specific to ATM in losing signal.
It's purely physical thing low level drivers should deal with.

> As for (5), this could be exported using sysfs.  Since it is a
> USB matter, I guess I could do this now.  So this could also go.

Yes.

	Regards
		Oliver