[at76c503a-develop] [PATCH 2/6] Remove last remains of LED support, it shouldn't be here

[Guido Guenther]

Hi Pavel,
On Sat, Apr 28, 2007 at 04:54:07PM -0400, Pavel Roskin wrote:
> Quoting Erik Hovland <erik at hovland.org>:
> 
> > If you could keep LED support, it would be helpful. We are carrying a
> > patch in our tree that uses LEDs. I will see what I can do about getting
> > an updated patch to this list for those LEDs.
> 
> OK, let's keep the LED code for now.  It was just a cleanup based on nm output. 
> We have bigger issues.
>
> But I'm sure we'll be asked about the LED code if the driver is submitted again
> for the kernel.  Something needs to be done about it.
I think Johannes suggested using a LED trigger, that's why I added the
code.

> What's really strange is that the LED code only indicates whether the driver is
> loaded.  It's completely separate from any hardware.  If the hardware is
> disconnected, the LED is still registered.
The trigger is registered during module load time and unregistered on
module removal, it triggers the led (if bound to any via sysfs) on
package receive - looks o.k. to me.
Cheers,
 -- Guido

> We can just as well have a separate module with the LED code that would indicate
> its own presence.  But that's a sign that the LED code is useless as it is now.
> 
> --
> Regards,
> Pavel Roskin
> _______________________________________________
> at76c503a-develop mailing list
> at76c503a-develop at lists.berlios.de
> https://lists.berlios.de/mailman/listinfo/at76c503a-develop