Re: [PATCH can-next v6] can: add tx/rx LED trigger support

[Fabio Baltieri]

On Tue, Sep 04, 2012 at 09:11:28AM +0200, Kurt Van Dijck wrote:
> On Mon, Sep 03, 2012 at 10:54:49PM +0200, Oliver Hartkopp wrote:
> > On 03.09.2012 20:29, Fabio Baltieri wrote:
> > 
> > > On Mon, Sep 03, 2012 at 08:13:35PM +0200, Kurt Van Dijck wrote:
> > >> On Mon, Sep 03, 2012 at 02:40:39PM +0200, Marc Kleine-Budde wrote:
> > >>> The net->ifindex is unique. But it's only an integer. Usually can0 has a
> > >>> ifindex != 0, so a simple can%d is contra productive here.
> > >>>
> > >>> Some pointers to related code:
> > >>> http://lxr.free-electrons.com/source/drivers/base/core.c#L1847
> > >>> http://lxr.free-electrons.com/source/drivers/base/core.c#L73
> > >>> http://lxr.free-electrons.com/source/include/linux/device.h#L695
> > >>>
> > >>> comments?
> > > 
> > > That would probabily makes really hard to choose the right
> > > default_trigger for led devices to get to the appropriate CAN LED in
> > > embedded systems, as trigger name would depend from other network
> > > devices and probing order (correct me if I'm wrong).
> > > 
> > > Something with device name would probaily be more appropriate here.
> > > 
> > >>
> > >> a very recent idea: something with netdevice notifiers and NETDEV_CHANGENAME ...
> > >> http://lxr.free-electrons.com/source/net/core/dev.c#L1030
> > >>
> > >> you could: rename the trigger, or if we think it's usefull,
> > >> block the netdev rename when its triggers are in use.
> > > 
> > > Blocking the rename looks overkill to me,
> 
> renaming a netdev _after_ first attaching led triggers looks stupid to me
> anyway.
> 
> > > what about using device name
> > > with an optional "port id" appended to it?  Sounds simpler...
> > 
> > 
> > The name of the device can only be changed when the interface is down.
> > Is it possible to put some scripting around it to detach and attach the leds
> > to the interfaces on ifup/ifdown triggers?
> 
> Are the led triggers available for using while the netdev is down then?

Sure!  On embedded systems triggers are usually attached to actual LEDs
at probe time using default_trigger field of struct led_classdev, and
that can be specified both in machine files or in device tree.

See

http://lxr.free-electrons.com/source/arch/powerpc/boot/dts/mpc8315erdb.dts#L477
http://lxr.free-electrons.com/source/arch/arm/mach-pxa/tosa.c#L576

that's why the trigger name should be predictable, at least at
boot/probe time.

The actual CAN LED trigger also reflects the actual state of the
interface (off if down, on if up, blink on activity).

Fabio
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/