Re: Request for Comment: LED device naming for netdev LEDs

[Marek Behun <marek.behun@nic.cz>]

On Mon, 28 Sep 2020 19:10:05 +0200
Alexander Dahl <post@lespocky.de> wrote:

> Hei hei,
> 
> On Mon, Sep 28, 2020 at 05:52:09PM +0200, Marek Behun wrote:
> > On Mon, 28 Sep 2020 15:04:10 +0200
> > Alexander Dahl <ada@thorsis.com> wrote:
> >   
> > > Hei Marek,
> > > 
> > > Am Sonntag, 27. September 2020, 02:52:58 CEST schrieb Marek Behun:  
> > > > On Sun, 27 Sep 2020 00:40:25 +0200
> > > > 
> > > > Marek Behun <marek.behun@nic.cz> wrote:    
> > > > > What I am wondering is how should we select a name for the device part
> > > > > of the LED for network devices, when network namespaces are enabled.
> > > > > 
> > > > > a) We could just use the interface name (eth0:yellow:activity). The
> > > > > 
> > > > >    problem is what should happen when the interface is renamed, or
> > > > >    moved to another network namespace.
> > > > >    Pavel doesn't want to complicate the LED subsystem with LED device
> > > > >    renaming, nor, I think, with namespace mechanism. I, for my part, am
> > > > >    not opposed to LED renaming, but do not know what should happen when
> > > > >    the interface is moved to another namespace.
> > > > > 
> > > > > b) We could use the device name, as in struct device *. But these names
> > > > > 
> > > > >    are often too long and may contain characters that we do not want in
> > > > >    LED name (':', or '/', for example).
> > > > > 
> > > > > c) We could create a new naming mechanism, something like
> > > > > 
> > > > >    device_pretty_name(dev), which some classes may implement somehow.
> > > > > 
> > > > > What are your ideas about this problem?
> > > > > 
> > > > > Marek    
> > > > 
> > > > BTW option b) and c) can be usable if we create a new utility, ledtool,
> > > > to report infromation about LEDs and configure LEDs.
> > > > 
> > > > In that case it does not matter if the LED is named
> > > >   ethernet-adapter0:red:activity
> > > > or
> > > >   ethernet-phy0:red:activity
> > > > because this new ledtool utility could just look deeper into sysfs to
> > > > find out that the LED corresponds to eth0, whatever it name is.    
> > > 
> > > I like the idea to have such a tool.  What do you have in mind?  Sounds for me 
> > > like it would be somehow similar to libgpiod with gpio* for GPIO devices or 
> > > like libevdev for input devices or like mtd-utils …
> > >   
> > As Pavel said, we have ethtool, maybe we could have ledtool.  
> 
> Yes.  IIRC ethtool uses libmnl for communicating with the kernel
> through the netlink interface.
> 
> > > Especially a userspace library could be helpful to avoid reinventing the wheel 
> > > on userspace developer side?  
> > 
> > If such a need arises, than yes. For most embedded systems though I
> > think ledtool would be enough, since mostly LEDs can be controlled from
> > shell scripts.  
> 
> I saw proprietary embedded C++ applications building on top of
> proprietary C libraries interacting with the sysfs leds interface …
> O:-)
> 

:-O okaaaay...

> > > Does anyone else know prior work for linux leds sysfs interface from 
> > > userspace?  
> > 
> > I am not aware of that, everyone just uses sysfs now.  
> 
> Sorry, that was maybe misleading.  What I wanted to know was if there
> already is some free library/tool using that sysfs interface?  I
> suppose not, otherwise "ledtool" would not be needed?  IIRC there are
> generic libs for abstracting sysfs access, but I did not like them.
> ;-)
> 
> Long story short, I would be interested in helping on a ledtool /
> libledtool in C in my spare time.  (No time to learn Rust at the
> moment though.)

Well, we can always work on something independent of Pavel, but I
don't think that would be wise...

> Greets
> Alex
>