Re: [PATCH 06/10] firmware: arm_scmi: Add System Telemetry driver

[Jonathan Cameron <jonathan.cameron@huawei.com>]

On Tue, 21 Oct 2025 17:03:36 +0100
Cristian Marussi <cristian.marussi@arm.com> wrote:

> On Tue, Oct 21, 2025 at 04:15:29PM +0100, Jonathan Cameron wrote:
> > On Tue, 21 Oct 2025 11:27:02 +0100
> > Cristian Marussi <cristian.marussi@arm.com> wrote:
> >   
> > > On Mon, Oct 20, 2025 at 05:23:28PM +0100, Jonathan Cameron wrote:  
> > > > On Thu, 25 Sep 2025 21:35:50 +0100
> > > > Cristian Marussi <cristian.marussi@arm.com> wrote:
> > > >     
> > > > > Add a new SCMI System Telemetry driver which gathers platform Telemetry
> > > > > data through the new the SCMI Telemetry protocol and expose all of the
> > > > > discovered Telemetry data events on a dedicated pseudo-filesystem that
> > > > > can be used to interactively configure SCMI Telemetry and access its
> > > > > provided data.    
> > > >    
> > > 
> > > Hi,
> > >    
> > > > I'm not a fan of providing yet another filesystem but you didn't  
> > 
> > "did" was what this was meant to say.
> > 
> > Sorry for the confusing garbage comment from me!
> >   
> > > > lay out reasoning in the cover letter.    
> > > 
> > > Sorry, I dont understand..you mean here that I did NOT provide enough reasons
> > > why I am adopting a new FS approach ? ... or I misunderstood the English ?
> > > 
> > > .. because I did provide a lot of reasons (for my point-of-view) to go
> > > for a new FS in the cover-letter...
> > >   
> > > > 
> > > > One non trivial issue is that you'll have to get filesystem review on this.
> > > > My review is rather superficial but a few things stood out.    
> > > 
> > > Well yes I would have expected that, but now the FS implementation
> > > internals of this series is definetely immature and to be reworked (to
> > > the extent of using a well-know deprecated FS mount api at first..)
> > > 
> > > So I posted this V1 to lay-out the ideas and the effective FS API layout
> > > but I was planning to extend the review audience once I have reworked fully
> > > the series FS bits in the next V2...  
> > 
> > I'd suggest ABI docs for v2. That will match what you have in the cover letter
> > but put it in the somewhat formal description format of Documentation/ABI/
> >   
> 
> Oh yes of course... the while docs/ stuff is still TBD...btw I am not even
> sure if the whole driver will be required to be moved into fs/ as a
> requirement while doing filesystem review...I suppose I will leave this
> sort of reworks for the next reviews cycles....
> 
> ...and...if I may ask... is it linux-fsdevel the ML for this fs-related
> stuff I suppose...not sure about maintainers looking at MAINTAINERS ...

Seems resonable but beyond that I have no idea.

Give it a go and see what happens.  Probably also include kernfs related folk
directly. They are likely to have opinions and might review if they have time.


Jonathan
> 
> Thanks a lot for having a look Jonathan.
> Cristian
>