Re: [STLinux Kernel] [PATCH 06/12] watchdog: bindings: Supply knowledge of a third supported device - clocksource

[Peter Griffin]

Hi Lee,

<snip>

On Fri, 08 May 2015, Lee Jones wrote:
> > > > >  
> > > > > -- compatible 	: Must be one of: "st,stih407-lpc" "st,stih416-lpc"
> > > > > -				  "st,stih415-lpc" "st,stid127-lpc"
> > > > > +- compatible 	: Must be one of: "st,stih407-lpc"
> > > > 
> > > > The same comment as the RTC DT patch, you are removing the compatibles
> > > > documentation for the other supported platforms like stih416-lpc.
> > > > AFAIK they are required in the driver to get the correct sysconfig register.
> > > 
> > > That's intentional.  I haven't yet tested any of this IP on
> > > STiH41{5,6} & STiH127.  Due to lack of documentation, I'm not even
> > > sure if this IP even exists on some of the other platforms.  I will
> > > add them back when support is added to both driver and DTB and I've
> > > been able to test them.
> > 
> > That was kind of my point, the driver code AFAIK already contains support
> > for these SoC's.
> > 
> > I would either expect the patch to remove support from the DT docs AND the
> > driver, or leave it as is.
> > 
> > It seems odd to only change the DT docs, and become unaligned to the
> > code (this assumes I'm looking at the latest patchset here 
> > https://lkml.org/lkml/2015/3/4/1088 which includes support for these SoCs).
> 
> The decision to remove these 'supported' platforms was made on the RTC
> side, where there is only support for "st,stih407-lpc" in the driver.

The change makes sense in the RTC case then.

> I thought it best to mirror that thought over to the Watchdog LPC
> bindings,

Why? This is the bit I disagree with, the wdt DT docs IMHO should match the wdt
driver code.

> but thought it not really worth ripping out existing support
> from the driver.  

I agree it is not worth ripping out existing support only to add it back in. Equally
I can't see the point in doing this with the DT documentation either.

> If others wish to test/use the Watchdog on other
> platforms and can read C code, they'll know what to do.

Yes, and without this change they could also read the documentation. Your doing a
v2 anyway, it seems trivial to fix this at the same time.

regards,

Peter.
--
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/