[PATCH] of/address: Don't throw errors on absent ranges properties

[PATCH] of/address: Don't throw errors on absent ranges properties

[Benjamin Herrenschmidt]

The core always tries to translate any "reg" property to construct the platform
device names. This results in a pile of "OF: no ranges; cannot translate" errors
in dmesg whenever we expose things like i2c devices that cannot directly translate
to the MMIO space.

Turn this into a pr_debug instead

Signed-off-by: Benjamin Herrenschmidt <benh-XVmvHMARGAS8U2dJNN8I7kB+6BGkLq7r@public.gmane.org>

diff --git a/drivers/of/address.c b/drivers/of/address.c
index f0541fd..bf1f79d 100644
--- a/drivers/of/address.c
+++ b/drivers/of/address.c
@@ -444,7 +444,7 @@ static int of_translate_one(struct device_node *parent, struct of_bus *bus,
 	 */
 	ranges = of_get_property(parent, rprop, &rlen);
 	if (ranges == NULL && !of_empty_ranges_quirk()) {
-		pr_err("OF: no ranges; cannot translate\n");
+		pr_debug("OF: no ranges; cannot translate\n");
 		return 1;
 	}
 	if (ranges == NULL || rlen == 0) {


--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] of/address: Don't throw errors on absent ranges properties

[Grant Likely]

On Fri, 14 Nov 2014 17:58:23 +1100
, Benjamin Herrenschmidt <benh@kernel.crashing.org>
 wrote:
> The core always tries to translate any "reg" property to construct the platform
> device names. This results in a pile of "OF: no ranges; cannot translate" errors
> in dmesg whenever we expose things like i2c devices that cannot directly translate
> to the MMIO space.

I don't have a problem with the change, but it seems to be catching an
odd usage of of_device_make_bus_id(). Why is of_device_make_bus_id()
being called on i2c devices? Those shouldn't be modelled as
platform_devices.

g.

> 
> Turn this into a pr_debug instead
> 
> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> 
> diff --git a/drivers/of/address.c b/drivers/of/address.c
> index f0541fd..bf1f79d 100644
> --- a/drivers/of/address.c
> +++ b/drivers/of/address.c
> @@ -444,7 +444,7 @@ static int of_translate_one(struct device_node *parent, struct of_bus *bus,
>  	 */
>  	ranges = of_get_property(parent, rprop, &rlen);
>  	if (ranges == NULL && !of_empty_ranges_quirk()) {
> -		pr_err("OF: no ranges; cannot translate\n");
> +		pr_debug("OF: no ranges; cannot translate\n");
>  		return 1;
>  	}
>  	if (ranges == NULL || rlen == 0) {
> 
> 

Re: [PATCH] of/address: Don't throw errors on absent ranges properties

[Benjamin Herrenschmidt]

On Tue, 2014-11-18 at 16:57 +0000, Grant Likely wrote:
> On Fri, 14 Nov 2014 17:58:23 +1100
> , Benjamin Herrenschmidt <benh-XVmvHMARGAS8U2dJNN8I7kB+6BGkLq7r@public.gmane.org>
>  wrote:
> > The core always tries to translate any "reg" property to construct the platform
> > device names. This results in a pile of "OF: no ranges; cannot translate" errors
> > in dmesg whenever we expose things like i2c devices that cannot directly translate
> > to the MMIO space.
> 
> I don't have a problem with the change, but it seems to be catching an
> odd usage of of_device_make_bus_id(). Why is of_device_make_bus_id()
> being called on i2c devices? Those shouldn't be modelled as
> platform_devices.

Sorry, this was my explanation being full of crap. It's my i2c
_controller_ which is a platform device, and is on the xscom bus which
isn't directly MMIO translatable.

The patch still stands :)

Cheers,
Ben.

> g.
> 
> > 
> > Turn this into a pr_debug instead
> > 
> > Signed-off-by: Benjamin Herrenschmidt <benh-XVmvHMARGAS8U2dJNN8I7kB+6BGkLq7r@public.gmane.org>
> > 
> > diff --git a/drivers/of/address.c b/drivers/of/address.c
> > index f0541fd..bf1f79d 100644
> > --- a/drivers/of/address.c
> > +++ b/drivers/of/address.c
> > @@ -444,7 +444,7 @@ static int of_translate_one(struct device_node *parent, struct of_bus *bus,
> >  	 */
> >  	ranges = of_get_property(parent, rprop, &rlen);
> >  	if (ranges == NULL && !of_empty_ranges_quirk()) {
> > -		pr_err("OF: no ranges; cannot translate\n");
> > +		pr_debug("OF: no ranges; cannot translate\n");
> >  		return 1;
> >  	}
> >  	if (ranges == NULL || rlen == 0) {
> > 
> > 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/


--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] of/address: Don't throw errors on absent ranges properties

[Grant Likely]

On Wed, 19 Nov 2014 13:32:31 +1100
, Benjamin Herrenschmidt <benh@kernel.crashing.org>
 wrote:
> On Tue, 2014-11-18 at 16:57 +0000, Grant Likely wrote:
> > On Fri, 14 Nov 2014 17:58:23 +1100
> > , Benjamin Herrenschmidt <benh@kernel.crashing.org>
> >  wrote:
> > > The core always tries to translate any "reg" property to construct the platform
> > > device names. This results in a pile of "OF: no ranges; cannot translate" errors
> > > in dmesg whenever we expose things like i2c devices that cannot directly translate
> > > to the MMIO space.
> > 
> > I don't have a problem with the change, but it seems to be catching an
> > odd usage of of_device_make_bus_id(). Why is of_device_make_bus_id()
> > being called on i2c devices? Those shouldn't be modelled as
> > platform_devices.
> 
> Sorry, this was my explanation being full of crap. It's my i2c
> _controller_ which is a platform device, and is on the xscom bus which
> isn't directly MMIO translatable.

Okay, that makes more sense, but if xscom is a different bus with
different access methods, then why is it using platform_bus_type? I
would expect it to have it's own bus_type and container for struct
device.

> The patch still stands :)

Indeed, I merged it yesterday. :)

> 
> Cheers,
> Ben.
> 
> > g.
> > 
> > > 
> > > Turn this into a pr_debug instead
> > > 
> > > Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> > > 
> > > diff --git a/drivers/of/address.c b/drivers/of/address.c
> > > index f0541fd..bf1f79d 100644
> > > --- a/drivers/of/address.c
> > > +++ b/drivers/of/address.c
> > > @@ -444,7 +444,7 @@ static int of_translate_one(struct device_node *parent, struct of_bus *bus,
> > >  	 */
> > >  	ranges = of_get_property(parent, rprop, &rlen);
> > >  	if (ranges == NULL && !of_empty_ranges_quirk()) {
> > > -		pr_err("OF: no ranges; cannot translate\n");
> > > +		pr_debug("OF: no ranges; cannot translate\n");
> > >  		return 1;
> > >  	}
> > >  	if (ranges == NULL || rlen == 0) {
> > > 
> > > 
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
> 
> 

Re: [PATCH] of/address: Don't throw errors on absent ranges properties

[Benjamin Herrenschmidt]

On Wed, 2014-11-19 at 10:31 +0000, Grant Likely wrote:
> was my explanation being full of crap. It's my i2c
> > _controller_ which is a platform device, and is on the xscom bus
> which
> > isn't directly MMIO translatable.
> 
> Okay, that makes more sense, but if xscom is a different bus with
> different access methods, then why is it using platform_bus_type? I
> would expect it to have it's own bus_type and container for struct
> device.

There is no point really. It's not really a bus we expose as such to
Linux (mostly the FW uses it) though we use the xscom nodes as chip
nodes.

We do expose selected devices we pick from underneath xscom such
as the LPC bus and the i2c controllers as platform devices which then
use FW interfaces to perform the actual LPC or i2c accesses.

Creating a dedicated bus type would be completely pointless.

The FW itself uses what's there more intensively.

> > The patch still stands :)
> 
> Indeed, I merged it yesterday. :)

Thanks!

Cheers,
Ben.


--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html