[STLinux Kernel] [PATCH v2 2/4] remoteproc: dt: Provide bindings for ST's Remote Processor Controller driver

[Lee Jones]

On Tue, 01 Sep 2015, Peter Griffin wrote:

> Hi Lee,
> 
> On Fri, 28 Aug 2015, Lee Jones wrote:
> 
> > Signed-off-by: Ludovic Barre <ludovic.barre at st.com>
> > Signed-off-by: Lee Jones <lee.jones at linaro.org>
> > ---
> >  .../devicetree/bindings/remoteproc/st-rproc.txt    | 35 ++++++++++++++++++++++
> >  1 file changed, 35 insertions(+)
> >  create mode 100644 Documentation/devicetree/bindings/remoteproc/st-rproc.txt
> 
> The patch documening the DT bindings should be ordered before the patch which adds
> the DT node to aid reviewing.
> > 
> > diff --git a/Documentation/devicetree/bindings/remoteproc/st-rproc.txt b/Documentation/devicetree/bindings/remoteproc/st-rproc.txt
> > new file mode 100644
> > index 0000000..fbd7d78
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/remoteproc/st-rproc.txt
> > @@ -0,0 +1,35 @@
> > +STMicroelectronics Remote Processor
> > +-----------------------------------
> > +
> > +This binding provides support for adjunct processors found on ST SoCs.
> > +
> > +The remote processors can be controlled from the bootloader or the primary OS.
> > +If the bootloader starts a remote processor processor the primary OS must detect
> > +its state and act accordingly.
> > +
> > +Required properties:
> > +- compatible		Should be one of:
> > +				"st,st231-rproc"
> > +				"st,st40-rproc"
> 
> st40-proc isn't used anywhere. The stih407 doesn't have a ST40 copro, and
> looking in the vendor tree remoteproc support isn't present for stih415/6 which
> are the the only upstream SoC's to have a ST40 co-pro.
> 
> So I think st40-rproc support can be removed.
> 
> > +- reg			Size and length of reserved co-processor memory
> > +- resets		Reset lines (See: ../reset/reset.txt)
> > +- reset-names		Must be "sw_reset" and "pwr_reset"
> 
> pwr_reset isn't used by any of the st231 co-processors. It seems to
> be related to ST40 support which I don't think is required upstream.
> Removing it would make the driver a fair bit smaller.
> 
> > +- clocks		Clock for co-processor (See: ../clock/clock-bindings.txt)
> > +- clock-names		Must be "rproc_clk"
> 
> I can't see any co-pro which uses more than one clock, so clock-names looks
> superflous.
> 
> > +- clock-frequency	Clock frequency to set co-processor at if the bootloader
> > +			hasn't already done so
> > +- st,syscfg-boot	The register that holds the boot vector for the co-processor
> 
> I would prefer to see this binding match how most other sti drivers reference syscfg
> registers which is: -
> 
> st,syscfg       = <&syscfg_core 0xf4>;
> 
> Description: phandle of sysconfig bank plus integer array containing register offsets.
> 
> It also means it is easily extendable if more than one syscfg register is
> required in the future to boot a co-pro.

Ack.  Good points, will fix.

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog