[PATCH v3] clk: shmobile: Add R8A7740-specific clock support
Geert Uytterhoeven
geert at linux-m68k.org
Thu May 22 23:47:46 PDT 2014
Hi Laurent,
On Fri, May 23, 2014 at 12:42 AM, Laurent Pinchart
<laurent.pinchart at ideasonboard.com> wrote:
> On Wednesday 21 May 2014 17:50:35 Geert Uytterhoeven wrote:
>> On Wed, May 21, 2014 at 5:41 PM, Laurent Pinchart wrote:
>> >> +cpg_clocks: cpg_clocks at e6150000 {
>> >> + compatible = "renesas,r8a7740-cpg-clocks";
>> >> + reg = <0xe6150000 0x10000>;
>> >
>> > This range covers the MSTP registers. Unlike other Renesas SoCs with CCF
>> > support in mainline, the MSTP registers are in the middle of the CPG
>> > address space here, which is quite problematic.
>>
>> This is also the case for R-Car Gen2?!?
>>
>> /* Special CPG clocks */
>> cpg_clocks: cpg_clocks at e6150000 {
>> compatible = "renesas,r8a7791-cpg-clocks",
>> "renesas,rcar-gen2-cpg-clocks";
>> reg = <0 0xe6150000 0 0x1000>;
>>
>> mstp0_clks: mstp0_clks at e6150130 {
>> compatible = "renesas,r8a7791-mstp-clocks",
>> "renesas,cpg-mstp-clocks";
>> reg = <0 0xe6150130 0 4>, <0 0xe6150030 0 4>;
>
> Indeed :-/ It looks like we're just lucky that of_iomap() doesn't request the
> memory resource.
>
> Do you think it would make sense to move the MSTP nodes inside the CPG node to
> better describe the hardware, and avoid future breakages if of_iomap() starts
> calling request_resource() tomorrow ?
That indeed makes sense.
In general (not limited to r8a7740), I think we should move a few more nodes.
Currently we have almost all device nodes at the top level, while they should be
under a bus node. Or in a tree of bus nodes, matching the block diagram
in the datasheet.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus TorvaldsH
More information about the linux-arm-kernel
mailing list