[PATCH 02/13] ARM: dts: r8a73a4: Remove unnecessary clock-output-names properties
Simon Horman
horms at verge.net.au
Tue Mar 22 17:52:57 PDT 2016
On Tue, Mar 22, 2016 at 11:17:09AM +0100, Geert Uytterhoeven wrote:
> On Tue, Mar 22, 2016 at 2:40 AM, Simon Horman
> <horms+renesas at verge.net.au> wrote:
> > * Fixed rate and fixed factor clocks do not require an
> > clock-output-names property.
> > * Since 07705583e920fef6 ("clk: shmobile: div6: Make clock-output-names
> > optional") Renesas div6 clocks do not require a clock-output-names
> > property.
> >
> > In the above cases there is only one clock output and its name is taken
> > from that of the clock node.
> >
> > Accordingly, remove the unnecessary clock-output-names properties and as
> > necessary update the node names. The zb_clk is excluded from this rename
> > as the MSTP clock driver (clk-mstp.c) explicitly looks for a clock
> > named zb_clk for the r8a73a4 and sh73a0 SoCs.
>
> Yeah, that was a bit hacky, but I didn't see a better solution with the MSTP
> driver. CPG/MSSR would solve that differently.
It made my eyes bleed but I'm over it now :)
> > Signed-off-by: Simon Horman <horms+renesas at verge.net.au>
> > ---
> > arch/arm/boot/dts/r8a73a4.dtsi | 76 ++++++++++++++----------------------------
> > 1 file changed, 25 insertions(+), 51 deletions(-)
> >
> > diff --git a/arch/arm/boot/dts/r8a73a4.dtsi b/arch/arm/boot/dts/r8a73a4.dtsi
> > index 6583a1dfca1f..6eb23221cf2d 100644
> > --- a/arch/arm/boot/dts/r8a73a4.dtsi
> > +++ b/arch/arm/boot/dts/r8a73a4.dtsi
>
> > @@ -538,173 +533,152 @@
> > clocks = <&pll1_div2_clk>, <0>,
> > <&cpg_clocks R8A73A4_CLK_PLL2S>, <0>;
> > #clock-cells = <0>;
> > - clock-output-names = "zb";
>
> I'm wondering whether we should keep this one for now?
> Dropping it changes the clock name from "zb_clk" to "zb".
>
> Nothing should rely on that, though, but keeping "zb" would allow us to e.g.
> change the MSTP driver to match on clock name instead of node name, and drop
> clock-output-names and rename the clock node later.
Good idea, there seems little point in painting ourselves into a corner.
I'll drop the zb change from this and the sh73a0 patch.
We can clean things up later once/if we have a good plan.
More information about the linux-arm-kernel
mailing list