Re: [PATCH v2 3/3] riscv: dts: thead: add basic spi node

[Drew Fustini <pdp7pdp7@xxxxxxxxx>]

On Wed, Jul 03, 2024 at 03:45:37PM +0100, Conor Dooley wrote:
> Kanak, Drew,
> 
> On Mon, Jul 01, 2024 at 05:43:54PM +0530, Kanak Shilledar wrote:
> > created spi0 node with fixed clock. the spi0 node
> > uses synopsis designware driver and has the following
> > compatible "snps,dw-apb-ssi". the spi0 node is connected
> > to a SPI NOR flash pad which is left unpopulated on the back
> > side of the board.
> > 
> > Signed-off-by: Kanak Shilledar <kanakshilledar@xxxxxxxxx>
> > ---
> > Changes in v2:
> > - Separated from a single patch file
> > ---
> >  .../boot/dts/thead/th1520-beaglev-ahead.dts      |  9 +++++++++
> >  .../boot/dts/thead/th1520-lichee-module-4a.dtsi  |  4 ++++
> >  .../riscv/boot/dts/thead/th1520-lichee-pi-4a.dts |  5 +++++
> 
> Didn't you say there was a flash on one of these two boards?
> 
> >  arch/riscv/boot/dts/thead/th1520.dtsi            | 16 ++++++++++++++++
> >  4 files changed, 34 insertions(+)
> > 
> > diff --git a/arch/riscv/boot/dts/thead/th1520-beaglev-ahead.dts b/arch/riscv/boot/dts/thead/th1520-beaglev-ahead.dts
> > index d9b4de9e4757..3103b74e0288 100644
> > --- a/arch/riscv/boot/dts/thead/th1520-beaglev-ahead.dts
> > +++ b/arch/riscv/boot/dts/thead/th1520-beaglev-ahead.dts
> > @@ -17,6 +17,7 @@ aliases {
> >  		gpio1 = &gpio1;
> >  		gpio2 = &gpio2;
> >  		gpio3 = &gpio3;
> > +		spi0 = &spi0;
> 
> "spi" would sort after "serial".
> 
> >  		serial0 = &uart0;
> >  		serial1 = &uart1;
> >  		serial2 = &uart2;
> > @@ -52,6 +53,10 @@ &sdhci_clk {
> >  	clock-frequency = <198000000>;
> >  };
> >  
> > +&spi_clk {
> > +	clock-frequency = <396000000>;
> > +};
> 
> I'm pretty sceptical about adding more of these fixed clocks, rather
> than waiting for the clock driver. Drew, what do you think? Should we
> just add one more to your fixup list or would you rather delay? Guess it
> depends on how long more you think that clock driver is likely to take.

I think the clk driver [1] is in good shape but it has not been reviewed
by the clk maintainer yet. Thus it is hard to predict any timeline for
it getting merged.

SPI support doesn't require any driver changes so I'd be inclined to
allow the fixed clock in this case. It will be simple to change it over
to a real clock once the clk driver is upstream.

Acked-by: Drew Fustini <drew@xxxxxxxx>

Thanks,
Drew