Re: [PATCH] arm64: dts: renesas: sparrow-hawk: Add overlay for WaveShare 2CH CANFD HAT
From: Geert Uytterhoeven
Date: Wed Sep 09 2026 - 04:04:16 EST
Hi Marek,
On Tue, 8 Sept 2026 at 19:02, Marek Vasut <marek.vasut@xxxxxxxxxxx> wrote:
> On 9/8/26 2:02 PM, Geert Uytterhoeven wrote:
> >> +&pfc {
> >> + msiof0_pins: spi-ws-2ch-canfd0 { /* SPI_0 */
> >
> > Node name should be "msiof0".
>
> Do we also have to fix up
> arch/arm64/boot/dts/renesas/r8a779g3-sparrow-hawk.dts this:
>
> "
> 755 msiof1_pins: sound {
> ^^^^^
> "
> ?
Might be a good idea, to increase consistency.
> But I wonder, shouldn't the DTOs contain unique pinmux node names, which
> cannot accidentally overlap with the base DT or other DTO node names ?
In general, the answer is yes.
In this case it is about pin control subnodes. If the base DTS already
uses msiof0, its original user needs to be disabled by the overlay
anyway.
> >> + groups = "msiof0_clk", "msiof0_txd", "msiof0_rxd";
> >> + function = "msiof0";
> >> + };
> >> +};
> >> +
> >> +/*
> >> + * PWM1 uses PWM1_B pin configuration for pin GP2_13. This conflicts
> >> + * with SPI_1 CE_1 signal which is also on pin GP2_13. Disable PWM1
> >> + * in case this hat is in use.
> >> + */
> >> +&pwm1 {
> >> + status = "disabled";
> >> +};
> >
> > Please move this node disabling pwm1 to the top, for compatibility
> > with dynamic DT overlays.
> Could you expand on that a bit ? What difference does the placement of
> this node make within the DTSO ? Please pardon my ignorance.
In case of a dynamic overlay, pwm1 must be disabled before SPI_1 is
probed. Otherwise the latter will fail with -EBUSY due to the
conflicting pin being still in use.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
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 Torvalds