Hi Florian,
On Tue, 2020-03-03 at 10:21 -0800, Florian Fainelli wrote:
> On 3/3/20 4:08 AM, Nicolas Saenz Julienne wrote:
> > Depending on bcm2711's revision its emmc2 controller might have
> > different DMA constraints. Raspberry Pi 4's firmware will take care of
> > updating those, but only if a certain alias is found in the device tree.
> > So, move emmc2 into its own bus, so as not to pollute other devices with
> > dma-ranges changes and create the emmc2bus alias.
> >
> > Based in Phil ELwell's downstream implementation.
> >
> > Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@xxxxxxx>
>
> Nit: the subject should be ARM: dts: bcm2711. Some more comments below.
Of course, should have known better.
> > ---
> > arch/arm/boot/dts/bcm2711-rpi-4-b.dts | 1 +
> > arch/arm/boot/dts/bcm2711.dtsi | 19 ++++++++++++++-----
> > 2 files changed, 15 insertions(+), 5 deletions(-)
> >
> > diff --git a/arch/arm/boot/dts/bcm2711-rpi-4-b.dts
> > b/arch/arm/boot/dts/bcm2711-rpi-4-b.dts
> > index 1d4b589fe233..e26ea9006378 100644
> > --- a/arch/arm/boot/dts/bcm2711-rpi-4-b.dts
> > +++ b/arch/arm/boot/dts/bcm2711-rpi-4-b.dts
> > @@ -20,6 +20,7 @@ memory@0 {
> > };
> >
> > aliases {
> > + emmc2bus = &emmc2bus;
> > ethernet0 = &genet;
> > pcie0 = &pcie0;
> > };
> > diff --git a/arch/arm/boot/dts/bcm2711.dtsi b/arch/arm/boot/dts/bcm2711.dtsi
> > index d1e684d0acfd..61ea8b44c51e 100644
> > --- a/arch/arm/boot/dts/bcm2711.dtsi
> > +++ b/arch/arm/boot/dts/bcm2711.dtsi
> > @@ -241,17 +241,26 @@ pwm1: pwm@7e20c800 {
> > status = "disabled";
> > };
> >
> > + hvs@7e400000 {
> > + interrupts = <GIC_SPI 97 IRQ_TYPE_LEVEL_HIGH>;
> > + };
> > + };
> > +
> > + emmc2bus: emmc2bus {
> > + compatible = "simple-bus";
> > + #address-cells = <2>;
> > + #size-cells = <1>;
> > +
> > + ranges = <0x0 0x7e000000 0x0 0xfe000000 0x01800000>;
> > + dma-ranges = <0x0 0xc0000000 0x0 0x00000000 0x40000000>;
>
> This deserves a comment for two reasons:
>
> - explaining which of these properties is getting patched by the
> firmware (and it would be really nice if we had a concept of annotation
> attributes for Device Tree such that you could express something like:
>
> dma-ranges = <> __patchable;
Something like this would've been useful to me some time ago while debugging
CMA issues on a random arm64 Board. I was left wondering if the memory nodes on
that specific board were set in stone or just a placeholder.
> - explaining why this is not collapsed in the soc bus node, because the
> dma-ranges constraint can be different based on the Pi4 revision
Noted
> With that fixed, this looks good to me!
Thanks!
>
> > +
> > emmc2: emmc2@7e340000 {
> > compatible = "brcm,bcm2711-emmc2";
> > - reg = <0x7e340000 0x100>;
> > + reg = <0x0 0x7e340000 0x100>;
> > interrupts = <GIC_SPI 126 IRQ_TYPE_LEVEL_HIGH>;
> > clocks = <&clocks BCM2711_CLOCK_EMMC2>;
> > status = "disabled";
> > };
> > -
> > - hvs@7e400000 {
> > - interrupts = <GIC_SPI 97 IRQ_TYPE_LEVEL_HIGH>;
> > - };
> > };
> >
> > arm-pmu {
> >
>
>
Attachment:
signature.asc
Description: This is a digitally signed message part
