Re: [PATCH 3/5] arm64: dts: qcom: Add base SC8380XP dtsi and the QCP dts

[Sibi Sankar]

Hey Konrad,

On 10/26/23 16:06, Konrad Dybcio wrote:
> On 10/25/23 16:24, Sibi Sankar wrote:
> > From: Rajendra Nayak <quic_rjendra@xxxxxxxxxxx>
> >
> > Add base dtsi and QCP board (Qualcomm Compute Platform) dts file for
> > SC8380XP SoC, describing the CPUs, GCC and RPMHCC clock controllers,
> > geni UART, interrupt controller, TLMM, reserved memory, interconnects,
> > SMMU and LLCC nodes.
> >
> > Co-developed-by: Abel Vesa <abel.vesa@xxxxxxxxxx>
> > Signed-off-by: Abel Vesa <abel.vesa@xxxxxxxxxx>
> > Signed-off-by: Rajendra Nayak <quic_rjendra@xxxxxxxxxxx>
> > Co-developed-by: Sibi Sankar <quic_sibis@xxxxxxxxxxx>
> > Signed-off-by: Sibi Sankar <quic_sibis@xxxxxxxxxxx>
> > ---
> [...]
>
> > +&tlmm {
> > +    gpio-reserved-ranges = <33 3>, <44 4>, <238 1>;
> It would be really cool if you added an explanation on why these
> GPIOs need to be reserved, especially since you can see what's
> connected on there on schematics. So, like:
>
> gpio-reserved-ranges = <33 3>, /* something */
>                 <44 4>, /* something else (fp scanner?)
>                 <238 1>; /* UFS reset? */

will do.

> [...]
>
> > +            compatible = "qcom,oryon";
> Again, this compatible won't fly unless all of these cores
> are totally identical and Oryon is only a name for this
> generation on this SoC (which I believe not to be the case).
>
> > +            reg = <0x0 0x0>;
> > +            enable-method = "psci";
> > +            next-level-cache = <&L1_0>;
> > +
> > +            L1_0: l1-cache {
> > +                compatible = "cache";
> I'm not sure if L1 is supposed to be described in the DT,
> Krzysztof should know.
>
> > +                next-level-cache = <&L2_0>;
> > +
> > +                L2_0: l2-cache-0 {
> > +                    compatible = "cache";
> cache-level?
> cache-unified?

ack will remove the l1 and use ^^ appropriately.

> [...]
>
> > +    memory@80000000 {
> > +        device_type = "memory";
> > +        /* We expect the bootloader to fill in the size */
> > +        reg = <0 0x80000000 0 0x80000000>;
> That contradicts the comment you made above. Plus, 2 GiB seems a
> bit low for this SoC :D

will fix this.

> [...]
>
> > +        gunyah_hyp_mem: gunyah-hyp-region@80000000 {
> you can probably strip the "-region" part, as this is implied by
> being a child of /reserved-memory

ok, will do but all the previous SoCs do it differently.

> > +        pld_pep_mem: pld-pep-region@81f30000 {
> What's PLD?
>
> What's this region used for? PEP is a Windows invention.

We list all the possible reserved memory regions from the reference doc.
I can remove the unused regions in the platforms dts later once this
series lands.

> [...]
>
> > +        av1_encoder_mem: av1-encoder-region@8e900000 {
> Is AV1enc hardware separate from iris?

no, it isn't separate from iris AFAIK.

> [...]
>
> > +        gcc: clock-controller@100000 {
> > +            compatible = "qcom,sc8380xp-gcc";
> > +            reg = <0 0x100000 0 0x200000>;
> The address part of reg should be padded to 8 hex digits.

thanks for catching this.

> > +
> > +                interconnects = <&clk_virt MASTER_QUP_CORE_2 0 
> > &clk_virt SLAVE_QUP_CORE_2 0>,
> QCOM_ICC_TAG_ALWAYS would be nicer than 0 (see sa8775p)

will do.

> [...]
>
> > +
> > +            interrupts =    <GIC_SPI 65 IRQ_TYPE_LEVEL_HIGH>,
> One space after and before '='

ack

> Konrad