Re: dma_mask limited to 32-bits with OF platform device

[Rob Herring]

On Tue, Mar 3, 2020 at 8:06 AM Robin Murphy <robin.murphy@xxxxxxx> wrote:
>
> On 03/03/2020 8:27 am, Roger Quadros wrote:
> [...]
> >> With the patch (in the end). dev->bus_dma_limit is still set to 0 and
> >> so is not being used.
> >>
> >> from of_dma_configure()
> >>          ret = of_dma_get_range(np, &dma_addr, &paddr, &size);
> >> ...
> >>          /* ...but only set bus limit if we found valid dma-ranges
> >> earlier */
> >>          if (!ret)
> >>                  dev->bus_dma_limit = end;
> >>
> >> There is no other place bus_dma_limit is set. Looks like every device
> >> should inherit that
> >> from it's parent right?
> >
> > Any ideas how to expect this to work?
>
> Is of_dma_get_range() actually succeeding, or is it tripping up on some
> aspect of the DT (in which case there should be errors in the log)?
>
> Looking again at the fragment below, are you sure it's correct? It
> appears to me like it might actually be defining a 1-byte-long DMA
> range, which indeed I wouldn't really expect to work.

Indeed, though it took me a minute to see why.

>
> Robin.
>
> >>
> >> diff --git a/arch/arm/boot/dts/dra7.dtsi b/arch/arm/boot/dts/dra7.dtsi
> >> index 64a0f90f5b52..5418c31d4da7 100644
> >> --- a/arch/arm/boot/dts/dra7.dtsi
> >> +++ b/arch/arm/boot/dts/dra7.dtsi
> >> @@ -680,15 +680,22 @@
> >>           };
> >>
> >>           /* OCP2SCP3 */
> >> -        sata: sata@4a141100 {
> >> -            compatible = "snps,dwc-ahci";
> >> -            reg = <0x4a140000 0x1100>, <0x4a141100 0x7>;

Based on this, the parent address size is 1 cell...

> >> -            interrupts = <GIC_SPI 49 IRQ_TYPE_LEVEL_HIGH>;
> >> -            phys = <&sata_phy>;
> >> -            phy-names = "sata-phy";
> >> -            clocks = <&l3init_clkctrl DRA7_L3INIT_SATA_CLKCTRL 8>;
> >> -            ti,hwmods = "sata";
> >> -            ports-implemented = <0x1>;
> >> +        sata_aux_bus {
> >> +            #address-cells = <2>;
> >> +            #size-cells = <2>;
> >> +            compatible = "simple-bus";
> >> +            ranges = <0x0 0x0 0x4a140000 0x0 0x1200>;
> >> +            dma-ranges = <0x0 0x0 0x0 0x0 0x1 0x00000000>;

So this is:
child addr: 0x0 0x0
parent addr: 0x0
size: 0x0 0x1

The last cell is just ignored I guess if you aren't seeing any errors.
We check this in dtc for ranges, but not dma-ranges. So I'm fixing
that.

> >> +            sata: sata@4a141100 {
> >> +                compatible = "snps,dwc-ahci";
> >> +                reg = <0x0 0x0 0x0 0x1100>, <0x0 0x1100 0x0 0x7>;
> >> +                interrupts = <GIC_SPI 49 IRQ_TYPE_LEVEL_HIGH>;
> >> +                phys = <&sata_phy>;
> >> +                phy-names = "sata-phy";
> >> +                clocks = <&l3init_clkctrl DRA7_L3INIT_SATA_CLKCTRL 8>;
> >> +                ti,hwmods = "sata";
> >> +                ports-implemented = <0x1>;
> >> +            };
> >>           };
> >>
> >>           /* OCP2SCP1 */
> >>
> >