Re: [PATCH 2/2] clk: bcm63xx-gate: add BCM6318 support

[Philippe Mathieu-DaudÃ]

On Wed, Jun 10, 2020 at 5:32 PM Florian Fainelli <f.fainelli@xxxxxxxxx> wrote:
> On 6/10/2020 1:29 AM, Philippe Mathieu-Daudà wrote:
> > Hi,
> >
> > On Wed, Jun 10, 2020 at 8:13 AM Ãlvaro FernÃndez Rojas
> > <noltari@xxxxxxxxx> wrote:
> >>
> >> Hi Florian,
> >>
> >>> El 10 jun 2020, a las 4:27, Florian Fainelli <f.fainelli@xxxxxxxxx> escribiÃ:
> >>>
> >>>
> >>>
> >>> On 6/9/2020 4:30 AM, Ãlvaro FernÃndez Rojas wrote:
> >>>> +static const struct clk_bcm63xx_table_entry bcm6318_clocks[] = {
> >>>> +    { .name = "adsl_asb", .bit = 0, },
> >>>> +    { .name = "usb_asb", .bit = 1, },
> >>>> +    { .name = "mips_asb", .bit = 2, },
> >>>> +    { .name = "pcie_asb", .bit = 3, },
> >>>> +    { .name = "phymips_asb", .bit = 4, },
> >>>> +    { .name = "robosw_asb", .bit = 5, },
> >>>> +    { .name = "sar_asb", .bit = 6, },
> >>>> +    { .name = "sdr_asb", .bit = 7, },
> >>>> +    { .name = "swreg_asb", .bit = 8, },
> >>>> +    { .name = "periph_asb", .bit = 9, },
> >>>> +    { .name = "cpubus160", .bit = 10, },
> >>>> +    { .name = "adsl", .bit = 11, },
> >>>> +    { .name = "sar124", .bit = 12, },
> >>>
> >>> Nit: this should be sar125
> >>
> >> Nice catch, I will fix this in v2.
> >>
> >>>
> >>>> +    { .name = "mips", .bit = 13, .flags = CLK_IS_CRITICAL, },
> >>>> +    { .name = "pcie", .bit = 14, },
> >>>> +    { .name = "robosw250", .bit = 16, },
> >>>> +    { .name = "robosw025", .bit = 17, },
> >>>> +    { .name = "sdr", .bit = 19, .flags = CLK_IS_CRITICAL, },
> >>>> +    { .name = "usb", .bit = 20, },
> >>>
> >>> This should probably be "usbd" to indicate this is the USB device clock
> >>> (not host)
> >>
> >> Ok, I will change it. I got confused by the fact that both (usbd and usbh) were present on 6318_map_part.h:
> >> #define USBD_CLK_EN         (1 << 20)
> >> #define USBH_CLK_EN         (1 << 20)
> >
> > Is there a datasheet to verify that?
>
> Not a public one, but I can confirm this is correct given the internal
> datasheet.

OK thank you Florian.

> --
> Florian