Re: [PATCH v3 43/44] dt: document HiSilicon SPMI controller and mfd/regulator properties

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Aug 18, 2020 at 4:18 PM Mauro Carvalho Chehab
<mchehab+huawei@xxxxxxxxxx> wrote:
>
> Em Tue, 18 Aug 2020 11:07:55 -0600
> Rob Herring <robh@xxxxxxxxxx> escreveu:
>
> > > > > +  spmi-channel:
> > > > > +    description: number of the SPMI channel where the PMIC is connected
> > > >
> > > > This looks like a common (to SPMI), but it's not something defined in
> > > > spmi.txt
> > >
> > > This one is not part of the SPMI core. It is stored inside a private
> > > structure inside at the HiSilicon spmi controller driver. It is stored
> > > there as ctrl_dev->channel, and it is used to calculate the register offset
> > > for readl():
> > >
> > >     offset  = SPMI_APB_SPMI_STATUS_BASE_ADDR;
> > >     offset += SPMI_CHANNEL_OFFSET * ctrl_dev->channel + SPMI_SLAVE_OFFSET * sid;
> > >     do {
> > >             status = readl(base + offset);
> > >     ...
> > >
> > > The SPMI bus is somewhat similar to I2C: it is a 2-wire serial bus
> > > with up to 16 devices connected to it.
> > >
> > > Now, most modern I2C chipsets provide multiple independent I2C
> > > channels, on different pins. Also, on some chipsets, certain
> > > GPIO pins can be used either as GPIO or as I2C.
> > >
> > > I strongly suspect that this is the case here: according with
> > > the Hikey 970 schematics:
> > >
> > >     https://www.96boards.org/documentation/consumer/hikey/hikey970/hardware-docs/files/hikey970-schematics.pdf
> > >
> > > The pins used by SPMI clock/data can also be used as GPIO.
> > >
> > > While I don't have access to the datasheets for Kirin 970 (or any other
> > > chipsets on this board), for me, it sounds that different GPIO pins
> > > are allowed to use SPMI. The "spmi-channel" property specifies
> > > what pins will be used for SPMI, among the ones that are
> > > compatible with MIPI SPMI specs.
> >
> > Based on this, I think it should be called 'hisilicon,spmi-channel' as
> > it is vendor specific.
>
> I'm fine with "hisilicon,spmi-channel".

Humm, QCom has a 'qcom,channel' property for SPMI. Seems like maybe it
should be a common thing.

Rob



[Index of Archives]     [Device Tree Compilter]     [Device Tree Spec]     [Linux Driver Backports]     [Video for Linux]     [Linux USB Devel]     [Linux PCI Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Yosemite Backpacking]


  Powered by Linux