RE: [PATCH V4 2/5] clocksource/drivers/sysctr: Add clock-frequency property

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

 



Hi, Rob

> On Mon, Jul 1, 2019 at 3:47 AM <Anson.Huang@xxxxxxx> wrote:
> >
> > From: Anson Huang <Anson.Huang@xxxxxxx>
> 
> 'dt-bindings: timer: ...' for the subject.

OK, I made a mistake.

> 
> >
> > Systems which use platform driver model for clock driver require the
> > clock frequency to be supplied via device tree when system counter
> > driver is enabled.
> 
> This is a DT binding. What's a platform driver?

It is just trying to explain why we need to introduce this "clock-frequency"
property, nothing about the binding and platform driver.

> 
> >
> > This is necessary as in the platform driver model the of_clk
> > operations do not work correctly because system counter driver is
> > initialized in early phase of system boot up, and clock driver using
> > platform driver model is NOT ready at that time, it will cause system
> > counter driver initialization failed.
> >
> > Add clock-frequency property to the device tree bindings of the NXP
> > system counter, so the driver can tell timer-of driver to get clock
> > frequency from DT directly instead of doing of_clk operations via clk
> > APIs.
> 
> While you've now given a good explanation why you need this, it all sounds
> like linux specific issues and a DT change should not be necessary.
> 
> Presumably, 'clocks' points to a fixed-clock node, right? Just parse the 'clocks'
> phandle and fetch the frequency from that node if you need to get the
> frequency 'early'.

Sound like a better solution, I will try that, since the system counter's clock is
from osc_24m and divided by 3, since different platforms may have different divider,
so maybe I can create a fixed-clock node in DT, then system counter driver can parse
the clock and fetch the frequency from that node, will redo a V5 patch.

> 
> > Signed-off-by: Anson Huang <Anson.Huang@xxxxxxx>
> > ---
> > No change.
> > ---
> >  .../devicetree/bindings/timer/nxp,sysctr-timer.txt        | 15 +++++++++------
> >  1 file changed, 9 insertions(+), 6 deletions(-)
> >
> > diff --git
> > a/Documentation/devicetree/bindings/timer/nxp,sysctr-timer.txt
> > b/Documentation/devicetree/bindings/timer/nxp,sysctr-timer.txt
> > index d576599..7088a0e 100644
> > --- a/Documentation/devicetree/bindings/timer/nxp,sysctr-timer.txt
> > +++ b/Documentation/devicetree/bindings/timer/nxp,sysctr-timer.txt
> > @@ -11,15 +11,18 @@ Required properties:
> >  - reg :             Specifies the base physical address and size of the comapre
> >                      frame and the counter control, read & compare.
> >  - interrupts :      should be the first compare frames' interrupt
> > -- clocks :         Specifies the counter clock.
> > -- clock-names:             Specifies the clock's name of this module
> > +- clocks :          Specifies the counter clock, mutually exclusive with clock-
> frequency.
> > +- clock-names :     Specifies the clock's name of this module, mutually
> exclusive with
> > +                   clock-frequency.
> > +- clock-frequency : Specifies system counter clock frequency, mutually
> exclusive with
> > +                   clocks/clock-names.
> 
> It doesn't really work to say one or the other is needed unless you make the
> OS support both cases.

The OS already support both cases now with this patch series.

Thanks,
Anson




[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