RE: [PATCH v2 2/3] media: i2c: ov5645: Switch to assigned-clock-rates

[Prabhakar Mahadev Lad]

Hi Laurent,

Thank you for the review.

> -----Original Message-----
> From: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx>
> Sent: 12 March 2020 23:03
> To: Prabhakar Mahadev Lad <prabhakar.mahadev-lad.rj@xxxxxxxxxxxxxx>
> Cc: Geert Uytterhoeven <geert@xxxxxxxxxxxxxx>; Mauro Carvalho Chehab
> <mchehab@xxxxxxxxxx>; Shawn Guo <shawnguo@xxxxxxxxxx>; Sascha
> Hauer <s.hauer@xxxxxxxxxxxxxx>; Pengutronix Kernel Team
> <kernel@xxxxxxxxxxxxxx>; Rob Herring <robh+dt@xxxxxxxxxx>; Mark
> Rutland <mark.rutland@xxxxxxx>; Sakari Ailus
> <sakari.ailus@xxxxxxxxxxxxxxx>; NXP Linux Team <linux-imx@xxxxxxx>;
> Magnus Damm <magnus.damm@xxxxxxxxx>; Ezequiel Garcia
> <ezequiel@xxxxxxxxxxxxx>; open list:OPEN FIRMWARE AND FLATTENED
> DEVICE TREE BINDINGS <devicetree@xxxxxxxxxxxxxxx>; Linux Kernel Mailing
> List <linux-kernel@xxxxxxxxxxxxxxx>; Linux-Renesas <linux-renesas-
> soc@xxxxxxxxxxxxxxx>; Fabio Estevam <festevam@xxxxxxxxx>; Linux ARM
> <linux-arm-kernel@xxxxxxxxxxxxxxxxxxx>; Linux Media Mailing List <linux-
> media@xxxxxxxxxxxxxxx>
> Subject: Re: [PATCH v2 2/3] media: i2c: ov5645: Switch to assigned-clock-
> rates
>
> Hi Prabhakar,
>
> On Thu, Mar 12, 2020 at 09:52:38PM +0000, Prabhakar Mahadev Lad wrote:
> > On 12 March 2020 21:42, Geert Uytterhoeven wrote:
> > > On Thu, Mar 12, 2020 at 10:13 PM Lad Prabhakar wrote:
> > > > This patch switches to assigned-clock-rates for specifying the clock rate.
> > > > The clk-conf.c internally handles setting the clock rate when
> > > > assigned-clock-rates is passed.
> > > >
> > > > The driver now sets the clock frequency only if the deprecated
> > > > property clock-frequency is defined instead assigned-clock-rates,
> > > > this is to avoid breakage with existing DT binaries.
> > > >
> > > > Signed-off-by: Lad Prabhakar
> > > > <prabhakar.mahadev-lad.rj@xxxxxxxxxxxxxx>
> > >
> > > Thanks for your patch!
> > >
> > > > --- a/drivers/media/i2c/ov5645.c
> > > > +++ b/drivers/media/i2c/ov5645.c
> > > > @@ -1055,6 +1055,7 @@ static int ov5645_probe(struct i2c_client
> *client)
> > > >         struct device_node *endpoint;
> > > >         struct ov5645 *ov5645;
> > > >         u8 chip_id_high, chip_id_low;
> > > > +       bool set_clk = false;
> > > >         unsigned int i;
> > > >         u32 xclk_freq;
> > > >         int ret;
> > > > @@ -1094,10 +1095,17 @@ static int ov5645_probe(struct i2c_client
> > > *client)
> > > >                 return PTR_ERR(ov5645->xclk);
> > > >         }
> > > >
> > > > -       ret = of_property_read_u32(dev->of_node, "clock-frequency",
> > > &xclk_freq);
> > > > +       ret = of_property_read_u32(dev->of_node, "assigned-clock-
> rates",
> > > > +                                  &xclk_freq);
> > > >         if (ret) {
> > >
> > > I think you can just leave out the above check...
> > >
> > > > -               dev_err(dev, "could not get xclk frequency\n");
> > > > -               return ret;
> > > > +               /* check if deprecated property clock-frequency is defined */
> > > > +               ret = of_property_read_u32(dev->of_node, "clock-
> frequency",
> > > > +                                          &xclk_freq);
> > > > +               if (ret) {
> > >
> > > ... and ignore the absence of the deprecated property.
> > >
> > > > +                       dev_err(dev, "could not get xclk frequency\n");
> > > > +                       return ret;
> > > > +               }
> > > > +               set_clk = true;
> > >
> > > I.e. just
> > >
> > >         /* check if deprecated property clock-frequency is defined */
> > >         xclk_freq = 0;
> > >         of_property_read_u32(dev->of_node, "clock-frequency",
> &xclk_freq);
> > >         if (xclk_freq) {
> > >                 ret = clk_set_rate(ov5645->xclk, xclk_freq);
> > >                 if (ret) {
> > >                         dev_err(dev, "could not set xclk frequency\n");
> > >                         return ret;
> > >                 }
> > >         } else {
> > >                 xclk_freq = clk_get_rate(ov5645->xclk, xclk_freq);
> > >         }
> > >
> > I did think about it initially, but realized the clk_get_rate may vary
> > platform to platform for example in G2E clk_get_rate() returns a
> > frequency of 24242424 with assigned-clock-rates set to  24000000 and
> > probe would fail due to a check for external clock must be 24MHz, with 1%
> tolerance.
>
> Then you need to handle the tolerance in this driver :-)
>
Sure I will increase the tolerance here.

Cheers,
--Prabhakar

> > > >         }
> > > >
> > > >         /* external clock must be 24MHz, allow 1% tolerance */ @@
> > > > -1107,10 +1115,12 @@ static int ov5645_probe(struct i2c_client *client)
> > > >                 return -EINVAL;
> > > >         }
> > > >
> > > > -       ret = clk_set_rate(ov5645->xclk, xclk_freq);
> > > > -       if (ret) {
> > > > -               dev_err(dev, "could not set xclk frequency\n");
> > > > -               return ret;
> > > > +       if (set_clk) {
> > > > +               ret = clk_set_rate(ov5645->xclk, xclk_freq);
> > > > +               if (ret) {
> > > > +                       dev_err(dev, "could not set xclk frequency\n");
> > > > +                       return ret;
> > > > +               }
> > > >
> > > >         }
> > > >
> > > >         for (i = 0; i < OV5645_NUM_SUPPLIES; i++)
>
> --
> Regards,
>
> Laurent Pinchart


Renesas Electronics Europe GmbH, Geschaeftsfuehrer/President: Carsten Jauch, Sitz der Gesellschaft/Registered office: Duesseldorf, Arcadiastrasse 10, 40472 Duesseldorf, Germany, Handelsregister/Commercial Register: Duesseldorf, HRB 3708 USt-IDNr./Tax identification no.: DE 119353406 WEEE-Reg.-Nr./WEEE reg. no.: DE 14978647