Re: [PATCH v2 2/3] rtc: Add support for the MSTAR MSC313 RTC

[Romain Perier <romain.perier@gmail.com>]

Hello,

Le ven. 6 août 2021 à 21:38, Alexandre Belloni
<alexandre.belloni@bootlin.com> a écrit :
>
> Hello,
>
> On 01/08/2021 18:09:20+0200, Romain Perier wrote:
> > +static int msc313_rtc_probe(struct platform_device *pdev)
> > +{
> > +     struct device *dev = &pdev->dev;
> > +     struct msc313_rtc *priv;
> > +     int ret;
> > +     int irq;
> > +     unsigned long rate;
> > +     u16 reg;
> > +
> > +     priv = devm_kzalloc(&pdev->dev, sizeof(struct msc313_rtc), GFP_KERNEL);
> > +     if (!priv)
> > +             return -ENOMEM;
> > +
> > +     priv->rtc_base = devm_platform_ioremap_resource(pdev, 0);
> > +     if (IS_ERR(priv->rtc_base))
> > +             return PTR_ERR(priv->rtc_base);
> > +
> > +     irq = platform_get_irq(pdev, 0);
> > +     if (irq < 0)
> > +             return -EINVAL;
> > +
> > +     priv->rtc_dev = devm_rtc_allocate_device(dev);
> > +     if (IS_ERR(priv->rtc_dev))
> > +             return PTR_ERR(priv->rtc_dev);
> > +
> > +     priv->rtc_dev->ops = &msc313_rtc_ops;
> > +     priv->rtc_dev->range_min = RTC_TIMESTAMP_BEGIN_0000;
>
> I'm pretty sure this doesn't fit in this RTC registers, you should
> probably leave range_min to 0 (i.e. not set it at all).

ack

>
> > +     priv->rtc_dev->range_max = U32_MAX - 1; /* 2106-02-07 06:28:14 */
>
> I guess this one should be U32_MAX

ack

> > +
> > +     ret = devm_request_irq(dev, irq, msc313_rtc_interrupt, IRQF_SHARED,
> > +                            dev_name(&pdev->dev), &pdev->dev);
> > +     if (ret) {
> > +             dev_err(dev, "Could not request IRQ\n");
> > +             return ret;
> > +     }
> > +
> > +     priv->clk = devm_clk_get(dev, NULL);
> > +     if (IS_ERR(priv->clk)) {
> > +             dev_err(dev, "No input reference clock\n");
> > +             return PTR_ERR(priv->clk);
> > +     }
> > +
> > +     ret = clk_prepare_enable(priv->clk);
> > +     if (ret) {
> > +             dev_err(dev, "Failed to enable the reference clock, %d\n", ret);
> > +             return ret;
> > +     }
> > +
> > +     ret = devm_add_action_or_reset(dev, (void (*) (void *))clk_disable_unprepare, priv->clk);
> > +     if (ret)
> > +             return ret;
> > +
> > +     rate = clk_get_rate(priv->clk);
> > +     writew(rate & 0xFFFF, priv->rtc_base + REG_RTC_FREQ_CW_L);
> > +     writew((rate >> 16) & 0xFFFF, priv->rtc_base + REG_RTC_FREQ_CW_H);
> > +
> > +     reg = readw(priv->rtc_base + REG_RTC_CTRL);
> > +     reg |= CNT_EN_BIT;
> > +     writew(reg, priv->rtc_base + REG_RTC_CTRL);
> > +
>
> If on POR, CNT_EN_BIT is not set, then it would be nice to use that to
> know whether the RTC is properly set. You can then check CNT_EN_BIT in
> .read_time and return -EINVAL if it is not set. Then you can set the bit
> in .set_time. It is anyway useless to let the RTC running if it is not
> set.

Yeah, this is to be sure that the RTC is alive with a valid value
(which makes sense).
Ok I will fix everything in v3, then.

Romain