Re: [PATCH 3/4] gpio: sifive: Add GPIO driver for SiFive SoCs

[Bartosz Golaszewski <bgolaszewski@xxxxxxxxxxxx>]

pon., 18 lis 2019 o 11:03 Yash Shah <yash.shah@xxxxxxxxxx> napisał(a):
>
> > -----Original Message-----
> > From: Bartosz Golaszewski <bgolaszewski@xxxxxxxxxxxx>
> > Sent: 13 November 2019 18:41
> > To: Yash Shah <yash.shah@xxxxxxxxxx>
> > Cc: linus.walleij@xxxxxxxxxx; robh+dt@xxxxxxxxxx; mark.rutland@xxxxxxx;
> > palmer@xxxxxxxxxxx; Paul Walmsley ( Sifive) <paul.walmsley@xxxxxxxxxx>;
> > aou@xxxxxxxxxxxxxxxxx; tglx@xxxxxxxxxxxxx; jason@xxxxxxxxxxxxxx;
> > maz@xxxxxxxxxx; bmeng.cn@xxxxxxxxx; atish.patra@xxxxxxx; Sagar Kadam
> > <sagar.kadam@xxxxxxxxxx>; linux-gpio@xxxxxxxxxxxxxxx;
> > devicetree@xxxxxxxxxxxxxxx; linux-riscv@xxxxxxxxxxxxxxxxxxx; linux-
> > kernel@xxxxxxxxxxxxxxx; Sachin Ghadi <sachin.ghadi@xxxxxxxxxx>
> > Subject: Re: [PATCH 3/4] gpio: sifive: Add GPIO driver for SiFive SoCs
> >
> > wt., 12 lis 2019 o 13:12 Yash Shah <yash.shah@xxxxxxxxxx> napisał(a):
> > >
> > > Adds the GPIO driver for SiFive RISC-V SoCs.
> > >
> > > Signed-off-by: Wesley W. Terpstra <wesley@xxxxxxxxxx>
> > > [Atish: Various fixes and code cleanup]
> > > Signed-off-by: Atish Patra <atish.patra@xxxxxxx>
> > > Signed-off-by: Yash Shah <yash.shah@xxxxxxxxxx>
>
> [...]
>
> > > +
> > > +static int sifive_gpio_probe(struct platform_device *pdev) {
> > > +       struct device *dev = &pdev->dev;
> > > +       struct device_node *node = pdev->dev.of_node;
> > > +       struct device_node *irq_parent;
> > > +       struct irq_domain *parent;
> > > +       struct gpio_irq_chip *girq;
> > > +       struct sifive_gpio *chip;
> > > +       struct resource *res;
> > > +       int ret, ngpio;
> > > +
> > > +       chip = devm_kzalloc(dev, sizeof(*chip), GFP_KERNEL);
> > > +       if (!chip)
> > > +               return -ENOMEM;
> > > +
> > > +       res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> > > +       chip->base = devm_ioremap_resource(dev, res);
> >
> > Use devm_platform_ioremap_resource() and drop the res variable.
> >
>
> Sure, will do that.
>
> > > +       if (IS_ERR(chip->base)) {
> > > +               dev_err(dev, "failed to allocate device memory\n");
> > > +               return PTR_ERR(chip->base);
> > > +       }
> > > +
> > > +       chip->regs = devm_regmap_init_mmio(dev, chip->base,
> > > +
> > > + &sifive_gpio_regmap_config);
> >
> > Why do you need this regmap here? You initialize a new regmap, then use
> > your own locking despite not having disabled the internal locking in regmap,
> > and then you initialize the mmio generic GPIO code which will use yet
> > another lock to operate on the same registers and in the end you write to
> > those registers without taking any lock anyway.
> > Doesn't make much sense to me.
> >
>
> As suggested in the comments received on the RFC version of this patch[0], I am trying to use regmap MMIO by looking at gpio-mvebu.c. I got your point regarding the usage of own locks is not making any sense.
> Here is what I will do in v2:
> 1. drop the usage of own locks
> 2. consistently use regmap_* apis for register access (replace all iowrites).
> Does this make sense now?

The thing is: the gpio-mmio code you're (correctly) reusing uses a
different lock - namely: bgpio_lock in struct gpio_chip. If you want
to use regmap for register operations, then you need to set
disable_locking in regmap_config to true and then take this lock
manually on every access.

Bart

>
> > > +       if (IS_ERR(chip->regs))
> > > +               return PTR_ERR(chip->regs);
> > > +
>
> [...]
>
> > > +
> > > +       ret = gpiochip_add_data(&chip->gc, chip);
> > > +       if (ret)
> > > +               return ret;
> > > +
> > > +       platform_set_drvdata(pdev, chip);
> > > +       dev_info(dev, "SiFive GPIO chip registered %d GPIOs\n",
> > > + ngpio);
> >
> > Core gpio library emits a very similar debug message from
> > gpiochip_setup_dev(), I think you can drop it and directly return
> > gpiochip_add_data().
> >
> > Bartosz
>
> Ok. Will directly return gpiochip_add_data().
> Thanks for your comments!
>
> - Yash
>
> [0] https://lore.kernel.org/linux-riscv/20181010123519.RVexDppaPFpIWl7QU_hpP8tc5qqWPJgeuLYn0FaGbeQ@z/