Re: [PATCH 2/6] pinctrl: Update clock handling for thepinctrl-nomadik GPIO driver

[Linus Walleij]

On Wed, Oct 24, 2012 at 4:45 PM, Lee Jones <lee.jones@xxxxxxxxxx> wrote:

> The clock framework has changed somewhat and it's now better to
> invoke clock_prepare_enable() and clk_disable_unprepare() rather
> than the legacy clk_enable() and clk_disable() calls. This patch
> converts the Nomadik Pin Control driver to the new framework.
>
> Signed-off-by: Lee Jones <lee.jones@xxxxxxxxxx>

(...)
> -               clk_enable(chip->clk);
> +               clk_prepare_enable(chip->clk);
(...)
> -               clk_disable(chip->clk);
> +               clk_disable_unprepare(chip->clk);

(Repeated for each occurence.)

Is this *really* causing a regression? I mean the driver
begin like this in nmk_gpio_probe():

        clk = devm_clk_get(&dev->dev, NULL);
        if (IS_ERR(clk)) {
                ret = PTR_ERR(clk);
                goto out;
        }
        clk_prepare(clk);

Then it leaves the clock prepared. So the clock is always
prepared. You would only need to enable/disable it at times.

And the semantics of the clk_enable/clk_disable call pair
is such that it is fastpath and should be real quick, and that
is exactly why we're using it repeatedly like that. Inserting
clk_unprepare() effectively could make the whole driver a
lot slower, so convince me on this one. ...

I suspect the real bug (if there is one) must be in the clock
implementation.

Yours,
Linus Walleij
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/