On Mon, Sep 19, 2022 at 11:14 PM Uwe Kleine-König
<u.kleine-koenig@xxxxxxxxxxxxxx> wrote:
>
> Hello,
>
> On Mon, Sep 19, 2022 at 03:39:15PM -0700, Olof Johansson wrote:
> > When the PWM driver was changed to disable clocks if no PWMs are enabled,
> > it ended up also disabling the shared parent with the UART, since the
> > UART doesn't do any clock enablement on its own.
> >
> > To avoid these surprises, add clk_prepare_enable/clk_disable_unprepare
> > calls.
> >
> > Fixes: ace41d7564e655 ("pwm: sifive: Ensure the clk is enabled exactly once per running PWM")
> > Cc: Uwe Kleine-König <u.kleine-koenig@xxxxxxxxxxxxxx>
> > Cc: Emil Renner Berthing <emil.renner.berthing@xxxxxxxxxxxxx>
> > Cc: Palmer Dabbelt <palmer@xxxxxxxxxxx>
> > Cc: Paul Walmsley <paul.walmsley@xxxxxxxxxx>
> > Signed-off-by: Olof Johansson <olof@xxxxxxxxx>
> > ---
> > drivers/tty/serial/sifive.c | 14 ++++++++++----
> > 1 file changed, 10 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/tty/serial/sifive.c b/drivers/tty/serial/sifive.c
> > index 5c3a07546a58..751f98068806 100644
> > --- a/drivers/tty/serial/sifive.c
> > +++ b/drivers/tty/serial/sifive.c
> > @@ -950,23 +950,28 @@ static int sifive_serial_probe(struct platform_device *pdev)
> > dev_err(&pdev->dev, "unable to find controller clock\n");
> > return PTR_ERR(clk);
> > }
> > + clk_prepare_enable(clk);
>
> The return code of clk_prepare_enable() must be checked. Also there is a
> simpler patch to fix that problem. I didn't test but I expect just doing
>
> diff --git a/drivers/tty/serial/sifive.c b/drivers/tty/serial/sifive.c
> index 5c3a07546a58..4b1d4fe8458e 100644
> --- a/drivers/tty/serial/sifive.c
> +++ b/drivers/tty/serial/sifive.c
> @@ -945,7 +945,7 @@ static int sifive_serial_probe(struct platform_device *pdev)
> return PTR_ERR(base);
> }
>
> - clk = devm_clk_get(&pdev->dev, NULL);
> + clk = devm_clk_get_enabled(&pdev->dev, NULL);
> if (IS_ERR(clk)) {
> dev_err(&pdev->dev, "unable to find controller clock\n");
> return PTR_ERR(clk);
>
> would be enough and also cares for disabling the clock on remove.
Yep, thanks for pointing that out. v2 sent out (after test).
-Olof
