2022-09-19 08:44:19

by Li Zhong

Subject: [PATCH v1] drivers/tty/serial/8250: check the return value of clk_set_rate()

Check the return of clk_set_rate() which could fail if there is error
setting the rate.

Signed-off-by: Li Zhong <[email protected]>
---
drivers/tty/serial/8250/8250_bcm7271.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/tty/serial/8250/8250_bcm7271.c b/drivers/tty/serial/8250/8250_bcm7271.c
index 8efdc271eb75..9014853d6e35 100644
--- a/drivers/tty/serial/8250/8250_bcm7271.c
+++ b/drivers/tty/serial/8250/8250_bcm7271.c
@@ -671,7 +671,9 @@ static void init_real_clk_rates(struct device *dev, struct brcmuart_priv *priv)
priv->real_rates[x] = clk_get_rate(priv->baud_mux_clk);
}
}
- clk_set_rate(priv->baud_mux_clk, priv->default_mux_rate);
+ rc = clk_set_rate(priv->baud_mux_clk, priv->default_mux_rate);
+ if (rc)
+ dev_err(dev, "Error restoring default BAUD MUX clock\n");
}

static void set_clock_mux(struct uart_port *up, struct brcmuart_priv *priv,
--
2.25.1

2022-09-19 09:07:27

by Greg KH

Subject: Re: [PATCH v1] drivers/tty/serial/8250: check the return value of clk_set_rate()

On Mon, Sep 19, 2022 at 01:15:23AM -0700, Li Zhong wrote:
> Check the return of clk_set_rate() which could fail if there is error
> setting the rate.
>
> Signed-off-by: Li Zhong <[email protected]>
> ---
> drivers/tty/serial/8250/8250_bcm7271.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/tty/serial/8250/8250_bcm7271.c b/drivers/tty/serial/8250/8250_bcm7271.c
> index 8efdc271eb75..9014853d6e35 100644
> --- a/drivers/tty/serial/8250/8250_bcm7271.c
> +++ b/drivers/tty/serial/8250/8250_bcm7271.c
> @@ -671,7 +671,9 @@ static void init_real_clk_rates(struct device *dev, struct brcmuart_priv *priv)
> priv->real_rates[x] = clk_get_rate(priv->baud_mux_clk);
> }
> }
> - clk_set_rate(priv->baud_mux_clk, priv->default_mux_rate);
> + rc = clk_set_rate(priv->baud_mux_clk, priv->default_mux_rate);
> + if (rc)
> + dev_err(dev, "Error restoring default BAUD MUX clock\n");

What does this actually help with? Shouldn't you recover properly from
an error instead of telling the user something bad went wrong and
nothing can be done?

This is not a correct fix at all, sorry.

thanks,

greg k-h