Re: [PATCH] PCI: amlogic: The check for devm_add_action_or_reset's return value added

[Christophe JAILLET]

Le 20/12/2022 à 10:21, Aleksandr Burakov a écrit :
> The return value of the function devm_add_action_or_reset() was not
> checked hence an error code would not be returned.
>
> Found by Linux Verification Center (linuxtesting.org) with SVACE.
>
> Fixes: 9c0ef6d34fdb ("PCI: amlogic: Add the Amlogic Meson PCIe controller driver")
> Signed-off-by: Aleksandr Burakov <a.burakov@xxxxxxxxxxxx>
> ---
>   drivers/pci/controller/dwc/pci-meson.c | 6 +++++-
>   1 file changed, 5 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/pci/controller/dwc/pci-meson.c b/drivers/pci/controller/dwc/pci-meson.c
> index c1527693bed9..3d82f0b65480 100644
> --- a/drivers/pci/controller/dwc/pci-meson.c
> +++ b/drivers/pci/controller/dwc/pci-meson.c
> @@ -187,9 +187,13 @@ static inline struct clk *meson_pcie_probe_clock(struct device *dev,
>   		return ERR_PTR(ret);
>   	}
>   
> -	devm_add_action_or_reset(dev,
> +	ret = devm_add_action_or_reset(dev,
>   				 (void (*) (void *))clk_disable_unprepare,

Hi,

using devm_clk_get_enabled() a few lines above would be IMHO much more 
elegant than these casts.

This also saves a few LoC and avoid some other warnings/issues. (see [1])

Such an approach has already been used in [2].

Just my 2c,

CJ

[1]: https://lore.kernel.org/all/20221118233101.never.215-kees@xxxxxxxxxx/
[2]: https://lore.kernel.org/all/20221202184525.gonna.423-kees@xxxxxxxxxx/

>   				 clk);
> +	if (ret) {
> +		dev_err(dev, "couldn't reset clk\n");
> +		return ERR_PTR(ret);
> +	}
>   
>   	return clk;
>   }