Re: [PATCH] [v2] wilc1000: Fix memleak in wilc_sdio_probe

[Claudiu.Beznea]

On 20.08.2020 08:48, Dinghao Liu wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> 
> When devm_clk_get() returns -EPROBE_DEFER, sdio_priv
> should be freed just like when wilc_cfg80211_init()
> fails.
> 
> Fixes: 8692b047e86cf ("staging: wilc1000: look for rtc_clk clock")
> Signed-off-by: Dinghao Liu <dinghao.liu@xxxxxxxxxx>
> ---
> 
> Changelog:
> 
> v2: - Remove 'staging' prefix in subject.
> ---
>  drivers/net/wireless/microchip/wilc1000/sdio.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/wireless/microchip/wilc1000/sdio.c b/drivers/net/wireless/microchip/wilc1000/sdio.c
> index 3ece7b0b0392..351ff909ab1c 100644
> --- a/drivers/net/wireless/microchip/wilc1000/sdio.c
> +++ b/drivers/net/wireless/microchip/wilc1000/sdio.c
> @@ -149,9 +149,10 @@ static int wilc_sdio_probe(struct sdio_func *func,
>         wilc->dev = &func->dev;
> 
>         wilc->rtc_clk = devm_clk_get(&func->card->dev, "rtc");
> -       if (PTR_ERR_OR_ZERO(wilc->rtc_clk) == -EPROBE_DEFER)
> +       if (PTR_ERR_OR_ZERO(wilc->rtc_clk) == -EPROBE_DEFER) {
> +               kfree(sdio_priv);

The proper way to free resources here would be to call
wilc_netdev_cleanup() which will free all objects allocated by
wilc_cfg80211_init() and will also free sdio_priv. I personally would go
further and remove the kfree(wilc->bus_data) from wilc_netdev_cleanup() and
keep it in the wilc_sdio_probe(), wilc_spi_probe() where it was actually
allocated.

>                 return -EPROBE_DEFER;
> -       else if (!IS_ERR(wilc->rtc_clk))
> +       } else if (!IS_ERR(wilc->rtc_clk))
>                 clk_prepare_enable(wilc->rtc_clk);
> 
>         dev_info(&func->dev, "Driver Initializing success\n");
> --
> 2.17.1
>