Re: [PATCH V2] rtlwifi: Return correct failure code on error

[Larry Finger <Larry.Finger@lwfinger.net>]

On 02/08/2012 03:14 PM, Simon Graham wrote:
> Callers of rtl_pci_init expect zero to be returned on error. Returning
> the error code leads to the data being used when it shouldn't causing,
> amongst other things, divide by zero panics attempting to use the
> ring size that is set to zero.
>
> Signed-off-by: Simon Graham<simon.graham@virtualcomputer.com>
> ---
>   drivers/net/wireless/rtlwifi/pci.c |    2 +-
>   1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/net/wireless/rtlwifi/pci.c b/drivers/net/wireless/rtlwifi/pci.c
> index fb84707..f0ce366 100644
> --- a/drivers/net/wireless/rtlwifi/pci.c
> +++ b/drivers/net/wireless/rtlwifi/pci.c
> @@ -1492,7 +1492,7 @@ static int rtl_pci_init(struct ieee80211_hw *hw, struct pci_dev *pdev)
>   	if (err) {
>   		RT_TRACE(rtlpriv, COMP_ERR, DBG_EMERG,
>   			 "tx ring initialization failed\n");
> -		return err;
> +		return 0;
>   	}
>
>   	return 1;

You must have an extreme memory limitation to find all these problems.

As long as you are reworking this section, I would prefer you leave the "return 
err" as it is now, change the "return 1" to "return 0", and remove the not 
operator in "err = !rtl_pci_init(hw, pdev);" call of this routine. I just 
discovered it a few days ago, and it bothers me, but I've been too busy to 
submit a fix. In addition, these changes will convert rtl_pci_init() into a more 
normal routine that returns 0 if OK and non-zero on errors.

Larry