On Sat, Aug 06, 2011 at 07:27:48AM -0700, Dan Carpenter wrote:
> There were some curly braces missing so the original code in
> wpa_supplicant_ioctl() pretty much always returned -EFAULT without
> doing anything.
>
> Signed-off-by: Dan Carpenter <error27@xxxxxxxxx>
>
> diff --git a/drivers/staging/rtl8712/rtl871x_ioctl_linux.c b/drivers/staging/rtl8712/rtl871x_ioctl_linux.c
> index 40e6b5c..c380a30 100644
> --- a/drivers/staging/rtl8712/rtl871x_ioctl_linux.c
> +++ b/drivers/staging/rtl8712/rtl871x_ioctl_linux.c
> @@ -2072,9 +2072,10 @@ static int wpa_supplicant_ioctl(struct net_device *dev, struct iw_point *p)
> param = (struct ieee_param *)_malloc(p->length);
> if (param == NULL)
> return -ENOMEM;
> - if (copy_from_user(param, p->pointer, p->length))
> + if (copy_from_user(param, p->pointer, p->length)) {
> kfree((u8 *)param);
> return -EFAULT;
> + }
Ouch! (Now emacs would've caught this! ;-))
My reading of the code suggests that (not surprisingly) this path does
not get executed. The frequent case is to go thru wext-core's
"standard" handler, instead of the above. If so, then that'd explain
why I've not noticed this bug during testing.
Good catch.
thanks,
ali
> switch (param->cmd) {
> case IEEE_CMD_SET_WPA_PARAM:
> ret = wpa_set_param(dev, param->u.wpa_param.name,
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
