Fw: [Bug 51901] New: mac80211/cfg.c:1995: possible bad call to memcpy ?

Fw: [Bug 51901] New: mac80211/cfg.c:1995: possible bad call to memcpy ?

[Stephen Hemminger]

Begin forwarded message:

Date: Sat, 22 Dec 2012 16:28:47 +0000 (UTC)
From: bugzilla-daemon@bugzilla.kernel.org
To: shemminger@linux-foundation.org
Subject: [Bug 51901] New: mac80211/cfg.c:1995: possible bad call to memcpy ?


https://bugzilla.kernel.org/show_bug.cgi?id=51901

           Summary: mac80211/cfg.c:1995: possible bad call to memcpy ?
           Product: Networking
           Version: 2.5
    Kernel Version: 3.8-rc1
          Platform: All
        OS/Version: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: Other
        AssignedTo: shemminger@linux-foundation.org
        ReportedBy: dcb314@hotmail.com
        Regression: Yes


The source code is 

static int ieee80211_set_mcast_rate(struct wiphy *wiphy, struct net_device
*dev,
                    int rate[IEEE80211_NUM_BANDS])
{
    struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev);

    memcpy(sdata->vif.bss_conf.mcast_rate, rate, sizeof(rate));

The compiler says

cfg.c: In function ‘int ieee80211_set_mcast_rate(wiphy*, net_device*, int*)’:
cfg.c:1995: warning: argument to ‘sizeof’ in ‘void* memcpy(void*, const void*,
size_t)’ call is the same expression as the source; did you mean to dereference
it? [-Wsizeof-pointer-memaccess]
     memcpy(rate2, rate, sizeof(rate));
                               ^

Here is static analyser cppcheck also finding the same problem

[linux-3.8-rc1/net/mac80211/cfg.c:1995]: (error) Using 'sizeof' on array given
as function argument returns size of a pointer.


Suggest code rework.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

Re: Fw: [Bug 51901] New: mac80211/cfg.c:1995: possible bad call to memcpy ?

[Johannes Berg]

On Sat, 2012-12-22 at 13:55 -0800, Stephen Hemminger wrote:

> static int ieee80211_set_mcast_rate(struct wiphy *wiphy, struct net_device
> *dev,
>                     int rate[IEEE80211_NUM_BANDS])
> {
>     struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev);
> 
>     memcpy(sdata->vif.bss_conf.mcast_rate, rate, sizeof(rate));

Seems fine to me. If anyone's bothered by the warning, patch welcome,
but the code is ok since it's an array.

johannes

Re: Fw: [Bug 51901] New: mac80211/cfg.c:1995: possible bad call to memcpy ?

[Felix Fietkau]

On 2012-12-27 8:45 AM, Johannes Berg wrote:
> On Sat, 2012-12-22 at 13:55 -0800, Stephen Hemminger wrote:
> 
>> static int ieee80211_set_mcast_rate(struct wiphy *wiphy, struct net_device
>> *dev,
>>                     int rate[IEEE80211_NUM_BANDS])
>> {
>>     struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev);
>> 
>>     memcpy(sdata->vif.bss_conf.mcast_rate, rate, sizeof(rate));
> 
> Seems fine to me. If anyone's bothered by the warning, patch welcome,
> but the code is ok since it's an array.
I think the warning is correct and the code needs to be fixed. Since
rate is a function parameter, sizeof(rate) == sizeof(int).
I didn't know about this weird aspect of the C standard either, but I
verified it with a simple user space test program. ;)

- Felix

Re: Fw: [Bug 51901] New: mac80211/cfg.c:1995: possible bad call to memcpy ?

[Johannes Berg]

On Thu, 2012-12-27 at 09:30 +0100, Felix Fietkau wrote:
> On 2012-12-27 8:45 AM, Johannes Berg wrote:
> > On Sat, 2012-12-22 at 13:55 -0800, Stephen Hemminger wrote:
> > 
> >> static int ieee80211_set_mcast_rate(struct wiphy *wiphy, struct net_device
> >> *dev,
> >>                     int rate[IEEE80211_NUM_BANDS])
> >> {
> >>     struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev);
> >> 
> >>     memcpy(sdata->vif.bss_conf.mcast_rate, rate, sizeof(rate));
> > 
> > Seems fine to me. If anyone's bothered by the warning, patch welcome,
> > but the code is ok since it's an array.
> I think the warning is correct and the code needs to be fixed. Since
> rate is a function parameter, sizeof(rate) == sizeof(int).
> I didn't know about this weird aspect of the C standard either, but I
> verified it with a simple user space test program. ;)

Yeah you're right, I even tried to test it but got it wrong. It should
be changed.

johannes