linux-wireless.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* static analysis issue in rtl8188de driver
@ 2019-05-04 21:50 Colin Ian King
  2019-05-06  7:56 ` Pkshih
  0 siblings, 1 reply; 2+ messages in thread
From: Colin Ian King @ 2019-05-04 21:50 UTC (permalink / raw)
  To: Ping-Ke Shih, Kalle Valo, Larry Finger, linux-wireless
  Cc: linux-kernel@vger.kernel.org

Hi,

Static analysis with Coverity has found an issue in the rtl8188de
wireless driver in drivers/net/wireless/realtek/rtlwifi/rtl8192de/dm.c
in function tl92d_dm_txpower_tracking_callback_thermalmeter.

The issue is that u8 array ofdm_index[3] is never initialized, however
it is decremented and incremented in two places resulting in garbage
value from the stack being updated in the following code:

	if (thermalvalue > rtlpriv->dm.thermalvalue) {
        	for (i = 0; i < rf; i++)
                	ofdm_index[i] -= delta;
        	cck_index -= delta;
	} else {
        	for (i = 0; i < rf; i++)
                	ofdm_index[i] += index;
        	cck_index += index;	
	}

At my first look at the code I believe ofdm_index should be just
zero-initialized at declaration time, but I suspect that I'm overlooking
something maybe a bit deeper. Any ideas?

Colin


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: static analysis issue in rtl8188de driver
  2019-05-04 21:50 static analysis issue in rtl8188de driver Colin Ian King
@ 2019-05-06  7:56 ` Pkshih
  0 siblings, 0 replies; 2+ messages in thread
From: Pkshih @ 2019-05-06  7:56 UTC (permalink / raw)
  To: linux-wireless@vger.kernel.org, colin.king@canonical.com,
	kvalo@codeaurora.org, Larry.Finger@lwfinger.net
  Cc: linux-kernel@vger.kernel.org

On Sat, 2019-05-04 at 21:50 +0000, Colin Ian King wrote:
> Hi,
> 
> Static analysis with Coverity has found an issue in the rtl8188de
> wireless driver in drivers/net/wireless/realtek/rtlwifi/rtl8192de/dm.c
> in function tl92d_dm_txpower_tracking_callback_thermalmeter.
> 
> The issue is that u8 array ofdm_index[3] is never initialized, however
> it is decremented and incremented in two places resulting in garbage
> value from the stack being updated in the following code:
> 
> 	if (thermalvalue > rtlpriv->dm.thermalvalue) {
>         	for (i = 0; i < rf; i++)
>                 	ofdm_index[i] -= delta;
>         	cck_index -= delta;
> 	} else {
>         	for (i = 0; i < rf; i++)
>                 	ofdm_index[i] += index;
>         	cck_index += index;	
> 	}
> 
> At my first look at the code I believe ofdm_index should be just
> zero-initialized at declaration time, but I suspect that I'm overlooking
> something maybe a bit deeper. Any ideas?
> 


Hi Colin,

Thanks for your report.

After my quick look, there are at least two obvious problems.
One is array size of ofdm_index[] should be two instead. Another is the value
of ofdm_index[] should be obtained from rtlpriv->dm.ofdm_index[]. 
Since the logic is quite complex, I need some time to fix it.

PK


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2019-05-06  7:56 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-05-04 21:50 static analysis issue in rtl8188de driver Colin Ian King
2019-05-06  7:56 ` Pkshih

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).