linux-kernel - Re: [PATCH] Staging: rtlwifi: Remove NULL pointer dereference

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20171012021647.GB30753@eros>
Date:   Thu, 12 Oct 2017 13:16:47 +1100
From:   "Tobin C. Harding" <me@...in.cc>
To:     Shreeya Patel <shreeya.patel23498@...il.com>
Cc:     gregkh@...uxfoundation.org, devel@...verdev.osuosl.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] Staging: rtlwifi: Remove NULL pointer dereference

On Wed, Oct 11, 2017 at 06:02:47PM +0530, Shreeya Patel wrote:
> On Tue, 2017-10-10 at 11:06 +1100, Tobin C. Harding wrote:
> > On Tue, Oct 10, 2017 at 02:48:58AM +0530, Shreeya Patel wrote:
> > > 
> > > Remove NULL pointer dereference as it results in undefined
> > > behaviour, and will usually lead to a runtime error.
> > The diff does not show any pointer dereference so it is hard to
> > understand what you are trying to do
> > with this patch.
> > 
> > > 
> > > Signed-off-by: Shreeya Patel <shreeya.patel23498@...il.com>
> > > ---
> > >  drivers/staging/rtlwifi/base.c | 2 +-
> > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/staging/rtlwifi/base.c
> > > b/drivers/staging/rtlwifi/base.c
> > > index b88b0e8..5bb8f98 100644
> > > --- a/drivers/staging/rtlwifi/base.c
> > > +++ b/drivers/staging/rtlwifi/base.c
> > > @@ -781,7 +781,7 @@ static void _rtl_txrate_selectmode(struct
> > > ieee80211_hw *hw,
> > >  
> > >  	struct rtl_priv *rtlpriv = rtl_priv(hw);
> > >  	struct rtl_mac *mac = rtl_mac(rtl_priv(hw));
> > > -	struct rtl_sta_info *sta_entry = NULL;
> > > +	struct rtl_sta_info *sta_entry;
> > Now the pointer just has garbage in it instead of the testable value
> > of NULL. If you are concerned
> > with the dereference perhaps you could add a NULL check, again it's
> > hard to say without seeing the
> > code.
> 
> Hello, 
> 
> Thanks for making me understand. 
> 
> Here is the code after declaration and initialization of sta_entry. 
> Will it be good to add a NULL check in this case? 
> 
> struct rtl_sta_info *sta_entry = NULL;
> 	u8 ratr_index = SET_RATE_ID(RATR_INX_WIRELESS_MC);
> 
> 	if (sta) {
> 		sta_entry = (struct rtl_sta_info *)sta->drv_priv;
> 		ratr_index = sta_entry->ratr_index;
> 	}

Later in this function the macro SET_RATE_ID() is called, it relies on sta_entry being NULL if it
was not explicitly set.

Here is the macro;

#define SET_RATE_ID(rate_id)					\
	((rtlpriv->cfg->spec_ver & RTL_SPEC_NEW_RATEID) ?	\
		rtl_mrate_idx_to_arfr_id(hw, rate_id,		\
			(sta_entry ? sta_entry->wireless_mode :	\
			 WIRELESS_MODE_G)) :			\
		rate_id)

> If we are making a pointer point to NULL then what if any other
> variable is already pointing to NULL for some other purpose.
> Instead, removing initialization will be good right?

A pointer does not _point_ to NULL as such. A NULL pointer has a value of all zero bytes. Have you
read (and completed all the exercises) in KnR

https://en.wikipedia.org/wiki/The_C_Programming_Language

It is, in my opinion, one of the best tech books ever written. If you have any holes in your C
knowledge, this is the place to start.

Good luck,
Tobin.