Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Sat, 23 Jun 2018 11:52:56 +0900
From:   Stafford Horne <shorne@gmail.com>
To:     Eric Biggers <ebiggers3@gmail.com>
Cc:     LKML <linux-kernel@vger.kernel.org>,
        Greg KH <gregkh@linuxfoundation.org>, arnd@arndb.de,
        linux-crypto@vger.kernel.org,
        Herbert Xu <herbert@gondor.apana.org.au>,
        "David S. Miller" <davem@davemloft.net>,
        Nick Desaulniers <nick.desaulniers@gmail.com>
Subject: Re: [RFC PATCH 1/2] crypto: Fix -Wstringop-truncation warnings
Message-ID: <20180623025256.GG24595@lianli.shorne-pla.net>
References: <20180623020753.27266-1-shorne@gmail.com>
 <20180623020753.27266-2-shorne@gmail.com>
 <20180623024149.GB880@sol.localdomain>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <20180623024149.GB880@sol.localdomain>
User-Agent: Mutt/1.9.5 (2018-04-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Jun 22, 2018 at 07:41:49PM -0700, Eric Biggers wrote:
> On Sat, Jun 23, 2018 at 11:07:52AM +0900, Stafford Horne wrote:
> > As of GCC 9.0.0 the build is reporting warnings like:
> > 
> >     crypto/ablkcipher.c: In function ‘crypto_ablkcipher_report’:
> >     crypto/ablkcipher.c:374:2: warning: ‘strncpy’ specified bound 64 equals destination size [-Wstringop-truncation]
> >       strncpy(rblkcipher.geniv, alg->cra_ablkcipher.geniv ?: "<default>",
> >       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> >        sizeof(rblkcipher.geniv));
> >        ~~~~~~~~~~~~~~~~~~~~~~~~~
> > 
> > This means the strnycpy might create a non null terminated string.  Fix this by
> > limiting the size of the string copy to include the null terminator.
> > 
> > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > Cc: Arnd Bergmann <arnd@arndb.de>
> > Signed-off-by: Stafford Horne <shorne@gmail.com>
> > ---
> >  crypto/ablkcipher.c | 4 ++--
> >  crypto/blkcipher.c  | 2 +-
> >  2 files changed, 3 insertions(+), 3 deletions(-)
> > 
> > diff --git a/crypto/ablkcipher.c b/crypto/ablkcipher.c
> > index d880a4897159..972cd7c879f6 100644
> > --- a/crypto/ablkcipher.c
> > +++ b/crypto/ablkcipher.c
> > @@ -372,7 +372,7 @@ static int crypto_ablkcipher_report(struct sk_buff *skb, struct crypto_alg *alg)
> >  
> >  	strncpy(rblkcipher.type, "ablkcipher", sizeof(rblkcipher.type));
> >  	strncpy(rblkcipher.geniv, alg->cra_ablkcipher.geniv ?: "<default>",
> > -		sizeof(rblkcipher.geniv));
> > +		sizeof(rblkcipher.geniv) - 1);
> >  
> >  	rblkcipher.blocksize = alg->cra_blocksize;
> >  	rblkcipher.min_keysize = alg->cra_ablkcipher.min_keysize;
> > @@ -446,7 +446,7 @@ static int crypto_givcipher_report(struct sk_buff *skb, struct crypto_alg *alg)
> >  
> >  	strncpy(rblkcipher.type, "givcipher", sizeof(rblkcipher.type));
> >  	strncpy(rblkcipher.geniv, alg->cra_ablkcipher.geniv ?: "<built-in>",
> > -		sizeof(rblkcipher.geniv));
> > +		sizeof(rblkcipher.geniv) - 1);
> >  
> >  	rblkcipher.blocksize = alg->cra_blocksize;
> >  	rblkcipher.min_keysize = alg->cra_ablkcipher.min_keysize;
> > diff --git a/crypto/blkcipher.c b/crypto/blkcipher.c
> > index 01c0d4aa2563..f1644b5cf68c 100644
> > --- a/crypto/blkcipher.c
> > +++ b/crypto/blkcipher.c
> > @@ -511,7 +511,7 @@ static int crypto_blkcipher_report(struct sk_buff *skb, struct crypto_alg *alg)
> >  
> >  	strncpy(rblkcipher.type, "blkcipher", sizeof(rblkcipher.type));
> >  	strncpy(rblkcipher.geniv, alg->cra_blkcipher.geniv ?: "<default>",
> > -		sizeof(rblkcipher.geniv));
> > +		sizeof(rblkcipher.geniv) - 1);
> >  
> >  	rblkcipher.blocksize = alg->cra_blocksize;
> >  	rblkcipher.min_keysize = alg->cra_blkcipher.min_keysize;
> 
> Your "fix" introduces an information disclosure bug, as it results in
> uninitialized memory being copied to userspace.  This same broken patch was sent
> by someone else too.
> 
> Maybe it would be best to just memset() the crypto_report_* structs to 0 after
> declaration and then replace the strncpy()'s with strscpy()'s, even if just to
> stop people from sending broken "fixes".  Do you want to do that?

Right, I didnt realize that we were using strncpy to also init the whole buffer.

I will do as suggest, and respic.

-Stafford