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  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:   Fri, 10 May 2019 09:06:19 -0700
From:   Jakub Kicinski <jakub.kicinski@...ronome.com>
To:     John Fastabend <john.fastabend@...il.com>
Cc:     davem@...emloft.net, netdev@...r.kernel.org,
        oss-drivers@...ronome.com, Dave Watson <davejwatson@...com>,
        Dirk van der Merwe <dirk.vandermerwe@...ronome.com>
Subject: Re: [PATCH net 2/2] net/tls: handle errors from padding_length()

On Fri, 10 May 2019 07:16:44 -0700, John Fastabend wrote:
> Jakub Kicinski wrote:
> > At the time padding_length() is called the record header
> > is still part of the message.  If malicious TLS 1.3 peer
> > sends an all-zero record padding_length() will stop at
> > the record header, and return full length of the data
> > including the tail_size.
> > 
> > Subsequent subtraction of prot->overhead_size from rxm->full_len
> > will cause rxm->full_len to turn negative.  skb accessors,
> > however, will always catch resulting out-of-bounds operation,
> > so in practice this fix comes down to returning the correct
> > error code.  It also fixes a set but not used warning.  
> 
> In practice returning incorrect error codes to users can confuse
> applications though so this seems important. I've observed apps
> hang with wrong codes for example and this error seems to be pushed
> all the way up to sw_recvmsg.
> 
> > 
> > This code was added by commit 130b392c6cd6 ("net: tls: Add tls 1.3 support").
> > 
> > CC: Dave Watson <davejwatson@...com>
> > Signed-off-by: Jakub Kicinski <jakub.kicinski@...ronome.com>
> > Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@...ronome.com>
> > ---  
> 
> Looks good to me but one question below,

Thanks for the reviews!

> >  	/* After using skb->sk to propagate sk through crypto async callback
> > @@ -1478,7 +1488,7 @@ static int decrypt_skb_update(struct sock *sk, struct sk_buff *skb,
> >  	struct tls_prot_info *prot = &tls_ctx->prot_info;
> >  	int version = prot->version;
> >  	struct strp_msg *rxm = strp_msg(skb);
> > -	int err = 0;
> > +	int pad, err = 0;
> >  
> >  	if (!ctx->decrypted) {
> >  #ifdef CONFIG_TLS_DEVICE
> > @@ -1501,7 +1511,11 @@ static int decrypt_skb_update(struct sock *sk, struct sk_buff *skb,
> >  			*zc = false;
> >  		}
> >  
> > -		rxm->full_len -= padding_length(ctx, tls_ctx, skb);
> > +		pad = padding_length(ctx, prot, skb);
> > +		if (pad < 0)
> > +			return pad;
> > +  
> 
> Need to review a bit closer on my side, but do we need to do any cleanup
> if this fails? It looks like the other padding_length call sites will
> but here we eventually return directly to recvmsg.

Please double check my thinking, but not as far as I could tell.  
At this point we must have a decrypted frame, meaning no async
operation can be in progress on this skb.  And since the error 
will kill the connection, we don't have to worry about advancing
the record sequence number etc.

> > +		rxm->full_len -= pad;
> >  		rxm->offset += prot->prepend_size;
> >  		rxm->full_len -= prot->overhead_size;
> >  		tls_advance_record_sn(sk, &tls_ctx->rx, version);
> > -- 
> > 2.21.0
> >   

Powered by blists - more mailing lists