netdev - Re: [PATCH net] net/sched: cls_flower: fix only mask bit check in the validate_ct

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] [day] [month] [year] [list]

Message-ID: <YFDnCeQOrOKiQdV9@horizon.localdomain>
Date:   Tue, 16 Mar 2021 14:12:41 -0300
From:   Marcelo Ricardo Leitner <mleitner@...hat.com>
To:     wenxu@...oud.cn
Cc:     kuba@...nel.org, netdev@...r.kernel.org, jhs@...atatu.com,
        davem@...emloft.net
Subject: Re: [PATCH net] net/sched: cls_flower: fix only mask bit check in
 the validate_ct_state

Hi,

On Tue, Mar 16, 2021 at 03:44:17PM +0800, wenxu@...oud.cn wrote:
> From: wenxu <wenxu@...oud.cn>
> 
> The ct_state validate should not only check the mask bit and also
> check the state bit.
> For the +new+est case example, The 'new' and 'est' bits should be
> set in both state_mask and state flags. Or the -new-est case also
> will be reject by kernel.

Please mention why +trk-new-est is expected.

> 
> Fixes: 	1bcc51ac0731 ("net/sched: cls_flower: Reject invalid ct_state flags rules")
> Fixes: 	3aed8b63336c ("net/sched: cls_flower: validate ct_state for invalid and reply flags")

checkpatch.pl doesn't complain but I'm not sure if a tab is allowed here, btw.

> Signed-off-by: wenxu <wenxu@...oud.cn>
> ---
>  net/sched/cls_flower.c | 20 +++++++++++++-------
>  1 file changed, 13 insertions(+), 7 deletions(-)
> 
> diff --git a/net/sched/cls_flower.c b/net/sched/cls_flower.c
> index d097b5c..92659e1 100644
> --- a/net/sched/cls_flower.c
> +++ b/net/sched/cls_flower.c
> @@ -1401,31 +1401,37 @@ static int fl_set_enc_opt(struct nlattr **tb, struct fl_flow_key *key,
>  	return 0;
>  }
>  
> -static int fl_validate_ct_state(u16 state, struct nlattr *tb,
> +static int fl_validate_ct_state(u16 state_mask, u16 state,
> +				struct nlattr *tb,

The key/mask ordering is becoming messy in flower.
As this function gets called from fl_set_key_ct, please lets keep what was used
there: key, mask. Seems it's still the dominant one.
  static int fl_set_key_ct(struct nlattr **tb,
                           struct flow_dissector_key_ct *key,
                           struct flow_dissector_key_ct *mask,

On a similar note, I'm wondering if it worth just doing:
	u16 effective = state & state_mask;
To avoid this many checks below against key and mask simultaneously.

>  				struct netlink_ext_ack *extack)
>  {
> -	if (state && !(state & TCA_FLOWER_KEY_CT_FLAGS_TRACKED)) {
> +	if (state_mask && !(state_mask & TCA_FLOWER_KEY_CT_FLAGS_TRACKED)) {
>  		NL_SET_ERR_MSG_ATTR(extack, tb,
>  				    "no trk, so no other flag can be set");
>  		return -EINVAL;
>  	}
>  
> -	if (state & TCA_FLOWER_KEY_CT_FLAGS_NEW &&
> +	if (state_mask & TCA_FLOWER_KEY_CT_FLAGS_NEW &&
> +	    state & TCA_FLOWER_KEY_CT_FLAGS_NEW &&
> +	    state_mask & TCA_FLOWER_KEY_CT_FLAGS_ESTABLISHED &&
>  	    state & TCA_FLOWER_KEY_CT_FLAGS_ESTABLISHED) {
>  		NL_SET_ERR_MSG_ATTR(extack, tb,
>  				    "new and est are mutually exclusive");
>  		return -EINVAL;
>  	}
>  
> -	if (state & TCA_FLOWER_KEY_CT_FLAGS_INVALID &&
> -	    state & ~(TCA_FLOWER_KEY_CT_FLAGS_TRACKED |
> +	if (state_mask & TCA_FLOWER_KEY_CT_FLAGS_INVALID &&
> +	    state & TCA_FLOWER_KEY_CT_FLAGS_INVALID &&
> +	    state_mask & ~(TCA_FLOWER_KEY_CT_FLAGS_TRACKED |
>  		      TCA_FLOWER_KEY_CT_FLAGS_INVALID)) {

An indent adjust here is welcomed.

Thanks,
Marcelo

>  		NL_SET_ERR_MSG_ATTR(extack, tb,
>  				    "when inv is set, only trk may be set");
>  		return -EINVAL;
>  	}
>  
> -	if (state & TCA_FLOWER_KEY_CT_FLAGS_NEW &&
> +	if (state_mask & TCA_FLOWER_KEY_CT_FLAGS_NEW &&
> +	    state & TCA_FLOWER_KEY_CT_FLAGS_NEW &&
> +	    state_mask & TCA_FLOWER_KEY_CT_FLAGS_REPLY &&
>  	    state & TCA_FLOWER_KEY_CT_FLAGS_REPLY) {
>  		NL_SET_ERR_MSG_ATTR(extack, tb,
>  				    "new and rpl are mutually exclusive");
> @@ -1451,7 +1457,7 @@ static int fl_set_key_ct(struct nlattr **tb,
>  			       &mask->ct_state, TCA_FLOWER_KEY_CT_STATE_MASK,
>  			       sizeof(key->ct_state));
>  
> -		err = fl_validate_ct_state(mask->ct_state,
> +		err = fl_validate_ct_state(mask->ct_state, key->ct_state,
>  					   tb[TCA_FLOWER_KEY_CT_STATE_MASK],
>  					   extack);
>  		if (err)
> -- 
> 1.8.3.1
>