lkml.org 
[lkml]   [2005]   [Jul]   [22]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: Is this a bug in linux-2.6.12 ipsec code function xfrm4_rcv_encap ??
I AM SORRY FOR THE PREVIOUS MAIL.
I am correcting my previous mail.
Infact I see only One race(not three as was wrongly pointed out).
I commented out the section once again where the race might be.


On 7/23/05, k8 s <uint32@gmail.com> wrote:
> Hello,
> I see a possible race in linux-2.6.12 ipsec code function xfrm4_rcv_encap.
> I want to double check with the group.
> The issue is with SMP(mostly) or Preemptible Kernels.
> The race comes when someone flushes the SA's
> (setkey -Fexecuting on another processor )
> while xfrm_rcv_encap is executing one processor.
>
> Below is the function code.
> I am putting comments in the code where probably the race comes.
> correct me if I am wrong.
>
> int xfrm4_rcv_encap(struct sk_buff *skb, __u16 encap_type)
> {
> int err;
> u32 spi, seq;
> struct sec_decap_state xfrm_vec[XFRM_MAX_DEPTH];
> struct xfrm_state *x;
> int xfrm_nr = 0;
> int decaps = 0;
>
> if ((err = xfrm4_parse_spi(skb, skb->nh.iph->protocol, &spi, &seq)) != 0)
> goto drop;
>
> do {
> struct iphdr *iph = skb->nh.iph;
>
> if (xfrm_nr == XFRM_MAX_DEPTH)
> goto drop;
>
> x = xfrm_state_lookup((xfrm_address_t *)&iph->daddr, spi,
> iph->protocol, AF_INET);
>

> if (x == NULL)
> goto drop;
>
> spin_lock(&x->lock);
> if (unlikely(x->km.state != XFRM_STATE_VALID))
> goto drop_unlock;
>
> if (x->props.replay_window && xfrm_replay_check(x, seq))
> goto drop_unlock;
>
> if (xfrm_state_check_expire(x))
> goto drop_unlock;
>
> xfrm_vec[xfrm_nr].decap.decap_type = encap_type;
> if (x->type->input(x, &(xfrm_vec[xfrm_nr].decap), skb))
> goto drop_unlock;
>
> /* only the first xfrm gets the encap type */
> encap_type = 0;
>
> if (x->props.replay_window)
> xfrm_replay_advance(x, seq);
>
> x->curlft.bytes += skb->len;
> x->curlft.packets++;
>
> spin_unlock(&x->lock);
>
> xfrm_vec[xfrm_nr++].xvec = x;
>
> iph = skb->nh.iph;
>
/********************************************************
Race Here . The Check(x->props.mode) is without Lock. What if setkey
-F is done at the same time on another processor freeing what x points
to.
********************************************************/
> if (x->props.mode) {
> if (iph->protocol != IPPROTO_IPIP)
> goto drop;
> if (!pskb_may_pull(skb, sizeof(struct iphdr)))
> goto drop;
> if (skb_cloned(skb) &&
> pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
> goto drop;
> if (x->props.flags & XFRM_STATE_DECAP_DSCP)
> ipv4_copy_dscp(iph, skb->h.ipiph);
> if (!(x->props.flags & XFRM_STATE_NOECN))
> ipip_ecn_decapsulate(skb);
> skb->mac.raw = memmove(skb->data - skb->mac_len,
> skb->mac.raw, skb->mac_len);
> skb->nh.raw = skb->data;
> memset(&(IPCB(skb)->opt), 0, sizeof(struct ip_options));
> decaps = 1;
> break;
> }
>
> if ((err = xfrm_parse_spi(skb, skb->nh.iph->protocol, &spi, &seq)) < 0)
> goto drop;
> } while (!err);
>
> /* Allocate new secpath or COW existing one. */
>
> if (!skb->sp || atomic_read(&skb->sp->refcnt) != 1) {
> struct sec_path *sp;
> sp = secpath_dup(skb->sp);
> if (!sp)
> goto drop;
> if (skb->sp)
> secpath_put(skb->sp);
> skb->sp = sp;
> }
> if (xfrm_nr + skb->sp->len > XFRM_MAX_DEPTH)
> goto drop;
>
> memcpy(skb->sp->x+skb->sp->len, xfrm_vec, xfrm_nr*sizeof(struct
> sec_decap_state));
> skb->sp->len += xfrm_nr;
>
> if (decaps) {
> if (!(skb->dev->flags&IFF_LOOPBACK)) {
> dst_release(skb->dst);
> skb->dst = NULL;
> }
> netif_rx(skb);
> return 0;
> } else {
> return -skb->nh.iph->protocol;
> }
>
> drop_unlock:
> spin_unlock(&x->lock);
> xfrm_state_put(x);
> drop:
> while (--xfrm_nr >= 0)
> xfrm_state_put(xfrm_vec[xfrm_nr].xvec);
>
> kfree_skb(skb);
> return 0;
> }
>
>
>
>
>
> I am just guessing.
> If I am wrong I request anyone to give me a reason why it is not a bug ?
> I haven't checked the IPv6 front and the IPSec outbound side.
> Once this proves to be a bug I will check them.
>
> S.Kartikeyan
>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-07-23 03:36    [W:0.033 / U:4.380 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site