summaryrefslogtreecommitdiffstats
path: root/net/xfrm
diff options
context:
space:
mode:
authorHerbert Xu <herbert@gondor.apana.org.au>2007-11-19 18:50:17 -0800
committerDavid S. Miller <davem@davemloft.net>2008-01-28 14:53:54 -0800
commit1bf06cd2e338fd6fc29169d30eaf0df982338285 (patch)
tree351bf1c8e5063fdfb8c820470972169d398729fe /net/xfrm
parent60d5fcfb19d8a958fc563e52240cd05ec23f36c9 (diff)
[IPSEC]: Add async resume support on input
This patch adds support for async resumptions on input. To do so, the transform would return -EINPROGRESS and subsequently invoke the function xfrm_input_resume to resume processing. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/xfrm')
-rw-r--r--net/xfrm/xfrm_input.c38
1 files changed, 33 insertions, 5 deletions
diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c
index cce9d458604..96f42c1d2e8 100644
--- a/net/xfrm/xfrm_input.c
+++ b/net/xfrm/xfrm_input.c
@@ -101,8 +101,17 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
int err;
__be32 seq;
struct xfrm_state *x;
+ xfrm_address_t *daddr;
int decaps = 0;
- unsigned int daddroff = XFRM_SPI_SKB_CB(skb)->daddroff;
+ int async = 0;
+
+ /* A negative encap_type indicates async resumption. */
+ if (encap_type < 0) {
+ async = 1;
+ x = skb->sp->xvec[skb->sp->len - 1];
+ seq = XFRM_SKB_CB(skb)->seq;
+ goto resume;
+ }
/* Allocate new secpath or COW existing one. */
if (!skb->sp || atomic_read(&skb->sp->refcnt) != 1) {
@@ -116,6 +125,9 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
skb->sp = sp;
}
+ daddr = (xfrm_address_t *)(skb_network_header(skb) +
+ XFRM_SPI_SKB_CB(skb)->daddroff);
+
seq = 0;
if (!spi && (err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0)
goto drop;
@@ -124,9 +136,7 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
if (skb->sp->len == XFRM_MAX_DEPTH)
goto drop;
- x = xfrm_state_lookup((xfrm_address_t *)
- (skb_network_header(skb) + daddroff),
- spi, nexthdr, AF_INET);
+ x = xfrm_state_lookup(daddr, spi, nexthdr, AF_INET);
if (x == NULL)
goto drop;
@@ -147,8 +157,14 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
spin_unlock(&x->lock);
+ XFRM_SKB_CB(skb)->seq = seq;
+
nexthdr = x->type->input(x, skb);
+ if (nexthdr == -EINPROGRESS)
+ return 0;
+
+resume:
spin_lock(&x->lock);
if (nexthdr <= 0) {
if (nexthdr == -EBADMSG)
@@ -177,6 +193,12 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
break;
}
+ /*
+ * We need the inner address. However, we only get here for
+ * transport mode so the outer address is identical.
+ */
+ daddr = &x->id.daddr;
+
err = xfrm_parse_spi(skb, nexthdr, &spi, &seq);
if (err < 0)
goto drop;
@@ -190,7 +212,7 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
netif_rx(skb);
return 0;
} else {
- return x->inner_mode->afinfo->transport_finish(skb, 0);
+ return x->inner_mode->afinfo->transport_finish(skb, async);
}
drop_unlock:
@@ -201,6 +223,12 @@ drop:
}
EXPORT_SYMBOL(xfrm_input);
+int xfrm_input_resume(struct sk_buff *skb, int nexthdr)
+{
+ return xfrm_input(skb, nexthdr, 0, -1);
+}
+EXPORT_SYMBOL(xfrm_input_resume);
+
void __init xfrm_input_init(void)
{
secpath_cachep = kmem_cache_create("secpath_cache",