On Fri, Dec 02, 2005 at 07:47:45AM +0100, Eric Dumazet wrote:
>
> Then maybe we could just do something like that :
>
> if (skb == skb_peek(&sk->sk_receive_queue)) {
> __skb_unlink(skb, &sk->sk_receive_queue);
> atomic_dec(&skb->users); /* drop reference */
> }
>
> This should avoid inlining kfree_skb() with 2 conditional branches...
That's a good idea. Dave, here is a patch for 2.6.16.
[IP]: Simplify and consolidate MSG_PEEK error handling
When a packet is obtained from skb_recv_datagram with MSG_PEEK enabled
it is left on the socket receive queue. This means that when we detect
a checksum error we have to be careful when trying to free the packet
as someone could have dequeued it in the time being.
Currently this delicate logic is duplicated three times between UDPv4,
UDPv6 and RAWv6. This patch moves them into a one place and simplifies
the code somewhat.
This is based on a suggestion by Eric Dumazet.
Signed-off-by: Herbert Xu <[EMAIL PROTECTED]>
Thanks,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <[EMAIL PROTECTED]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -1239,6 +1239,8 @@ extern int skb_copy_and_csum_data
int hlen,
struct iovec *iov);
extern void skb_free_datagram(struct sock *sk, struct sk_buff *skb);
+extern void skb_kill_datagram(struct sock *sk, struct sk_buff *skb,
+ unsigned int flags);
extern unsigned int skb_checksum(const struct sk_buff *skb, int offset,
int len, unsigned int csum);
extern int skb_copy_bits(const struct sk_buff *skb, int offset,
diff --git a/net/core/datagram.c b/net/core/datagram.c
--- a/net/core/datagram.c
+++ b/net/core/datagram.c
@@ -47,6 +47,7 @@
#include <linux/rtnetlink.h>
#include <linux/poll.h>
#include <linux/highmem.h>
+#include <linux/spinlock.h>
#include <net/protocol.h>
#include <linux/skbuff.h>
@@ -200,6 +201,41 @@ void skb_free_datagram(struct sock *sk,
}
/**
+ * skb_kill_datagram - Free a datagram skbuff forcibly
+ * @sk: socket
+ * @skb: datagram skbuff
+ * @flags: MSG_ flags
+ *
+ * This function frees a datagram skbuff that was received by
+ * skb_recv_datagram. The flags argument must match the one
+ * used for skb_recv_datagram.
+ *
+ * If the MSG_PEEK flag is set, and the packet is still on the
+ * receive queue of the socket, it will be taken off the queue
+ * before it is freed.
+ *
+ * This function currently only disables BH when acquiring the
+ * sk_receive_queue lock. Therefore it must not be used in a
+ * context where that lock is acquired in an IRQ context.
+ */
+
+void skb_kill_datagram(struct sock *sk, struct sk_buff *skb, unsigned int
flags)
+{
+ if (flags & MSG_PEEK) {
+ spin_lock_bh(&sk->sk_receive_queue.lock);
+ if (skb == skb_peek(&sk->sk_receive_queue)) {
+ __skb_unlink(skb, &sk->sk_receive_queue);
+ atomic_dec(&skb->users);
+ }
+ spin_unlock_bh(&sk->sk_receive_queue.lock);
+ }
+
+ kfree_skb(skb);
+}
+
+EXPORT_SYMBOL(skb_kill_datagram);
+
+/**
* skb_copy_datagram_iovec - Copy a datagram to an iovec.
* @skb: buffer to copy
* @offset: offset in the buffer to start copying from
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -846,20 +846,7 @@ out:
csum_copy_err:
UDP_INC_STATS_BH(UDP_MIB_INERRORS);
- /* Clear queue. */
- if (flags&MSG_PEEK) {
- int clear = 0;
- spin_lock_bh(&sk->sk_receive_queue.lock);
- if (skb == skb_peek(&sk->sk_receive_queue)) {
- __skb_unlink(skb, &sk->sk_receive_queue);
- clear = 1;
- }
- spin_unlock_bh(&sk->sk_receive_queue.lock);
- if (clear)
- kfree_skb(skb);
- }
-
- skb_free_datagram(sk, skb);
+ skb_kill_datagram(sk, skb, flags);
if (noblock)
return -EAGAIN;
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -32,6 +32,7 @@
#include <linux/icmpv6.h>
#include <linux/netfilter.h>
#include <linux/netfilter_ipv6.h>
+#include <linux/skbuff.h>
#include <asm/uaccess.h>
#include <asm/ioctls.h>
#include <asm/bug.h>
@@ -433,25 +434,14 @@ out:
return err;
csum_copy_err:
- /* Clear queue. */
- if (flags&MSG_PEEK) {
- int clear = 0;
- spin_lock_bh(&sk->sk_receive_queue.lock);
- if (skb == skb_peek(&sk->sk_receive_queue)) {
- __skb_unlink(skb, &sk->sk_receive_queue);
- clear = 1;
- }
- spin_unlock_bh(&sk->sk_receive_queue.lock);
- if (clear)
- kfree_skb(skb);
- }
+ skb_kill_datagram(sk, skb, flags);
/* Error for blocking case is chosen to masquerade
as some normal condition.
*/
err = (flags&MSG_DONTWAIT) ? -EAGAIN : -EHOSTUNREACH;
/* FIXME: increment a raw6 drops counter here */
- goto out_free;
+ goto out;
}
static int rawv6_push_pending_frames(struct sock *sk, struct flowi *fl,
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -36,6 +36,7 @@
#include <linux/ipv6.h>
#include <linux/icmpv6.h>
#include <linux/init.h>
+#include <linux/skbuff.h>
#include <asm/uaccess.h>
#include <net/sock.h>
@@ -300,20 +301,7 @@ out:
return err;
csum_copy_err:
- /* Clear queue. */
- if (flags&MSG_PEEK) {
- int clear = 0;
- spin_lock_bh(&sk->sk_receive_queue.lock);
- if (skb == skb_peek(&sk->sk_receive_queue)) {
- __skb_unlink(skb, &sk->sk_receive_queue);
- clear = 1;
- }
- spin_unlock_bh(&sk->sk_receive_queue.lock);
- if (clear)
- kfree_skb(skb);
- }
-
- skb_free_datagram(sk, skb);
+ skb_kill_datagram(sk, skb, flags);
if (flags & MSG_DONTWAIT) {
UDP6_INC_STATS_USER(UDP_MIB_INERRORS);
