diff mbox

forcedeth skb_gso_segment warning

Message ID 20090629040641.GA3357@gondor.apana.org.au
State Accepted, archived
Delegated to: David Miller
Headers show

Commit Message

Herbert Xu June 29, 2009, 4:06 a.m. UTC
On Mon, Jun 29, 2009 at 12:03:30PM +0800, Herbert Xu wrote:
> 
> tcp: Stop non-TSO packets morphing into TSO

Here's a follow-up patch to optimise this boundary case.

tcp: Do not tack on TSO data to non-TSO packet

If a socket starts out on a non-TSO route, and then switches to
a TSO route, then we will tack on data to the tail of the tx queue
even if it started out life as non-TSO.  This is suboptimal because
all of it will then be copied and checksummed unnecessarily.

This patch fixes this by ensuring that skb->ip_summed is set to
CHECKSUM_PARTIAL before appending extra data beyond the MSS.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>


Cheers,

Comments

David Miller June 30, 2009, 2:43 a.m. UTC | #1
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Mon, 29 Jun 2009 12:06:41 +0800

> On Mon, Jun 29, 2009 at 12:03:30PM +0800, Herbert Xu wrote:
>> 
>> tcp: Stop non-TSO packets morphing into TSO
> 
> Here's a follow-up patch to optimise this boundary case.
> 
> tcp: Do not tack on TSO data to non-TSO packet
> 
> If a socket starts out on a non-TSO route, and then switches to
> a TSO route, then we will tack on data to the tail of the tx queue
> even if it started out life as non-TSO.  This is suboptimal because
> all of it will then be copied and checksummed unnecessarily.
> 
> This patch fixes this by ensuring that skb->ip_summed is set to
> CHECKSUM_PARTIAL before appending extra data beyond the MSS.
> 
> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Applied.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 17b89c5..7870a53 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -903,13 +903,17 @@  int tcp_sendmsg(struct kiocb *iocb, struct socket *sock, struct msghdr *msg,
 		iov++;
 
 		while (seglen > 0) {
-			int copy;
+			int copy = 0;
+			int max = size_goal;
 
 			skb = tcp_write_queue_tail(sk);
+			if (tcp_send_head(sk)) {
+				if (skb->ip_summed == CHECKSUM_NONE)
+					max = mss_now;
+				copy = max - skb->len;
+			}
 
-			if (!tcp_send_head(sk) ||
-			    (copy = size_goal - skb->len) <= 0) {
-
+			if (copy <= 0) {
 new_segment:
 				/* Allocate new segment. If the interface is SG,
 				 * allocate skb fitting to single page.
@@ -930,6 +934,7 @@  new_segment:
 
 				skb_entail(sk, skb);
 				copy = size_goal;
+				max = size_goal;
 			}
 
 			/* Try to append data to the end of skb. */
@@ -1028,7 +1033,7 @@  new_segment:
 			if ((seglen -= copy) == 0 && iovlen == 0)
 				goto out;
 
-			if (skb->len < size_goal || (flags & MSG_OOB))
+			if (skb->len < max || (flags & MSG_OOB))
 				continue;
 
 			if (forced_push(tp)) {