Hi Keir:
Here are the GSO changes again which should address your concerns. Let me
know if you have any other problems.
[NET] back: Fix off-by-one error in netbk_tx_err
The generalised extra request info patch introduced a bug with the use
of netbk_tx_err since it advanced the req_cons pointer by one. This
patch fixes thing by delaying the increment in netbk_tx_err.
Signed-off-by: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>
Cheers,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@xxxxxxxxxxxxxxxxxxx>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
--
diff -r 3fe11185adfb -r 3656a2985ae1 
linux-2.6-xen-sparse/drivers/xen/netback/netback.c
--- a/linux-2.6-xen-sparse/drivers/xen/netback/netback.c Sat Jul 01 
09:37:24 2006 +0100
+++ b/linux-2.6-xen-sparse/drivers/xen/netback/netback.c Mon Jul 03 
14:18:54 2006 +1000
@@ -496,9 +496,9 @@ static void netbk_tx_err(netif_t *netif,
 
 do {
 make_tx_response(netif, txp, NETIF_RSP_ERROR);
- if (++cons >= end)
+ if (cons >= end)
 break;
- txp = RING_GET_REQUEST(&netif->tx, cons);
+ txp = RING_GET_REQUEST(&netif->tx, cons++);
 } while (1);
 netif->tx.req_cons = cons;
 netif_schedule_work(netif);
@@ -764,11 +764,11 @@ static void net_tx_action(unsigned long 
 if (txreq.flags & NETTXF_extra_info) {
 work_to_do = netbk_get_extras(netif, extras,
 work_to_do);
+ i = netif->tx.req_cons;
 if (unlikely(work_to_do < 0)) {
- netbk_tx_err(netif, &txreq, 0);
+ netbk_tx_err(netif, &txreq, i);
 continue;
 }
- i = netif->tx.req_cons;
 }
 
 ret = netbk_count_requests(netif, &txreq, work_to_do);
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel