Cloning the big skbs we use for USB buffering chokes up TCP and
SCTP because the socket memory limits are hitting earlier than
they should. It is better to unconditionally copy the unwrapped
packets to freshly allocated skbs.

Reported-by: Jim Baxter <jim_bax...@mentor.com>
Acked-by: Eric Dumazet <eduma...@google.com>
Signed-off-by: Bjørn Mork <bj...@mork.no>
---
 drivers/net/usb/cdc_ncm.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/drivers/net/usb/cdc_ncm.c b/drivers/net/usb/cdc_ncm.c
index 93c9ca9924eb..2bbbd65591c7 100644
--- a/drivers/net/usb/cdc_ncm.c
+++ b/drivers/net/usb/cdc_ncm.c
@@ -1289,12 +1289,11 @@ next_ndp:
                        break;
 
                } else {
-                       skb = skb_clone(skb_in, GFP_ATOMIC);
+                       /* create a fresh copy to reduce truesize */
+                       skb = netdev_alloc_skb_ip_align(dev->net,  len);
                        if (!skb)
                                goto error;
-                       skb->len = len;
-                       skb->data = ((u8 *)skb_in->data) + offset;
-                       skb_set_tail_pointer(skb, len);
+                       memcpy(skb_put(skb, len), skb_in->data + offset, len);
                        usbnet_skb_return(dev, skb);
                        payload += len; /* count payload bytes in this NTB */
                }
-- 
2.0.0.rc4

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

Reply via email to