[Xen-changelog] [xen-unstable] Adjust truesize in netfront to more closely reflect real data payload

[Xen patchbot-unstable <patchbot-unstable@xxxxxxxxxxxxxxxxxxx>]

# HG changeset patch
# User ack@xxxxxxxxxxxxxxxxxxxxx
# Node ID d48842f924d0e0c8e6e52887914850f4c7c02585
# Parent  b53b154a5126f55cccee374c71d33145cb1ef600
Adjust truesize in netfront to more closely reflect real data payload
plus supervisor overheads but keeping out gross hypervisor overheads.
Signed-off-by: Emmanuel Ackaouy <ack@xxxxxxxxxxxxx>
---
 linux-2.6-xen-sparse/drivers/xen/netfront/netfront.c |   23 ++++++++++++++++++-
 1 files changed, 22 insertions(+), 1 deletion(-)

diff -r b53b154a5126 -r d48842f924d0 
linux-2.6-xen-sparse/drivers/xen/netfront/netfront.c
--- a/linux-2.6-xen-sparse/drivers/xen/netfront/netfront.c      Tue Aug 15 
10:06:13 2006 +0100
+++ b/linux-2.6-xen-sparse/drivers/xen/netfront/netfront.c      Tue Aug 15 
11:13:04 2006 +0100
@@ -1199,7 +1199,28 @@ err:
                }
 
                i = xennet_fill_frags(np, skb, &tmpq);
-               skb->truesize += skb->data_len;
+
+        /*
+         * Truesize must approximates the size of true data plus
+         * any supervisor overheads. Adding hypervisor overheads
+         * has been shown to significantly reduce achievable
+         * bandwidth with the default receive buffer size. It is
+         * therefore not wise to account for it here.
+         *
+         * After alloc_skb(RX_COPY_THRESHOLD), truesize is set to
+         * RX_COPY_THRESHOLD + the supervisor overheads. Here, we
+         * add the size of the data pulled in xennet_fill_frags().
+         *
+         * We also adjust for any unused space in the main data
+         * area by subtracting (RX_COPY_THRESHOLD - len). This is
+         * especially important with drivers which split incoming
+         * packets into header and data, using only 66 bytes of
+         * the main data area (see the e1000 driver for example.)
+         * On such systems, without this last adjustement, our
+         * achievable receive throughout using the standard receive
+         * buffer size was cut by 25%(!!!).
+         */
+               skb->truesize += skb->data_len - (RX_COPY_THRESHOLD - len);
                skb->len += skb->data_len;
 
                /*

_______________________________________________
Xen-changelog mailing list
Xen-changelog@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-changelog