From mboxrd@z Thu Jan 1 00:00:00 1970 From: Luis Henriques Subject: Re: [Xen-devel] [PATCH] xen-netfront: Fix handling packets on compound pages with skb_linearize Date: Mon, 8 Dec 2014 10:19:36 +0000 Message-ID: <20141208101936.GA7491@hercules> References: <1407778343-13622-1-git-send-email-zoltan.kiss@citrix.com> <547C2CFC.7060908@canonical.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Zoltan Kiss , Konrad Rzeszutek Wilk , Boris Ostrovsky , David Vrabel , Wei Liu , Ian Campbell , netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Paul Durrant , xen-devel@lists.xenproject.org, Kamal Mostafa To: Stefan Bader Return-path: Content-Disposition: inline In-Reply-To: <547C2CFC.7060908@canonical.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Mon, Dec 01, 2014 at 09:55:24AM +0100, Stefan Bader wrote: > On 11.08.2014 19:32, Zoltan Kiss wrote: > > There is a long known problem with the netfront/netback interface: = if the guest > > tries to send a packet which constitues more than MAX_SKB_FRAGS + 1= ring slots, > > it gets dropped. The reason is that netback maps these slots to a f= rag in the > > frags array, which is limited by size. Having so many slots can occ= ur since > > compound pages were introduced, as the ring protocol slice them up = into > > individual (non-compound) page aligned slots. The theoretical worst= case > > scenario looks like this (note, skbs are limited to 64 Kb here): > > linear buffer: at most PAGE_SIZE - 17 * 2 bytes, overlapping page b= oundary, > > using 2 slots > > first 15 frags: 1 + PAGE_SIZE + 1 bytes long, first and last bytes = are at the > > end and the beginning of a page, therefore they use 3 * 15 =3D 45 s= lots > > last 2 frags: 1 + 1 bytes, overlapping page boundary, 2 * 2 =3D 4 s= lots > > Although I don't think this 51 slots skb can really happen, we need= a solution > > which can deal with every scenario. In real life there is only a fe= w slots > > overdue, but usually it causes the TCP stream to be blocked, as the= retry will > > most likely have the same buffer layout. > > This patch solves this problem by linearizing the packet. This is n= ot the > > fastest way, and it can fail much easier as it tries to allocate a = big linear > > area for the whole packet, but probably easier by an order of magni= tude than > > anything else. Probably this code path is not touched very frequent= ly anyway. > >=20 > > Signed-off-by: Zoltan Kiss > > Cc: Wei Liu > > Cc: Ian Campbell > > Cc: Paul Durrant > > Cc: netdev@vger.kernel.org > > Cc: linux-kernel@vger.kernel.org > > Cc: xen-devel@lists.xenproject.org >=20 > This does not seem to be marked explicitly as stable. Has someone alr= eady asked > David Miller to put it on his stable queue? IMO it qualifies quite we= ll and the > actual change should be simple to pick/backport. >=20 Thank you Stefan, I'm queuing this for the next 3.16 kernel release. Cheers, -- Lu=EDs > -Stefan >=20 > >=20 > > diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.= c > > index 055222b..23359ae 100644 > > --- a/drivers/net/xen-netfront.c > > +++ b/drivers/net/xen-netfront.c > > @@ -628,9 +628,10 @@ static int xennet_start_xmit(struct sk_buff *s= kb, struct net_device *dev) > > slots =3D DIV_ROUND_UP(offset + len, PAGE_SIZE) + > > xennet_count_skb_frag_slots(skb); > > if (unlikely(slots > MAX_SKB_FRAGS + 1)) { > > - net_alert_ratelimited( > > - "xennet: skb rides the rocket: %d slots\n", slots); > > - goto drop; > > + net_dbg_ratelimited("xennet: skb rides the rocket: %d slots, %d = bytes\n", > > + slots, skb->len); > > + if (skb_linearize(skb)) > > + goto drop; > > } > > =20 > > spin_lock_irqsave(&queue->tx_lock, flags); > >=20 > > _______________________________________________ > > Xen-devel mailing list > > Xen-devel@lists.xen.org > > http://lists.xen.org/xen-devel > >=20 >=20 >=20