From mboxrd@z Thu Jan 1 00:00:00 1970 From: Luis Henriques Subject: Re: [3.16.y-ckt stable] Patch "xen-netfront: Fix handling packets on compound pages with skb_linearize" has been added to staging queue Date: Wed, 10 Dec 2014 17:54:52 +0000 Message-ID: <20141210175452.GE10463@hercules> References: <1418232388-16853-1-git-send-email-luis.henriques@canonical.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Wei Liu , Ian Campbell , Paul Durrant , netdev@vger.kernel.org, linux-kernel@vger.kernel.org, xen-devel@lists.xenproject.org, "David S. Miller" , Stefan Bader , kernel-team@lists.ubuntu.com To: Zoltan Kiss Return-path: Content-Disposition: inline In-Reply-To: <1418232388-16853-1-git-send-email-luis.henriques@canonical.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Wed, Dec 10, 2014 at 05:26:28PM +0000, Luis Henriques wrote: > This is a note to let you know that I have just added a patch titled >=20 > xen-netfront: Fix handling packets on compound pages with skb_lin= earize >=20 > to the linux-3.16.y-queue branch of the 3.16.y-ckt extended stable tr= ee=20 > which can be found at: >=20 > http://kernel.ubuntu.com/git?p=3Dubuntu/linux.git;a=3Dshortlog;h=3Dr= efs/heads/linux-3.16.y-queue >=20 > This patch is scheduled to be released in version 3.16.7-ckt3. >=20 > If you, or anyone else, feels it should not be added to this tree, pl= ease=20 > reply to this email. >=20 Ups, sorry! I thought I had this one dropped after David Vrabel comment. Now dropped for sure. Cheers, -- Lu=EDs > For more information about the 3.16.y-ckt tree, see > https://wiki.ubuntu.com/Kernel/Dev/ExtendedStable >=20 > Thanks. > -Luis >=20 > ------ >=20 > From 079f869ca5082866d276b686721e89a649e622fe Mon Sep 17 00:00:00 200= 1 > From: Zoltan Kiss > Date: Mon, 11 Aug 2014 18:32:23 +0100 > Subject: xen-netfront: Fix handling packets on compound pages with > skb_linearize >=20 > commit 97a6d1bb2b658ac85ed88205ccd1ab809899884d upstream. >=20 > 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 r= ing slots, > it gets dropped. The reason is that netback maps these slots to a fra= g in the > frags array, which is limited by size. Having so many slots can occur= since > compound pages were introduced, as the ring protocol slice them up in= to > individual (non-compound) page aligned slots. The theoretical worst c= ase > scenario looks like this (note, skbs are limited to 64 Kb here): > linear buffer: at most PAGE_SIZE - 17 * 2 bytes, overlapping page bou= ndary, > using 2 slots > first 15 frags: 1 + PAGE_SIZE + 1 bytes long, first and last bytes ar= e at the > end and the beginning of a page, therefore they use 3 * 15 =3D 45 slo= ts > last 2 frags: 1 + 1 bytes, overlapping page boundary, 2 * 2 =3D 4 slo= ts > 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 few = slots > overdue, but usually it causes the TCP stream to be blocked, as the r= etry will > most likely have the same buffer layout. > This patch solves this problem by linearizing the packet. This is not= the > fastest way, and it can fail much easier as it tries to allocate a bi= g linear > area for the whole packet, but probably easier by an order of magnitu= de than > anything else. Probably this code path is not touched very frequently= 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 > Signed-off-by: David S. Miller > Cc: Stefan Bader > Signed-off-by: Luis Henriques > --- > drivers/net/xen-netfront.c | 7 ++++--- > 1 file changed, 4 insertions(+), 3 deletions(-) >=20 > diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c > index 055222bae6e4..23359aeb1ba0 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 *skb= , 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 by= tes\n", > + slots, skb->len); > + if (skb_linearize(skb)) > + goto drop; > } >=20 > spin_lock_irqsave(&queue->tx_lock, flags);