| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <37D7C6CF3E00A74B8858931C1DB2F07750BBF524@SHSMSX103.ccr.corp.intel.com>
Date: Tue, 13 Sep 2016 12:22:52 +0000
From: "Liang, Kan" <kan.liang@...el.com>
To: Tom Herbert <tom@...bertland.com>
CC: "David S. Miller" <davem@...emloft.net>,
LKML <linux-kernel@...r.kernel.org>,
Linux Kernel Network Developers <netdev@...r.kernel.org>,
"Kirsher, Jeffrey T" <jeffrey.t.kirsher@...el.com>,
Ingo Molnar <mingo@...hat.com>,
"peterz@...radead.org" <peterz@...radead.org>,
Alexey Kuznetsov <kuznet@....inr.ac.ru>,
James Morris <jmorris@...ei.org>,
Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>,
"Patrick McHardy" <kaber@...sh.net>,
"akpm@...ux-foundation.org" <akpm@...ux-foundation.org>,
Kees Cook <keescook@...omium.org>,
"viro@...iv.linux.org.uk" <viro@...iv.linux.org.uk>,
"gorcunov@...nvz.org" <gorcunov@...nvz.org>,
John Stultz <john.stultz@...aro.org>,
Alexander Duyck <aduyck@...antis.com>,
David Decotigny <decot@...glers.com>,
Alexander Duyck <alexander.duyck@...il.com>,
Daniel Borkmann <daniel@...earbox.net>,
"rdunlap@...radead.org" <rdunlap@...radead.org>,
Hannes Frederic Sowa <hannes@...essinduktion.org>,
Stephen Hemminger <stephen@...workplumber.org>,
Alexei Starovoitov <alexei.starovoitov@...il.com>,
"Brandeburg, Jesse" <jesse.brandeburg@...el.com>,
Andi Kleen <andi@...stfloor.org>
Subject: RE: [RFC V3 PATCH 18/26] net/netpolicy: set tx queues according to
policy
> -----Original Message-----
> From: Tom Herbert [mailto:tom@...bertland.com]
> Sent: Monday, September 12, 2016 4:23 PM
> To: Liang, Kan <kan.liang@...el.com>
> Cc: David S. Miller <davem@...emloft.net>; LKML <linux-
> kernel@...r.kernel.org>; Linux Kernel Network Developers
> <netdev@...r.kernel.org>; Kirsher, Jeffrey T <jeffrey.t.kirsher@...el.com>;
> Ingo Molnar <mingo@...hat.com>; peterz@...radead.org; Alexey Kuznetsov
> <kuznet@....inr.ac.ru>; James Morris <jmorris@...ei.org>; Hideaki
> YOSHIFUJI <yoshfuji@...ux-ipv6.org>; Patrick McHardy <kaber@...sh.net>;
> akpm@...ux-foundation.org; Kees Cook <keescook@...omium.org>;
> viro@...iv.linux.org.uk; gorcunov@...nvz.org; John Stultz
> <john.stultz@...aro.org>; Alexander Duyck <aduyck@...antis.com>; Ben
> Hutchings <ben@...adent.org.uk>; David Decotigny <decot@...glers.com>;
> Florian Westphal <fw@...len.de>; Alexander Duyck
> <alexander.duyck@...il.com>; Daniel Borkmann <daniel@...earbox.net>;
> rdunlap@...radead.org; Cong Wang <xiyou.wangcong@...il.com>; Hannes
> Frederic Sowa <hannes@...essinduktion.org>; Stephen Hemminger
> <stephen@...workplumber.org>; Alexei Starovoitov
> <alexei.starovoitov@...il.com>; Brandeburg, Jesse
> <jesse.brandeburg@...el.com>; Andi Kleen <andi@...stfloor.org>
> Subject: Re: [RFC V3 PATCH 18/26] net/netpolicy: set tx queues according to
> policy
>
> On Mon, Sep 12, 2016 at 7:55 AM, <kan.liang@...el.com> wrote:
> > From: Kan Liang <kan.liang@...el.com>
> >
> > When the device tries to transmit a packet, netdev_pick_tx is called
> > to find the available tx queues. If the net policy is applied, it
> > picks up the assigned tx queue from net policy subsystem, and redirect
> > the traffic to the assigned queue.
> >
> > Signed-off-by: Kan Liang <kan.liang@...el.com>
> > ---
> > include/net/sock.h | 9 +++++++++
> > net/core/dev.c | 20 ++++++++++++++++++--
> > 2 files changed, 27 insertions(+), 2 deletions(-)
> >
> > diff --git a/include/net/sock.h b/include/net/sock.h index
> > e1e9e3d..ca97f35 100644
> > --- a/include/net/sock.h
> > +++ b/include/net/sock.h
> > @@ -2280,4 +2280,13 @@ extern int sysctl_optmem_max; extern __u32
> > sysctl_wmem_default; extern __u32 sysctl_rmem_default;
> >
> > +/* Return netpolicy instance information from socket. */ static
> > +inline struct netpolicy_instance *netpolicy_find_instance(struct sock
> > +*sk) { #ifdef CONFIG_NETPOLICY
> > + if (is_net_policy_valid(sk->sk_netpolicy.policy))
> > + return &sk->sk_netpolicy; #endif
> > + return NULL;
> > +}
> > #endif /* _SOCK_H */
> > diff --git a/net/core/dev.c b/net/core/dev.c index 34b5322..b9a8044
> > 100644
> > --- a/net/core/dev.c
> > +++ b/net/core/dev.c
> > @@ -3266,6 +3266,7 @@ struct netdev_queue *netdev_pick_tx(struct
> net_device *dev,
> > struct sk_buff *skb,
> > void *accel_priv) {
> > + struct sock *sk = skb->sk;
> > int queue_index = 0;
> >
> > #ifdef CONFIG_XPS
> > @@ -3280,8 +3281,23 @@ struct netdev_queue *netdev_pick_tx(struct
> net_device *dev,
> > if (ops->ndo_select_queue)
> > queue_index = ops->ndo_select_queue(dev, skb, accel_priv,
> > __netdev_pick_tx);
> > - else
> > - queue_index = __netdev_pick_tx(dev, skb);
> > + else {
> > +#ifdef CONFIG_NETPOLICY
> > + struct netpolicy_instance *instance;
> > +
> > + queue_index = -1;
> > + if (dev->netpolicy && sk) {
> > + instance = netpolicy_find_instance(sk);
> > + if (instance) {
> > + if (!instance->dev)
> > + instance->dev = dev;
> > + queue_index = netpolicy_pick_queue(instance, false);
> > + }
> > + }
> > + if (queue_index < 0) #endif
>
> I doubt this produces the intended effect. Several drivers use
> ndo_select_queue (such as mlx4) where there might do something special
> for a few packets but end up called the default handler which
> __netdev_pick_tx for most packets. So in such cases the netpolicy path would
> be routinely bypassed. Maybe this code should be in __netdev_pick_tx.
I will move the code to __netdev_pick_tx in next version.
Thanks,
Kan
>
> Tom
>
> > + queue_index = __netdev_pick_tx(dev, skb);
> > + }
> >
> > if (!accel_priv)
> > queue_index = netdev_cap_txqueue(dev,
> > queue_index);
> > --
> > 2.5.5
> >
Powered by blists - more mailing lists