sk_attach_filter() and sk_detach_filter() are run with socket locked.

Use the appropriate rcu_dereference_protected() instead of blocking BH,
and rcu_dereference_bh().
There is no point adding BH prevention and memory barrier.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

It seems we dont use appropriate refcount increment in an
rcu_read_lock() protected section.

fib_rule_get() might increment a null refcount and bad things could
happen.

While fib_nl_delrule() respects an rcu grace period before calling
fib_rule_put(), fib_rules_cleanup_ops() calls fib_rule_put() without a
grace period.

Note : after this patch, we might avoid the synchronize_rcu() call done
in fib_nl_delrule()
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.

Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.

This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.

Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.

This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :

> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> >  	if (parms->name[0])
> >  		strlcpy(name, parms->name, IFNAMSIZ);
> >  	else
> > -		sprintf(name, "gre%%d");
> > +		strcpy(name, "gre%d");
> >
> >  	dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> >  	if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>

Sorry ? It was not a fix, but at most a cleanup ;)

Anyway I forgot the gretap case...

[PATCH 2/4 v2] ip_gre: percpu stats accounting

Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.

Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.

This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Tunnels are going to use percpu for their accounting.

They are going to use a new tstats field in net_device.

skb_tunnel_rx() is changed to be a wrapper around __skb_tunnel_rx()

IPTUNNEL_XMIT() is changed to be a wrapper around __IPTUNNEL_XMIT()
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Phonet stack assumes the presence of Pipe Controller, either in Modem or
on Application Processing Engine user-space for the Pipe data.
Nokia Slim Modems like WG2.5 used in ST-Ericsson U8500 platform do not
implement Pipe controller in them.
This patch adds Pipe Controller implemenation to Phonet stack to support
Pipe data over Phonet stack for Nokia Slim Modems.
Signed-off-by: Kumar Sanghvi <kumar.sanghvi@stericsson.com>
Acked-by: Linus Walleij <linus.walleij@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Conflicts:
	drivers/net/qlcnic/qlcnic_init.c
	net/ipv4/ip_output.c

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: David S. Miller <davem@davemloft.net>

Clean up a missing exit path in the ipv6 module init routines.  In
addrconf_init we call ipv6_addr_label_init which calls register_pernet_subsys
for the ipv6_addr_label_ops structure.  But if module loading fails, or if the
ipv6 module is removed, there is no corresponding unregister_pernet_subsys call,
which leaves a now-bogus address on the pernet_list, leading to oopses in
subsequent registrations.  This patch cleans up both the failed load path and
the unload path.  Tested by myself with good results.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>

 include/net/addrconf.h |    1 +
 net/ipv6/addrconf.c    |   11 ++++++++---
 net/ipv6/addrlabel.c   |    5 +++++
 3 files changed, 14 insertions(+), 3 deletions(-)
Signed-off-by: David S. Miller <davem@davemloft.net>

loopback driver uses dev->ml_priv to store its percpu stats pointer.
It uses ugly casts "(void __percpu __force *)" to shut up sparse
complains.

Define an union to better document we use ml_priv in loopback driver and
define a lstats field with appropriate types.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

__in_dev_get_rtnl(dev_out) is called while RTNL is not held, thus
triggers a lockdep fault.

At this point, we only perform a raw test of dev_out->ip_ptr being NULL,
we dont need to make sure ip_ptr cant changed right after.

We can use rcu_dereference_raw() for this.
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Instead of having two places were we allocate dev->_rx, introduce
netif_alloc_rx_queues() helper and call it only from
register_netdevice(), not from alloc_netdev_mq()

Goal is to let drivers change dev->num_rx_queues after allocating netdev
and before registering it.

This also removes a lot of ifdefs in net/core/dev.c
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Freeing netdev without free_netdev() leads to net, tx leaks.
I might lead to dereferencing freed pointer.

The semantic match that finds this problem is as follows:
(http://coccinelle.lip6.fr/)

@@
struct net_device* dev;
@@

-kfree(dev)
+free_netdev(dev)
Signed-off-by: David S. Miller <davem@davemloft.net>

Freeing netdev without free_netdev() leads to net, tx leaks.
I might lead to dereferencing freed pointer.

The semantic match that finds this problem is as follows:
(http://coccinelle.lip6.fr/)

@@
struct net_device* dev;
@@

-kfree(dev)
+free_netdev(dev)
Signed-off-by: David S. Miller <davem@davemloft.net>

Freeing netdev without free_netdev() leads to net, tx leaks.
I might lead to dereferencing freed pointer.

The semantic match that finds this problem is as follows:
(http://coccinelle.lip6.fr/)

@@
struct net_device* dev;
@@

-kfree(dev)
+free_netdev(dev)
Signed-off-by: David S. Miller <davem@davemloft.net>

Freeing netdev without free_netdev() leads to net, tx leaks.
I might lead to dereferencing freed pointer.

The semantic match that finds this problem is as follows:
(http://coccinelle.lip6.fr/)

@@
struct net_device* dev;
@@

-kfree(dev)
+free_netdev(dev)
Signed-off-by: David S. Miller <davem@davemloft.net>

SOCK_MIN_RCVBUF current value is 256 bytes

It doesnt permit to receive the smallest possible frame, considering
socket sk_rmem_alloc/sk_rcvbuf account skb truesizes. On 64bit arches,
sizeof(struct sk_buff) is 240 bytes. Add the typical 64 bytes of
headroom, and we go over the limit.

With old kernels and 32bit arches, we were under the limit, if netdriver
was doing copybreak.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This enables auto loading for the smsc911x ethernet driver.
Signed-off-by: Vincent Stehlé <vincent.stehle@laposte.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Reset queue mapping when an skb is reentering the stack via a tunnel.
On second pass, the queue mapping from the original device is no
longer valid.
Signed-off-by: Tom Herbert <therbert@google.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Change "return (EXPR);" to "return EXPR;"

return is not a function, parentheses are not required.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

On 32bit arches, if PAGE_SIZE is smaller than 65536, we can use 16bit
offset and size fields. This patch saves 72 bytes per skb on i386, or
128 bytes after rounding.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

You can't call atomic_notifier_chain_unregister() while in atomic context.

Fix, call un/register_atmdevice_notifier in module __init and __exit.

Bug report:
http://comments.gmane.org/gmane.linux.network/172603Reported-by: Mikko Vinni <mmvinni@yahoo.com>
Tested-by: Mikko Vinni <mmvinni@yahoo.com>
Signed-off-by: Karl Hiramoto <karl@hiramoto.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Automatically allows vlans to get NETIF_F_HIGHDMA if underlying device
supports it.

On 32bit arches (and more precisely if CONFIG_HIGHMEM is enabled), it
can help to reduce cost of illegal_highdma() and __skb_linearize()
calls.

Tested on tg3 , bnx2, bonding, this worked very well.

This is a generalization of a patch provided by Yi Zou & Jeff Kirsher.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Compex FreedomLine 32 PnP-PCI2 cards have only TP and BNC connectors but the
SROM contains AUI port too. When TP loses link, the driver switches to
non-existing AUI port (which reports that carrier is always present).

Connecting TP back generates LinkPass interrupt but de_media_interrupt() is
broken - it only updates the link state of currently connected media, ignoring
the fact that LinkPass and LinkFail bits of MacStatus register belong to the
TP port only (the chip documentation says that).

This patch changes de_media_interrupt() to switch media to TP when link goes
up (and media type is not locked) and also to update the link state only when
the TP port is used.

Also the NonselPortActive (and also SelPortActive) bits of SIAStatus register
need to be cleared (by writing 1) after reading or they're useless.
Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Acked-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

At least my 21041 cards come out of suspend with bus mastering disabled so
they did not work after resume(no data transferred).
After adding pci_set_master(), the driver oopsed immediately on resume -
because de_clean_rings() is called on suspend but de_init_rings() call
was missing in resume.

Also disable link (reset SIA) before sleep (de4x5 does this too).
Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Acked-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The function to resize the Tx/Rx rings had the potential to
dereference a NULL pointer and the code would attempt to resize
the Tx ring even if the Rx ring allocation had failed.  This
would cause some confusion in the return code semantics.  Fixed
up to just unwind the allocations if any of them fail and return
an error.
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Tested-by: Emil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

At least on older 21041-AA chips (mine is rev. 11), TP duplex autonegotiation
causes the card not to work at all (link is up but no packets are transmitted).

de4x5 disables autonegotiation completely. But it seems to work on newer
(21041-PA rev. 21) so disable it only on rev<20 chips.
Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Acked-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

We have for each socket :

One spinlock (sk_slock.slock)
One rwlock (sk_callback_lock)

Possible scenarios are :

(A) (this is used in net/sunrpc/xprtsock.c)
read_lock(&sk->sk_callback_lock) (without blocking BH)
<BH>
spin_lock(&sk->sk_slock.slock);
...
read_lock(&sk->sk_callback_lock);
...

(B)
write_lock_bh(&sk->sk_callback_lock)
stuff
write_unlock_bh(&sk->sk_callback_lock)

(C)
spin_lock_bh(&sk->sk_slock)
...
write_lock_bh(&sk->sk_callback_lock)
stuff
write_unlock_bh(&sk->sk_callback_lock)
spin_unlock_bh(&sk->sk_slock)

This (C) case conflicts with (A) :

CPU1 [A]                         CPU2 [C]
read_lock(callback_lock)
<BH>                             spin_lock_bh(slock)
<wait to spin_lock(slock)>
                                 <wait to write_lock_bh(callback_lock)>

We have one problematic (C) use case in inet_csk_listen_stop() :

local_bh_disable();
bh_lock_sock(child); // spin_lock_bh(&sk->sk_slock)
WARN_ON(sock_owned_by_user(child));
...
sock_orphan(child); // write_lock_bh(&sk->sk_callback_lock)

lockdep is not happy with this, as reported by Tetsuo Handa

It seems only way to deal with this is to use read_lock_bh(callbacklock)
everywhere.

Thanks to Jarek for pointing a bug in my first attempt and suggesting
this solution.
Reported-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Tested-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Jarek Poplawski <jarkao2@gmail.com>
Tested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If the PM support is available this is passed
through the platform instead to be hard-coded
in the core files.
WoL on Magic Frame can be enabled by using
the ethtool support.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Masayuki Ohtake <masa-korg@dsn.okisemi.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

While investigating a bit, I found ip_fragment() slow path was taken
because ip_append_data() provides following layout for a send(MTU +
N*(MTU - 20)) syscall :

- one skb with 1500 (mtu) bytes
- N fragments of 1480 (mtu-20) bytes (before adding IP header)
last fragment gets 17 bytes of trail data because of following bit:

	if (datalen == length + fraggap)
		alloclen += rt->dst.trailer_len;

Then esp4 adds 16 bytes of data (while trailer_len is 17... hmm...
another bug ?)

In ip_fragment(), we notice last fragment is too big (1496 + 20) > mtu,
so we take slow path, building another skb chain.

In order to avoid taking slow path, we should correct ip_append_data()
to make sure last fragment has real trail space, under mtu...
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Change "return (EXPR);" to "return EXPR;"

return is not a function, parentheses are not required.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

E1000 can benefit from calling the GRO receive functions.
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Net drivers in general have an issue where timers fired
by mod_timer or work threads with schedule_work are running
outside of the rtnl_lock.

With no other lock protection these routines are vulnerable
to races with driver unload or reset paths.

The longer term solution to this might be a redesign with
safer locks being taken in the driver to guarantee no
reentrance, but for now a safe and effective fix is
to take the rtnl_lock in these routines.
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

E1000 is using several timers that in a follow on patch
will need to acquire the rtnl_lock in order to be safe.

This patch moves the timer bodies into work queues which
will allow the next patch to add rtnl_lock.
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If the netdev->features is set with NETIF_F_HIGHDMA, we should set the
corresponding netdev->vlan_features as well to allow VLAN netdev created
on top of the real netdev to be able to also benefit from HIGHDMA on 32bit
system, reducing the performance hit that is caused by __skb_linearize(),
particularly for large send. This is fixed in this patch for all Intel e1000,
e1000e, igb, ixgbe, and ixgbe drivers since this should be beneficial
to all devices supported by these drivers.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Tested-by: Emil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds support for the Intel(r) DH89xxCC series. The new
device will be using Intel(r) i347-AT4 and Marvell(r) M88E1322 and
M88E1112 PHYs. Support for these devices has also been added here.
Signed-off-by: Joseph Gasparakis <joseph.gasparakis@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This change corrects an issue in which we were setting all flag bits except
for promisc instead of clearing the promisc bits due to the incorrect use
of an |= instead of an &=.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>