sock_alloc() currently uses following code to update sockets_in_use

get_cpu_var(sockets_in_use)++;
put_cpu_var(sockets_in_use);

This translates to :

c0436274:       b8 01 00 00 00          mov    $0x1,%eax
c0436279:       e8 42 40 df ff          call   c022a2c0 <add_preempt_count>
c043627e:       bb 20 4f 6a c0          mov    $0xc06a4f20,%ebx
c0436283:       e8 18 ca f0 ff          call   c0342ca0 <debug_smp_processor_id>
c0436288:       03 1c 85 60 4a 65 c0    add    -0x3f9ab5a0(,%eax,4),%ebx
c043628f:       ff 03                   incl   (%ebx)
c0436291:       b8 01 00 00 00          mov    $0x1,%eax
c0436296:       e8 75 3f df ff          call   c022a210 <sub_preempt_count>
c043629b:       89 e0                   mov    %esp,%eax
c043629d:       25 00 e0 ff ff          and    $0xffffe000,%eax
c04362a2:       f6 40 08 08             testb  $0x8,0x8(%eax)
c04362a6:       75 07                   jne    c04362af <sock_alloc+0x7f>
c04362a8:       8d 46 d8                lea    -0x28(%esi),%eax
c04362ab:       5b                      pop    %ebx
c04362ac:       5e                      pop    %esi
c04362ad:       c9                      leave
c04362ae:       c3                      ret
c04362af:       e8 cc 5d 09 00          call   c04cc080 <preempt_schedule>
c04362b4:       8d 74 26 00             lea    0x0(%esi,%eiz,1),%esi
c04362b8:       eb ee                   jmp    c04362a8 <sock_alloc+0x78>

While percpu_add(sockets_in_use, 1) translates to a single instruction :

c0436275:   64 83 05 20 5f 6a c0    addl   $0x1,%fs:0xc06a5f20
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

VIRTIO_NET_F_MAC indicates the presence of the mac field in config
space, not the validity of the value it contains.  Allow the mac to be
changed at runtime, but only push the change into config space with the
VIRTIO_NET_F_MAC feature present.
Signed-off-by: NAlex Williamson <alex.williamson@hp.com>
Acked-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Firmware blob looks like this...

        __be16 lanai4_data_size
        unsigned char lanai4_code[]
Signed-off-by: NJaswinder Singh Rajput <jaswinderrajput@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

e1000 (and e1000e, igb, ixgbe, ixgb) all do a series of
operations each time a multicast address is added.  The flow goes
something like

1) stack adds one multicast address
2) stack passes whole current list of unicast and multicast
   addresses to driver
3) driver clears entire list in hardware
4) driver programs each multicast address using iomem in a loop

This was causing multicast packets to be lost during the
reprogramming process.

reference with test program:
http://kerneltrap.org/mailarchive/linux-netdev/2009/3/14/5160514/thread

Thanks to Dave Boutcher for his report and test program.

This driver fix prepares an array all at once in memory and
programs it in one shot to the hardware, not requiring an "erase"
cycle.
Signed-off-by: NJesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix this build error:

drivers/net/vxge/vxge-main.c: In function 'vxge_get_vpath_no':
drivers/net/vxge/vxge-main.c:695: error: dereferencing pointer to
incomplete type
...
Signed-off-by: NAlexander Beregalov <a.beregalov@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Remove unused #include <version.h> in drivers/net/qlge/qlge_ethtool.
Signed-off-by: NHuang Weiyi <weiyi.huang@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Remove unused #include <version.h> in drivers/net/dnet.c.
Signed-off-by: NHuang Weiyi <weiyi.huang@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It seems that trivial reset of pcount to one was not sufficient
in tcp_retransmit_skb. Multiple counters experience a positive
miscount when skb's pcount gets lowered without the necessary
adjustments (depending on skb's sacked bits which exactly), at
worst a packets_out miscount can crash at RTO if the write queue
is empty!

Triggering this requires mss change, so bidir tcp or mtu probe or
like.
Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Reported-by: NMarkus Trippelsdorf <markus@trippelsdorf.de>
Tested-by: NUwe Bugla <uwe.bugla@gmx.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We need full-scale adjustment to fix a TCP miscount in the next
patch, so just move it into a helper and call for that from the
other places.
Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some devices cannot send very short usb transfers. To get around this the
firmware adds a known pattern and flags the driver that it should check for
this pattern on short transfers. This flag was not taken into account by
the driver.
Signed-off-by: NJan Dumon <j.dumon@option.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Changed the order in which things are freed. This fixes an oops when
unplugging the device while network traffic is ongoing.
Signed-off-by: NJan Dumon <j.dumon@option.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add EXPORT_SYMBOL_GPL(fsl_pq_mdio_bus_name) for module builds
Signed-off-by: NSegher Boessenkool <segher@kernel.crashing.org>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

commit 1577ecef ("netdev: Merge UCC
and gianfar MDIO bus drivers") introduced a regression so that UCC
MDIO buses no longer work.

This is because fsl_pq_mdio driver wrongly masks all non-TBI PHYs
for !fsl,gianfar-mdio buses, while it should do that only for
fsl,gianfar-tbi buses.
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The driver should pass a device that actually specifies internal DMA
ops, but currently it passes netdev's device, which is wrong and that
causes following oops:

Kernel BUG at c01c4df8 [verbose debug info unavailable]
Oops: Exception in kernel mode, sig: 5 [#1]
[...]
NIP [c01c4df8] get_new_skb+0x7c/0xf8
LR [c01c4da4] get_new_skb+0x28/0xf8
Call Trace:
[ef82be00] [c01c4da4] get_new_skb+0x28/0xf8 (unreliable)
[ef82be20] [c01c4eb8] rx_bd_buffer_set+0x44/0x98
[ef82be40] [c01c62bc] ucc_geth_startup+0x11b0/0x147c
[ef82be80] [c01c6674] ucc_geth_open+0xec/0x2a4
[ef82bea0] [c02288a4] dev_open+0xc0/0x11c
[...]

Fix this by passing of_device's device that specifies DMA ops in its
archdata.
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is a fix to get cs89x0 network driver working on i.MX31ADS
Signed-off-by: NAlan Carvalho de Assis <acassis@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This driver contains experimental NAPI code disabled by default.
The commit bea3348e ("[NET]: Make NAPI polling independent of struct
net_device objects.") converted the NAPI path of this driver but that
conversion was not complete.  This patch fixes a build error
introduced by the commit.
Signed-off-by: NAtsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add Vendor/Product ID's for new devices.
Removed duplicate product ID 0x7361.
Signed-off-by: NJan Dumon <j.dumon@option.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Now that the driver is exclusively an of_platform driver we no longer
use the structs and #defines in fsl_devices.h
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Now that the driver is exclusively an of_platform driver we no longer
use the structs and #defines in fsl_devices.h
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

On an SMP system, the following message is printed. The patch below gets
fixes the problem.

=================================
[ INFO: inconsistent lock state ]
2.6.29-Linus-05093-gc31f403d #57
---------------------------------
inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
bash/4105 [HC1[1]:SC0[0]:HE0:SE1] takes:
 (&kaweth->device_lock){+...}, at: [<ffffffffa01aa286>]
                 kaweth_usb_receive+0x77/0x1af [kaw eth]
{hardirq-on-W} state was registered at:
  [<ffffffff80260503>] __lock_acquire+0x753/0x1685
  [<ffffffff8026148a>] lock_acquire+0x55/0x71
  [<ffffffff80461ba6>] _spin_lock+0x31/0x3d
  [<ffffffffa01aaa0c>] kaweth_start_xmit+0x2b/0x1e1 [kaweth]
  [<ffffffff803eccd3>] dev_hard_start_xmit+0x22e/0x2ad
  [<ffffffff803fe120>] __qdisc_run+0xf2/0x203
  [<ffffffff803ed0cd>] dev_queue_xmit+0x263/0x39b
  [<ffffffffa03a47cb>] packet_sendmsg_spkt+0x1c4/0x20a [af_packet]
  [<ffffffff803de0c2>] sock_sendmsg+0xe4/0xfd
  [<ffffffff803dec8f>] sys_sendto+0xe4/0x10c
  [<ffffffff8020bccb>] system_call_fastpath+0x16/0x1b
  [<ffffffffffffffff>] 0xffffffffffffffff
irq event stamp: 1280
hardirqs last  enabled at (1279): [<ffffffff80461a71>]
                  _spin_unlock_irqrestore+0x44/0x4c
hardirqs last disabled at (1280): [<ffffffff8020bad7>]
                  save_args+0x67/0x70
softirqs last  enabled at (660): [<ffffffff8024192c>]
                  __do_softirq+0x14d/0x15d
softirqs last disabled at (651): [<ffffffff8020ce9c>]
                  call_softirq+0x1c/0x28
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

GRO assumes that there is a one-to-one relationship between NAPI
structure and network device. Some devices like sky2 share multiple
devices on a single interrupt so only have one NAPI handler. Rather than
split GRO from NAPI, just have GRO assume if device changes that
it is a different flow.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Original comment (Karsten):
On a MSI MS-6702E mainboard, when in rtl8169_init_one() for the first time
after BIOS has run, IntrStatus reads 5 after chip has been reset.
IntrStatus should equal 0 there, so patch changes IntrStatus reset to happen
after chip reset instead of before.

Remark (Francois):
Assuming that the loglevel of the driver is increased above NETIF_MSG_INTR,
the bug reveals itself with a typical "interrupt 0025 in poll" message
at startup. In retrospect, the message should had been read as an hint of
an unexpected hardware state several months ago :o(

Fixes (at least part of) https://bugzilla.redhat.com/show_bug.cgi?id=460747Signed-off-by: NKarsten Wiese <fzu@wemgehoertderstaat.de>
Signed-off-by: NFrancois Romieu <romieu@fr.zoreil.com>
Tested-by: NJosep <josep.puigdemont@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While setting up the ring parameters using ethtool the driver can
panic or leak memory as ixgbe_open tries to setup tx & rx resources.
The updated logic will use ixgbe_down/up after successful allocation of
tx & rx resources
Signed-off-by: NMallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com>
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
CC: stable@kernel.org
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We were basicly ignoring ethtool users request for FC autoneg
and replying to queries with a "best guess".  This patch
enables the driver to store if we want to enable/disable
autoneg FC and do the correct behavior.
Signed-off-by: NDon Skidmore <donald.c.skidmore@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When network interface is made active we were not handling the error
scenarios properly to clean up rx & tx resources which might result in
a driver panic.
Signed-off-by: NMallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com>
Acked-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch resolves an issue with map single being used to map a buffer and
then unmap page being used to unmap it.  In addition it handles any error
conditions that may be detected using skb_dma_map.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Acked-by: NMallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com>
Acked-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

MSI-X allocation broke after the 82599 merge on systems with more than 8
CPU cores.  82598 drops back into MSI mode, which isn't sufficient to run
full, efficient 10G line rate.
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In the past flow control wasn't enabled by default under the
incorrect assumption that this opened up us to a denial of
service attack.  However since any switch that forwarded flow
control would be extremely msiconfigured and/or buggy, this
concern no longer out weighs the preformance gains from
having FC enabled.
Signed-off-by: NDon Skidmore <donald.c.skidmore@intel.com>
Acked-by: NMallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com>
Acked-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The priority flow control settings from the netlink layer aren't taking
effect in the base driver.  The boolean pfc_mode_enable in the dcb_config
struct isn't being set, so the hardware configuration code is never
reached.
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Ethtool tries to get advertised speed from phy.autoneg_advertised.
However for copper media this wasn't happening until later do to
an other fix which moved mac.ops.setup_link_speed placement in
ixgbe_link_config(). This patch will display the default advertised
speeds if it can't yet get this information from phy.autoneg_advertised.
Signed-off-by: NDon Skidmore <donald.c.skidmore@intel.com>
Acked-by: NMallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com>
Acked-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The ixgbe driver had issues when DEBUG was defined because the hw_dbg macro
was incomplete.  This patch completes the code based off of the code that
already existed in the igb module.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Acked-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Impact: Fix

Yinghai Lu found one system with 82575EB where, in the kernel that is
kexeced, probe igb failed with -2, the reason being that the adapter
could not be brought back from D3 by the kexec kernel, most probably
due to quirky hardware (it looks like the same behavior happened on
forcedeth).

Prevent igb from putting the adapter into D3 during shutdown except
when we going to power off the system.  For this purpose, seperate
igb_shutdown() from igb_suspend() and use the appropriate PCI PM
callbacks in both of them.
Signed-off-by: N"Rafael J. Wysocki" <rjw@sisk.pl>
Reported-by: NYinghai Lu <yinghai@kernel.org>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The code path for setting up phy loopback testing was out of date and was
setting bits it didn't need to.  This change cleans up the code path and
removes some code that has no effect on teh driver.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Increase the delay for copper phy init from 15ms to 100ms.  This is to
address issues seen in which ethtool -t was failing in some cases on 82576
based adapters.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

An issue was seen on suspend in which the system reported a page fault.  This
was due to the new reg_idx code being called after the queues were freed.

This update prevents any for loops from going through the queues by setting
the number of queues to 0 when they are freed.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When device is on PCIe link trained as x2 the driver is currently reporting
link width as "unknown".  The original patch provided by Myron adds the x2
link support and my changes are cosmetic to clean up the readability of the
conditional operators.

Based on work by: Myron Stowe <myron.stowe@hp.com>
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The FEC Ethernet device isn't always attached to a phy.  Be careful
not to dereference phy_device if it is NULL.  Also eliminates an
unnecessary extra function from the ioctl path.
Reported-by: NHenk Stegeman <henk.stegeman@gmail.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since not using net_device_ops gets you shunned out the cool crowd,
this patch modifies the fec_mpc52xx Ethernet driver to provide the
management hooks via a struct net_device_ops.
Reported-by: NHenk Stegeman <henk.stegeman@gmail.com>
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The driver triggers a BUG_ON() when allocating DMA buffers because the
arch/powerpc dma_ops aren't in the net_device's struct device.  This
patch fixes the problem by using the parent of_device which does have
the correct dma_ops set.
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>
Reviewed-by: NBecky Bruce <beckyb@kernel.crashing.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 78454473
(netfilter: iptables: lock free counters) forgot to disable BH
in arpt_do_table(), ipt_do_table() and  ip6t_do_table()

Use rcu_read_lock_bh() instead of rcu_read_lock() cures the problem.
Reported-and-bisected-by: NRoman Mindalev <r000n@r000n.net>
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Acked-by: NPatrick McHardy <kaber@trash.net>
Acked-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>