openwrt/staging/blogic.git
12 years agocpsw: Hook up default ndo_change_mtu.
David S. Miller [Tue, 20 Mar 2012 04:33:59 +0000 (00:33 -0400)]
cpsw: Hook up default ndo_change_mtu.

Reported-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet: qmi_wwan: fix build error due to cdc-wdm dependecy
Bjørn Mork [Mon, 19 Mar 2012 07:48:29 +0000 (07:48 +0000)]
net: qmi_wwan: fix build error due to cdc-wdm dependecy

Fixes:

drivers/built-in.o: In function `qmi_wwan_bind_shared':
qmi_wwan.c:(.text+0x25b686): undefined reference to `usb_cdc_wdm_register'
make[1]: *** [.tmp_vmlinux1] Error 1

Reported-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Acked-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetdev: driver: ethernet: Add TI CPSW driver
Mugunthan V N [Sun, 18 Mar 2012 20:17:54 +0000 (20:17 +0000)]
netdev: driver: ethernet: Add TI CPSW driver

This patch adds support for TI's CPSW driver.

The three port switch gigabit ethernet subsystem provides ethernet packet
communication and can be configured as an ethernet switch. Supports
10/100/1000 Mbps.

Signed-off-by: Cyril Chemparathy <cyril@ti.com>
Signed-off-by: Sriramakrishnan A G <srk@ti.com>
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetdev: driver: ethernet: add cpsw address lookup engine support
Mugunthan V N [Sun, 18 Mar 2012 20:17:53 +0000 (20:17 +0000)]
netdev: driver: ethernet: add cpsw address lookup engine support

TI CPSW ethernet switch has a built-in address lookup engine.  This patch adds
the code necessary for programming this module.

Signed-off-by: Cyril Chemparathy <cyril@ti.com>
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agophy: add am79c874 PHY support
Heiko Schocher [Sun, 18 Mar 2012 11:03:05 +0000 (11:03 +0000)]
phy: add am79c874 PHY support

Signed-off-by: Heiko Schocher <hs@denx.de>
Signed-off-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agomlx4_core: fix race on comm channel
Eugenia Emantayev [Sun, 18 Mar 2012 04:32:08 +0000 (04:32 +0000)]
mlx4_core: fix race on comm channel

Prevent race condition between commands on comm channel.
Happened while unloading the driver when switching from
event to polling mode. VF got completion on the last command
before switching to polling mode, but toggle was not changed.
After the fix - VF will not write the next command before
toggle is updated.

Signed-off-by: Eugenia Emantayev <eugenia@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobonding: send igmp report for its master
Peter Pan(潘卫平) [Sat, 17 Mar 2012 17:23:27 +0000 (17:23 +0000)]
bonding: send igmp report for its master

Liang Zheng(lzheng@redhat.com) found that in the following topo,
bonding does not send igmp report when we trigger a fail-over of bonding.

eth0--
      |-- bond0 -- br0
eth1--

modprobe bonding mode=1 miimon=100 resend_igmp=10
ifconfig bond0 up
ifenslave bond0 eth0 eth1

brctl addbr br0
ifconfig br0 192.168.100.2/24 up
brctl addif br0 bond0

Add 192.168.100.2(br0) into a multicast group, like 224.10.10.10,
then trigger a fali-over in bonding.
You can see that parameter "resend_igmp" does not work.

The reason is that when we add br0 into a multicast group,
it does not propagate multicast knowledge down to its ports.

If we choose to propagate multicast knowledge down to all ports for bridge,
then we have to track every change that is done to bridge, and keep a backup
for all ports. It is hard to track, I think.

Instead I choose to modify bonding to send igmp report for its master.

Changelog:
V2: correct comments
V3: move this check into bond_resend_igmp_join_requests()
V4: only send igmp reports if bond is enslaved to a bridge

Signed-off-by: Weiping Pan <panweiping3@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agofs_enet: Add MPC5125 FEC support and PHY interface selection
Vladimir Ermakov [Sat, 17 Mar 2012 13:10:50 +0000 (13:10 +0000)]
fs_enet: Add MPC5125 FEC support and PHY interface selection

Add compatible string for MPC5125 FEC. The FEC on MPC5125 additionally
supports RMII PHY interface. Configure controller/PHY interface type
according to the optional phy-connection-type property in the ethernet
node. This property should be either "rmii" or "mii".

Signed-off-by: Vladimir Ermakov <vooon341@gmail.com>
Signed-off-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet: bpf_jit: fix BPF_S_LDX_B_MSH compilation
Eric Dumazet [Sun, 18 Mar 2012 02:40:48 +0000 (02:40 +0000)]
net: bpf_jit: fix BPF_S_LDX_B_MSH compilation

Matt Evans spotted that x86 bpf_jit was incorrectly handling negative
constant offsets in BPF_S_LDX_B_MSH instruction.

We need to abort JIT compilation like we do in common_load so that
filter uses the interpreter code and can call __load_pointer()

Reference: http://lists.openwall.net/netdev/2011/07/19/11

Thanks to Indan Zupancic to bring back this issue.

Reported-by: Matt Evans <matt@ozlabs.org>
Reported-by: Indan Zupancic <indan@nul.nu>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet: update the usage of CHECKSUM_UNNECESSARY
Yi Zou [Mon, 19 Mar 2012 11:12:41 +0000 (11:12 +0000)]
net: update the usage of CHECKSUM_UNNECESSARY

As suggested by Ben, this adds the clarification on the usage of
CHECKSUM_UNNECESSARY on the outgoing patch. Also add the usage
description of NETIF_F_FCOE_CRC and CHECKSUM_UNNECESSARY
for the kernel FCoE protocol driver.

This is a follow-up to the following:
http://patchwork.ozlabs.org/patch/147315/

Signed-off-by: Yi Zou <yi.zou@intel.com>
Cc: Ben Hutchings <bhutchings@solarflare.com>
Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Cc: www.Open-FCoE.org <devel@open-fcoe.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agofcoe: use CHECKSUM_UNNECESSARY instead of CHECKSUM_PARTIAL on tx
Yi Zou [Fri, 16 Mar 2012 23:08:12 +0000 (23:08 +0000)]
fcoe: use CHECKSUM_UNNECESSARY instead of CHECKSUM_PARTIAL on tx

Fix a bug when using 'ethtool -K ethx tx off' to turn off tx ip checksum,
FCoE CRC offload should not be impacte. The skb_checksum_help() is needed
only if it's not FCoE traffic for ip checksum, regardless of ethtool toggling
the tx ip checksum on or off. Instead of using CHECKSUM_PARTIAL, we will
use CHECKSUM_UNNECESSARY as a proper indication to avoid sw ip checksum
on FCoE frames.

Ref. to original discussion thread:
http://patchwork.ozlabs.org/patch/146567/

CC: "James E.J. Bottomley" <JBottomley@parallels.com>
CC: Robert Love <robert.w.love@intel.com>
Signed-off-by: Yi Zou <yi.zou@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet: do not do gso for CHECKSUM_UNNECESSARY in netif_needs_gso
Yi Zou [Fri, 16 Mar 2012 23:08:11 +0000 (23:08 +0000)]
net: do not do gso for CHECKSUM_UNNECESSARY in netif_needs_gso

This is related to fixing the bug of dropping FCoE frames when disabling tx ip
checksum by 'ethtool -K ethx tx off'. The FCoE protocol stack driver would
use CHECKSUM_UNNECESSARY on tx path instead of CHECKSUM_PARTIAL (as indicated in
the 2/2 of this series). To do so, netif_needs_gso() has to be changed here to
not do gso for both CHECKSUM_PARTIAL and CHECKSUM_UNNECESSARY.

Ref. to original discussion thread:
http://patchwork.ozlabs.org/patch/146567/

Signed-off-by: Yi Zou <yi.zou@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoixgbe: Fix issues with SR-IOV loopback when flow control is disabled
Alexander Duyck [Fri, 16 Mar 2012 23:07:34 +0000 (23:07 +0000)]
ixgbe: Fix issues with SR-IOV loopback when flow control is disabled

This patch allows us to avoid a Tx hang when SR-IOV is enabled.  This hang
can be triggered by sending small packets at a rate that was triggering Rx
missed errors from the adapter while the internal Tx switch and at least
one VF are enabled.

This was all due to the fact that under heavy stress the Rx FIFO never
drained below the flow control high water mark.  This resulted in the Tx
FIFO being head of line blocked due to the fact that it relies on the flow
control high water mark to determine when it is acceptable for the Tx to
place a packet in the Rx FIFO.

The resolution for this is to set the FCRTH value to the RXPBSIZE - 32 so
that even if the ring is almost completely full we can still place Tx
packets on the Rx ring and drop incoming Rx traffic if we do not have
sufficient space available in the Rx FIFO.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet/hyperv: Fix the code handling tx busy
Haiyang Zhang [Mon, 19 Mar 2012 21:27:06 +0000 (17:27 -0400)]
net/hyperv: Fix the code handling tx busy

Instead of dropping the packet, we keep the skb buffer, and return
NETDEV_TX_BUSY to let upper layer retry send. This will not cause
endless loop, because the host is taking data away from ring buffer,
and we have called the stop_queue before returning NETDEV_TX_BUSY.

The stop_queue was called in the function netvsc_send() in file
netvsc.c, then it returns to rndis_filter_send(), which returns to
netvsc_start_xmit() in file netvsc_drv.c. So the NETDEV_TX_BUSY is
indeed returned AFTER queue is stopped.

Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: K. Y. Srinivasan <kys@microsoft.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net...
David S. Miller [Mon, 19 Mar 2012 21:24:27 +0000 (17:24 -0400)]
Merge branch 'master' of git://git./linux/kernel/git/jkirsher/net-next

12 years agoixgbe: fix namespace issues when FCoE/DCB is not enabled
Jeff Kirsher [Sat, 18 Feb 2012 07:08:14 +0000 (07:08 +0000)]
ixgbe: fix namespace issues when FCoE/DCB is not enabled

Resolve namespace issues when FCoE or DCB is not enabled.
The issue is with certain configurations we end up with namespace
problems. A simple example:

ixgbe_main.c
 - defines func A()
 - uses func A()

ixgbe_fcoe.c
 - uses func A()

ixgbe.h
 - has prototype for func A()

For default (FCoE included) all is good.  But when it isn't the namespace
checker complains about how func A() could be static.

To resolve this, created a ixgbe_lib file to contain functions used
by DCB/FCoE and their helper functions so that they are always in
namespace whether or not DCB/FCoE is enabled.

Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
12 years agortlwifi: Remove unused ETH_ADDR_LEN defines
Joe Perches [Sun, 18 Mar 2012 17:37:59 +0000 (17:37 +0000)]
rtlwifi: Remove unused ETH_ADDR_LEN defines

Just neatening.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoigbvf: Use ETH_ALEN
Joe Perches [Sun, 18 Mar 2012 17:37:58 +0000 (17:37 +0000)]
igbvf: Use ETH_ALEN

Remove an unnecessary #define and use memcpy
instead of a loop to copy an ethernet address.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoatlx: Use ETH_ALEN
Joe Perches [Sun, 18 Mar 2012 17:37:57 +0000 (17:37 +0000)]
atlx: Use ETH_ALEN

No need for yet another #define for this.

Convert NODE_ADDRESS_SIZE use to ETH_ALEN and remove #define.
Use memcpy instead of a loop to copy an address.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoif_vlan: Remove VLAN_ETH_ALEN define and the 1 use of it
Joe Perches [Sun, 18 Mar 2012 17:37:56 +0000 (17:37 +0000)]
if_vlan: Remove VLAN_ETH_ALEN define and the 1 use of it

Just use ETH_ALEN.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agotcp: reduce out_of_order memory use
Eric Dumazet [Sun, 18 Mar 2012 11:07:47 +0000 (11:07 +0000)]
tcp: reduce out_of_order memory use

With increasing receive window sizes, but speed of light not improved
that much, out of order queue can contain a huge number of skbs, waiting
to be moved to receive_queue when missing packets can fill the holes.

Some devices happen to use fat skbs (truesize of 4096 + sizeof(struct
sk_buff)) to store regular (MTU <= 1500) frames. This makes highly
probable sk_rmem_alloc hits sk_rcvbuf limit, which can be 4Mbytes in
many cases.

When limit is hit, tcp stack calls tcp_collapse_ofo_queue(), a true
latency killer and cpu cache blower.

Doing the coalescing attempt each time we add a frame in ofo queue
permits to keep memory use tight and in many cases avoid the
tcp_collapse() thing later.

Tested on various wireless setups (b43, ath9k, ...) known to use big skb
truesize, this patch removed the "packets collapsed in receive queue due
to low socket buffer" I had before.

This also reduced average memory used by tcp sockets.

With help from Neal Cardwell.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: H.K. Jerry Chu <hkchu@google.com>
Cc: Tom Herbert <therbert@google.com>
Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agotcp: introduce tcp_data_queue_ofo
Eric Dumazet [Sun, 18 Mar 2012 11:06:44 +0000 (11:06 +0000)]
tcp: introduce tcp_data_queue_ofo

Split tcp_data_queue() in two parts for better readability.

tcp_data_queue_ofo() is responsible for queueing incoming skb into out
of order queue.

Change code layout so that the skb_set_owner_r() is performed only if
skb is not dropped.

This is a preliminary patch before "reduce out_of_order memory use"
following patch.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: H.K. Jerry Chu <hkchu@google.com>
Cc: Tom Herbert <therbert@google.com>
Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: validate FW trace prior to its printing
Dmitry Kravkov [Sun, 18 Mar 2012 10:33:45 +0000 (10:33 +0000)]
bnx2x: validate FW trace prior to its printing

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: consistent statistics for old FW
Yuval Mintz [Sun, 18 Mar 2012 10:33:44 +0000 (10:33 +0000)]
bnx2x: consistent statistics for old FW

Previously applied patch making the bnx2x statistics consistent
did not apply to old FWs. This remedies it, extending the consistent
behaviour to all drivers.

Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Reported-by: Michal Schmidt <mschmidt@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: changed iscsi/fcoe mac init and macros
Dmitry Kravkov [Sun, 18 Mar 2012 10:33:43 +0000 (10:33 +0000)]
bnx2x: changed iscsi/fcoe mac init and macros

This includes changes in macros to better distinguish between the two
protocols, and slightly changed the way their macs are set.
Notice this file contains string print lines with more than 80 characters,
as to not break prints.

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: added TLV_NOT_FOUND flags to the dcb
Dmitry Kravkov [Sun, 18 Mar 2012 10:33:42 +0000 (10:33 +0000)]
bnx2x: added TLV_NOT_FOUND flags to the dcb

The new error flags are supported by the bnx2x FW.

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: changed initial dcb configuration
Dmitry Kravkov [Sun, 18 Mar 2012 10:33:41 +0000 (10:33 +0000)]
bnx2x: changed initial dcb configuration

The changes were mostly made to enable back-to-back data flow with dcb.
Other changes were simply deemed as a better 'clean' initial configuration.

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: removed dcb unused code
Yuval Mintz [Sun, 18 Mar 2012 10:33:40 +0000 (10:33 +0000)]
bnx2x: removed dcb unused code

Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: reduced sparse warnings
Yuval Mintz [Sun, 18 Mar 2012 10:33:39 +0000 (10:33 +0000)]
bnx2x: reduced sparse warnings

This patch reduces sparse warnings in the bnx2x code,
mostly by changing functions into static and changing
initialization of structures.

Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: revised driver prints
Merav Sicron [Sun, 18 Mar 2012 10:33:38 +0000 (10:33 +0000)]
bnx2x: revised driver prints

We've revised driver prints, changing the mask of existing prints
to allow better control over the debug messages, added prints to
error scenarios, removed unnecessary prints and corrected some spelling.
Please note that this patch contains lines with over 80 characters,
as string messages were kept in a single line.

Signed-off-by: Merav Sicron <meravs@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: added 'likely' to fast-path skb existence
Yuval Mintz [Sun, 18 Mar 2012 10:33:37 +0000 (10:33 +0000)]
bnx2x: added 'likely' to fast-path skb existence

Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobe2net: fix programming of VLAN tags for VF
Ajit Khaparde [Sun, 18 Mar 2012 06:23:41 +0000 (06:23 +0000)]
be2net: fix programming of VLAN tags for VF

Signed-off-by: Ajit Khaparde <ajit.khaparde@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobe2net: Fix number of vlan slots in flex mode
Ajit Khaparde [Sun, 18 Mar 2012 06:23:31 +0000 (06:23 +0000)]
be2net: Fix number of vlan slots in flex mode

In flex10 mode the number of vlan slots supported is halved.

Signed-off-by: Ajit Khaparde <ajit.khaparde@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobe2net: Program secondary UC MAC address into MAC filter
Ajit Khaparde [Sun, 18 Mar 2012 06:23:21 +0000 (06:23 +0000)]
be2net: Program secondary UC MAC address into MAC filter

Signed-off-by: Ajit Khaparde <ajit.khaparde@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobe2net: enable WOL by default if h/w supports it
Ajit Khaparde [Sun, 18 Mar 2012 06:23:11 +0000 (06:23 +0000)]
be2net: enable WOL by default if h/w supports it

Signed-off-by: Ajit Khaparde <ajit.khaparde@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge branch 'gianfar-bql' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg...
David S. Miller [Mon, 19 Mar 2012 20:46:22 +0000 (16:46 -0400)]
Merge branch 'gianfar-bql' of git://git./linux/kernel/git/paulg/linux

12 years agoixgbe: Correct flag values set by ixgbe_fix_features
Alexander Duyck [Sat, 11 Feb 2012 07:18:57 +0000 (07:18 +0000)]
ixgbe: Correct flag values set by ixgbe_fix_features

This patch replaces the variable name data with the variable name features
for ixgbe_fix_features and ixgbe_set_features.  This helps to make some
issues more obvious such as the fact that we were disabling Rx VLAN tag
stripping when we should have been forcing it to be enabled when DCB is
enabled.

In addition there was deprecated code present that was disabling the LRO
flag if we had the itr value set too low.  I have updated this logic so
that we will now allow the LRO flag to be set, but will not enable RSC
until the rx-usecs value is high enough to allow enough time for Rx packet
coalescing.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Add support for enabling UDP RSS via the ethtool rx-flow-hash command
Alexander Duyck [Wed, 8 Feb 2012 07:51:53 +0000 (07:51 +0000)]
ixgbe: Add support for enabling UDP RSS via the ethtool rx-flow-hash command

This patch adds support for enabling or disabling UDP RSS via the
ethtool -N rx-flow-hash command.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Whitespace cleanups
Alexander Duyck [Wed, 8 Feb 2012 07:51:47 +0000 (07:51 +0000)]
ixgbe: Whitespace cleanups

This patch contains several fixes for formatting in regards to whitespace.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Two minor fixes for RSS and FDIR set queues functions
Alexander Duyck [Wed, 8 Feb 2012 07:51:42 +0000 (07:51 +0000)]
ixgbe: Two minor fixes for RSS and FDIR set queues functions

This change fixes two minor issues. The first was the fact that we were
setting the return value to false twice in the set_rss_queues function.
The second is the fact that we should have been using "min_t(int," instead
of "min((int)" in set_fdir_queues.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: drop err_eeprom tag which is at same location as err_sw_init
Alexander Duyck [Wed, 8 Feb 2012 07:51:37 +0000 (07:51 +0000)]
ixgbe: drop err_eeprom tag which is at same location as err_sw_init

The err_eeprom and err_sw_init tags both go to the same location.  So
instead of maintaining two tags this patch combines them so we only use
err_sw_init.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Move poll routine in order to improve readability
Alexander Duyck [Wed, 8 Feb 2012 07:51:27 +0000 (07:51 +0000)]
ixgbe: Move poll routine in order to improve readability

This change relocates the ixgbe_poll routine so it is right next to the
interrupt routine that schedules and calls it.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: cleanup logic for the service timer and VF hang detection
Alexander Duyck [Wed, 8 Feb 2012 07:51:22 +0000 (07:51 +0000)]
ixgbe: cleanup logic for the service timer and VF hang detection

This change just cleans up some of the logic in the service_timer function
so that we can avoid unnecessary swapping of the ready value between true to
false and back to true.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Update layout of ixgbe_ring structure to improve cache performance
Alexander Duyck [Wed, 8 Feb 2012 07:51:16 +0000 (07:51 +0000)]
ixgbe: Update layout of ixgbe_ring structure to improve cache performance

This change makes it so that only the 2nd cache line in the ring structure
should see frequent updates.  The advantage to this is that it should
reduce the amount of cross CPU cache bouncing since only the 2nd cache line
will be changing between most network transactions.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Store Tx flags and protocol information to tx_buffer sooner
Alexander Duyck [Wed, 8 Feb 2012 07:51:11 +0000 (07:51 +0000)]
ixgbe: Store Tx flags and protocol information to tx_buffer sooner

This change makes it so that we store the tx_flags and protocol information
to the tx_buffer_info structure sooner. This allows us to avoid unnecessary
read/write transactions since we are placing the data in the final location
earlier.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
David S. Miller [Mon, 19 Mar 2012 03:29:41 +0000 (23:29 -0400)]
Merge git://git./linux/kernel/git/davem/net

12 years agoLinux 3.3
Linus Torvalds [Sun, 18 Mar 2012 23:15:34 +0000 (16:15 -0700)]
Linux 3.3

12 years agogianfar: use netif_tx_queue_stopped instead of __netif_subqueue_stopped
Paul Gortmaker [Sun, 18 Mar 2012 21:11:22 +0000 (17:11 -0400)]
gianfar: use netif_tx_queue_stopped instead of __netif_subqueue_stopped

The __netif_subqueue_stopped() just does the following:

        struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
        return netif_tx_queue_stopped(txq);

and since we already have the txq in scope, we can just call that
directly in this case.

Suggested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
12 years agoDon't limit non-nested epoll paths
Jason Baron [Fri, 16 Mar 2012 20:34:03 +0000 (16:34 -0400)]
Don't limit non-nested epoll paths

Commit 28d82dc1c4ed ("epoll: limit paths") that I did to limit the
number of possible wakeup paths in epoll is causing a few applications
to longer work (dovecot for one).

The original patch is really about limiting the amount of epoll nesting
(since epoll fds can be attached to other fds). Thus, we probably can
allow an unlimited number of paths of depth 1. My current patch limits
it at 1000. And enforce the limits on paths that have a greater depth.

This is captured in: https://bugzilla.redhat.com/show_bug.cgi?id=681578

Signed-off-by: Jason Baron <jbaron@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Linus Torvalds [Sun, 18 Mar 2012 02:22:24 +0000 (19:22 -0700)]
Merge git://git./linux/kernel/git/davem/net

Pull networking changes from David Miller:
 "1) icmp6_dst_alloc() returns NULL instead of ERR_PTR() leading to
     crashes, particularly during shutdown.  Reported by Dave Jones and
     fixed by Eric Dumazet.

  2) hyperv and wimax/i2400m return NETDEV_TX_BUSY when they have
     already freed the SKB, which causes crashes as to the caller this
     means requeue the packet.  Fixes from Eric Dumazet.

  3) usbnet driver doesn't allocate the right amount of headroom on
     fresh RX SKBs, fix from Eric Dumazet.

  4) Fix regression in ip6_mc_find_dev_rcu(), as an RCU lookup it
     abolutely should not take a reference to 'dev', this leads to
     leaks.  Fix from RonQing Li.

  5) Fix netfilter ctnetlink race between delete and timeout expiration.
     From Pablo Neira Ayuso.

  6) Revert SFQ change which causes regressions, specifically queueing
     to tail can lead to unavoidable flow starvation.  From Eric
     Dumazet.

  7) Fix a memory leak and a crash on corrupt firmware files in bnx2x,
     from Michal Schmidt."

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net:
  netfilter: ctnetlink: fix race between delete and timeout expiration
  ipv6: Don't dev_hold(dev) in ip6_mc_find_dev_rcu.
  wimax/i2400m: fix erroneous NETDEV_TX_BUSY use
  net/hyperv: fix erroneous NETDEV_TX_BUSY use
  net/usbnet: reserve headroom on rx skbs
  bnx2x: fix memory leak in bnx2x_init_firmware()
  bnx2x: fix a crash on corrupt firmware file
  sch_sfq: revert dont put new flow at the end of flows
  ipv6: fix icmp6_dst_alloc()

12 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 17 Mar 2012 16:54:16 +0000 (09:54 -0700)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull perf fixes from Ingo Molnar.

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf tools, x86: Build perf on older user-space as well
  perf tools: Use scnprintf where applicable
  perf tools: Incorrect use of snprintf results in SEGV

12 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net...
David S. Miller [Sat, 17 Mar 2012 09:02:26 +0000 (02:02 -0700)]
Merge branch 'master' of git://git./linux/kernel/git/jkirsher/net-next

12 years agonetfilter: ctnetlink: fix race between delete and timeout expiration
Pablo Neira Ayuso [Fri, 16 Mar 2012 02:00:34 +0000 (02:00 +0000)]
netfilter: ctnetlink: fix race between delete and timeout expiration

Kerin Millar reported hardlockups while running `conntrackd -c'
in a busy firewall. That system (with several processors) was
acting as backup in a primary-backup setup.

After several tries, I found a race condition between the deletion
operation of ctnetlink and timeout expiration. This patch fixes
this problem.

Tested-by: Kerin Millar <kerframil@gmail.com>
Reported-by: Kerin Millar <kerframil@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoixgbe: always write DMA for single_mapped value with skb
Alexander Duyck [Wed, 8 Feb 2012 07:51:06 +0000 (07:51 +0000)]
ixgbe: always write DMA for single_mapped value with skb

This change makes it so that we always write the DMA address for the skb
itself on the same tx_buffer struct that the skb is written on.  This way
we don't need the MAPPED_AS_PAGE flag and we always know it will be the
first DMA value that we will have to unmap.

In addition I have found an issue in which we were leaking a DMA mapping if
the value happened to be 0 which is possible on some platforms.  In order
to resolve that I have updated the transmit path to use the length instead
of the DMA mapping in order to determine if a mapping is actually present.

One other tweak in this patch is that it only writes the olinfo information
on the first descriptor.  As it turns out it isn't necessary to write it
for anything but the first descriptor so there is no need to carry it
forward.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agophc: Update author's email address.
Richard Cochran [Fri, 16 Mar 2012 22:39:29 +0000 (22:39 +0000)]
phc: Update author's email address.

This commit brings the author email address macros up to date for four
modules in the PTP Hardware Clock subsystem.

Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoixgbe: Write gso_segs and bytcount to the ring sooner
Alexander Duyck [Wed, 8 Feb 2012 07:51:01 +0000 (07:51 +0000)]
ixgbe: Write gso_segs and bytcount to the ring sooner

This change makes it so that gso_segs and bytecount are written to the ring
sooner.  This helps to simplify the logic for the two since segmentation
offloads can now update them within their own function.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Place skb on first buffer_info structure to avoid using stack space
Alexander Duyck [Wed, 8 Feb 2012 07:50:56 +0000 (07:50 +0000)]
ixgbe: Place skb on first buffer_info structure to avoid using stack space

Instead of keeping a local copy of the skb on the stack for as long as long
as we do it makes sense to instead just place it on the first tx_buffer
structure so that we can save space on the stack and avoid unnecessary
read/write operations copying the pointer out of the stack and onto the
ring later.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Use packets to track Tx completions instead of a seperate value
Alexander Duyck [Wed, 8 Feb 2012 07:50:51 +0000 (07:50 +0000)]
ixgbe: Use packets to track Tx completions instead of a seperate value

A separate value was added to track Tx completions in order to determine if
the Tx unit was hung.  However we can do the same thing using the number of
packets completed without having to add another stat to the Tx ring.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Modify setup of descriptor flags to avoid conditional jumps
Alexander Duyck [Wed, 8 Feb 2012 07:50:45 +0000 (07:50 +0000)]
ixgbe: Modify setup of descriptor flags to avoid conditional jumps

This change makes it more likely that the descriptor flags setup will use
cmov instructions instead of conditional jumps when setting up the flags.
The advantage to this is that the code should just flow a bit more
smoothly.

To do this it is necessary to set the TX_FLAGS_CSUM bit in tx_flags when
doing TSO so that we also do the checksum in addition to the segmentation
offload.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Make certain that all frames fit minimum size requirements
Alexander Duyck [Wed, 8 Feb 2012 07:50:40 +0000 (07:50 +0000)]
ixgbe: Make certain that all frames fit minimum size requirements

This change makes certain that any packet we attempt to transmit will meet
minimum size requirements for the hardware.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: cleanup logic in ixgbe_change_mtu
Alexander Duyck [Wed, 8 Feb 2012 07:50:35 +0000 (07:50 +0000)]
ixgbe: cleanup logic in ixgbe_change_mtu

This change is meant to just cleanup the logic in ixgbe_change_mtu since we
are making it unnecessarily complex due to a workaround required for 82599
when SR-IOV is enabled.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Replace standard receive path with a page based receive
Alexander Duyck [Sat, 3 Mar 2012 02:35:52 +0000 (02:35 +0000)]
ixgbe: Replace standard receive path with a page based receive

This patch replaces the existing Rx hot-path in the ixgbe driver with a new
implementation that is based on performing a double buffered receive.  The
ixgbe driver already had something similar in place for its' packet split
path, however in that case we were still receiving the header for the
packet into the sk_buff.  The big change here is the entire receive path
will receive into pages only, and then pull the header out of the page and
copy it into the sk_buff data.  There are several motivations behind this
approach.

First, this allows us to avoid several cache misses as we were taking a
set of cache misses for allocating the sk_buff and then another set for
receiving data into the sk_buff.  We are able to avoid these misses on
receive now as we allocate the sk_buff when data is available.

Second we are able to see a considerable performance gain when an IOMMU is
enabled because we are no longer unmapping every buffer on receive.
Instead we can delay the unmap until we are unable to use the page, and
instead we can simply call sync_single_range on the half of the page that
contains new data.

Finally we are able to drop a considerable amount of code from the driver
as we no longer have to support 2 different receive modes, packet split and
one buffer.  This allows us to optimize the Rx path further since less
branching is required.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Support RX-ALL feature flag.
Ben Greear [Thu, 8 Mar 2012 08:28:41 +0000 (08:28 +0000)]
ixgbe: Support RX-ALL feature flag.

This allows the NIC to receive all frames available, including
those with bad FCS, ethernet control frames, and more.

Tested by sending frames with bad FCS.

Signed-off-by: Ben Greear <greearb@candelatech.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: Support sending custom Ethernet FCS.
Ben Greear [Tue, 6 Mar 2012 09:42:04 +0000 (09:42 +0000)]
ixgbe: Support sending custom Ethernet FCS.

Including bad FCS, used generate frames with bad FCS
to test other system's handling of RX of bad packets.

Signed-off-by: Ben Greear <greearb@candelatech.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoigb: Support RX-ALL feature flag.
Ben Greear [Tue, 6 Mar 2012 09:41:58 +0000 (09:41 +0000)]
igb: Support RX-ALL feature flag.

This allows the NIC to receive all frames available, including
those with bad FCS, un-matched vlans, ethernet control frames,
and more.

Tested by sending frames with bad FCS.

Signed-off-by: Ben Greear <greearb@candelatech.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoigb: Support sending custom Ethernet FCS.
Ben Greear [Tue, 6 Mar 2012 09:41:53 +0000 (09:41 +0000)]
igb: Support sending custom Ethernet FCS.

Including bad FCS, used generate frames with bad FCS
to test other system's handling of RX of bad packets.

Signed-off-by: Ben Greear <greearb@candelatech.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoeni: fix driver remove function and driver probe error path.
françois romieu [Fri, 16 Mar 2012 01:52:04 +0000 (01:52 +0000)]
eni: fix driver remove function and driver probe error path.

- add eni_do_release() to balance eni_do_init
- turn the zeroes DMA area into a per device data

Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet/irda: add clk_prepare/clk_unprepare to pxaficp_ir
Philipp Zabel [Thu, 15 Mar 2012 08:19:29 +0000 (08:19 +0000)]
net/irda: add clk_prepare/clk_unprepare to pxaficp_ir

This patch adds clk_prepare/clk_unprepare calls to the pxaficp_ir
driver by using the helper functions clk_prepare_enable and
clk_disable_unprepare.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Cc: Eric Miao <eric.y.miao@gmail.com>
Cc: Haojian Zhuang <haojian.zhuang@marvell.com>
Cc: Samuel Ortiz <samuel@sortiz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoarp: allow arp processing to honor per interface arp_accept sysctl
Neil Horman [Thu, 15 Mar 2012 05:25:58 +0000 (05:25 +0000)]
arp: allow arp processing to honor per interface arp_accept sysctl

I found recently that the arp_process function which handles all of our received
arp frames, is using IPV4_DEVCONF_ALL macro to check the state of the arp_process
flag.  This seems wrong, as it implies that either none or all of the network
interfaces accept gratuitous arps.  This patch corrects that, allowing
per-interface arp_accept configuration to deviate from the all setting.  Note
this also brings us into line with the way the arp_filter setting is handled
during arp_process execution.

Tested this myself on my home network, and confirmed it works as expected.

Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
CC: "David S. Miller" <davem@davemloft.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agousbnet: use netif_tx_wake_queue instead of netif_start_queue
Alexey Orishko [Wed, 14 Mar 2012 04:00:24 +0000 (04:00 +0000)]
usbnet: use netif_tx_wake_queue instead of netif_start_queue

If host is going to autosuspend function with two interfaces and
if IP packet has arrived in-between of two usbnet_suspend() callbacks,
i.e usbnet_resume() is called in-between, tx data flow is stopped.
When autosuspend timer expires and device is put to autosuspend
again, tx queue is waked up and data can be sent again.
This behavior might be repeated several times in a row.

Tested on Intel/ARM.

Reviewed-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Tested-by: Dmitry Tarnyagin <dmitry.tarnyagin@stericsson.com>
Signed-off-by: Alexey Orishko <alexey.orishko@stericsson.com>
Acked-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agostmmac: Add device-tree support
Stefan Roese [Tue, 13 Mar 2012 04:56:37 +0000 (04:56 +0000)]
stmmac: Add device-tree support

This patch adds support to configure the STMMAC ethernet driver via
device-tree instead of platform_data.

Currently, only the properties needed on SPEAr600 are provided. All
other properties should be added once needed on other platforms.

Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetxen: qlogic ethernet : Fix endian bug.
Santosh Nayak [Mon, 12 Mar 2012 22:58:24 +0000 (22:58 +0000)]
netxen: qlogic ethernet : Fix endian bug.

Change the datatype of "ip_addr" to __be32 as 'ip' should be in
big endian format.

Adapter needs "ip address" in big endian format stored at lower 32bit
of req.word[1]. netxen_config_ipaddr() now receives 'ip' in big endian
format. To satisfy adapter's need, use memcpy() to copy byte by byte
of 'ip' into lower 32bit of req.word[1].

Mac address and serial number of adapter need to be in little endian format.
Change the data type of the related  variables to __le32 / __le64 or cast it
explicitly to __le32 / __le64 depending upon the requirement.

Signed-off-by: Santosh Nayak <santoshprasadnayak@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoipv6: Don't dev_hold(dev) in ip6_mc_find_dev_rcu.
RongQing.Li [Thu, 15 Mar 2012 22:54:14 +0000 (22:54 +0000)]
ipv6: Don't dev_hold(dev) in ip6_mc_find_dev_rcu.

ip6_mc_find_dev_rcu() is called with rcu_read_lock(), so don't
need to dev_hold().
With dev_hold(), not corresponding dev_put(), will lead to leak.

[ bug introduced in 96b52e61be1 (ipv6: mcast: RCU conversions) ]

Signed-off-by: RongQing.Li <roy.qing.li@gmail.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agopxa168: remove unused stats member.
Rami Rosen [Fri, 16 Mar 2012 05:19:36 +0000 (05:19 +0000)]
pxa168: remove unused stats member.

The patch removes unused stats member in pxa168 network driver.

Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge branch 'akpm' (more patches from Andrew)
Linus Torvalds [Sat, 17 Mar 2012 00:14:55 +0000 (17:14 -0700)]
Merge branch 'akpm' (more patches from Andrew)

Merge some more email patches from Andrew Morton:
 "A couple of nilfs fixes"

* emailed from Andrew Morton <akpm@linux-foundation.org>:
  nilfs2: fix NULL pointer dereference in nilfs_load_super_block()
  nilfs2: clamp ns_r_segments_percentage to [1, 99]

12 years agonilfs2: fix NULL pointer dereference in nilfs_load_super_block()
Ryusuke Konishi [Sat, 17 Mar 2012 00:08:39 +0000 (17:08 -0700)]
nilfs2: fix NULL pointer dereference in nilfs_load_super_block()

According to the report from Slicky Devil, nilfs caused kernel oops at
nilfs_load_super_block function during mount after he shrank the
partition without resizing the filesystem:

 BUG: unable to handle kernel NULL pointer dereference at 00000048
 IP: [<d0d7a08e>] nilfs_load_super_block+0x17e/0x280 [nilfs2]
 *pde = 00000000
 Oops: 0000 [#1] PREEMPT SMP
 ...
 Call Trace:
  [<d0d7a87b>] init_nilfs+0x4b/0x2e0 [nilfs2]
  [<d0d6f707>] nilfs_mount+0x447/0x5b0 [nilfs2]
  [<c0226636>] mount_fs+0x36/0x180
  [<c023d961>] vfs_kern_mount+0x51/0xa0
  [<c023ddae>] do_kern_mount+0x3e/0xe0
  [<c023f189>] do_mount+0x169/0x700
  [<c023fa9b>] sys_mount+0x6b/0xa0
  [<c04abd1f>] sysenter_do_call+0x12/0x28
 Code: 53 18 8b 43 20 89 4b 18 8b 4b 24 89 53 1c 89 43 24 89 4b 20 8b 43
 20 c7 43 2c 00 00 00 00 23 75 e8 8b 50 68 89 53 28 8b 54 b3 20 <8b> 72
 48 8b 7a 4c 8b 55 08 89 b3 84 00 00 00 89 bb 88 00 00 00
 EIP: [<d0d7a08e>] nilfs_load_super_block+0x17e/0x280 [nilfs2] SS:ESP 0068:ca9bbdcc
 CR2: 0000000000000048

This turned out due to a defect in an error path which runs if the
calculated location of the secondary super block was invalid.

This patch fixes it and eliminates the reported oops.

Reported-by: Slicky Devil <slicky.dvl@gmail.com>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Tested-by: Slicky Devil <slicky.dvl@gmail.com>
Cc: <stable@vger.kernel.org> [2.6.30+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agonilfs2: clamp ns_r_segments_percentage to [1, 99]
Haogang Chen [Sat, 17 Mar 2012 00:08:38 +0000 (17:08 -0700)]
nilfs2: clamp ns_r_segments_percentage to [1, 99]

ns_r_segments_percentage is read from the disk.  Bogus or malicious
value could cause integer overflow and malfunction due to meaningless
disk usage calculation.  This patch reports error when mounting such
bogus volumes.

Signed-off-by: Haogang Chen <haogangchen@gmail.com>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris...
Linus Torvalds [Sat, 17 Mar 2012 00:04:02 +0000 (17:04 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/jmorris/linux-security

Pull maintainer update from James Morris:
 "Please pull this patch which adds Serge as maintainer of the
  capabilities code, as discussed on lwn and the lsm list.

  New capabilities must be signed off by the maintainer, and new uses of
  any capabilities should at be cc'd to the maintainer."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
  MAINTAINERS: Add Serge as maintainer of capabilities

12 years agoMerge tag 'for-linus' of git://linux-c6x.org/git/projects/linux-c6x-upstreaming
Linus Torvalds [Sat, 17 Mar 2012 00:03:15 +0000 (17:03 -0700)]
Merge tag 'for-linus' of git://linux-c6x.org/git/projects/linux-c6x-upstreaming

Pull c6x bugfix from Mark Salter:
 "Remove dead code from entry.S which causes a build failure when using
  a newer assembler (v2.22 complains about it, v2.20 ignores it)."

* tag 'for-linus' of git://linux-c6x.org/git/projects/linux-c6x-upstreaming:
  C6X: remove dead code from entry.S

12 years agoafs: Remote abort can cause BUG in rxrpc code
Anton Blanchard [Fri, 16 Mar 2012 10:28:19 +0000 (10:28 +0000)]
afs: Remote abort can cause BUG in rxrpc code

When writing files to afs I sometimes hit a BUG:

kernel BUG at fs/afs/rxrpc.c:179!

With a backtrace of:

afs_free_call
afs_make_call
afs_fs_store_data
afs_vnode_store_data
afs_write_back_from_locked_page
afs_writepages_region
afs_writepages

The cause is:

ASSERT(skb_queue_empty(&call->rx_queue));

Looking at a tcpdump of the session the abort happens because we
are exceeding our disk quota:

rx abort fs reply store-data error diskquota exceeded (32)

So the abort error is valid. We hit the BUG because we haven't
freed all the resources for the call.

By freeing any skbs in call->rx_queue before calling afs_free_call
we avoid hitting leaking memory and avoid hitting the BUG.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Cc: <stable@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoafs: Read of file returns EBADMSG
Anton Blanchard [Fri, 16 Mar 2012 10:28:07 +0000 (10:28 +0000)]
afs: Read of file returns EBADMSG

A read of a large file on an afs mount failed:

# cat junk.file > /dev/null
cat: junk.file: Bad message

Looking at the trace, call->offset wrapped since it is only an
unsigned short. In afs_extract_data:

        _enter("{%u},{%zu},%d,,%zu", call->offset, len, last, count);
...

        if (call->offset < count) {
                if (last) {
                        _leave(" = -EBADMSG [%d < %zu]", call->offset, count);
                        return -EBADMSG;
                }

Which matches the trace:

[cat   ] ==> afs_extract_data({65132},{524},1,,65536)
[cat   ] <== afs_extract_data() = -EBADMSG [0 < 65536]

call->offset went from 65132 to 0. Fix this by making call->offset an
unsigned int.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Cc: <stable@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wirel...
John W. Linville [Fri, 16 Mar 2012 17:45:25 +0000 (13:45 -0400)]
Merge branch 'master' of git://git./linux/kernel/git/linville/wireless-next into for-davem

Conflicts:
drivers/net/wireless/ath/ath9k/hw.c

12 years agoC6X: remove dead code from entry.S
Mark Salter [Fri, 16 Mar 2012 13:27:57 +0000 (09:27 -0400)]
C6X: remove dead code from entry.S

The ENDPROC() on sys_fadvise64_c6x() in arch/c6x/kernel/entry.S is
outside of the conditional block with the matching ENTRY() macro. This
leads a newer (v2.22 vs. v2.20) assembler to complain:

  /tmp/ccGZBaPT.s: Assembler messages:
  /tmp/ccGZBaPT.s: Error: .size expression for sys_fadvise64_c6x does not evaluate to a constant

The conditional block became dead code when c6x switched to generic
unistd.h and should be removed along with the offending ENDPROC().

Signed-off-by: Mark Salter <msalter@redhat.com>
Acked-by: David Howells <dhowells@redhat.com>
12 years agocdc_ncm: avoid discarding datagrams in rx path
Alexey Orishko [Wed, 14 Mar 2012 11:26:12 +0000 (11:26 +0000)]
cdc_ncm: avoid discarding datagrams in rx path

Changes:
- removed a limit for amount of datagrams for IN NTB
- using pointer to traverse NTB in rx_fixup()
- renamed "temp" to "len" in rx_fixup()
- do NTB sequence number check in rx path
Tested on Intel/ARM.

Reviewed-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Tested-by: Dmitry Tarnyagin <Dmitry.Tarnyagin@stericsson.com>
Signed-off-by: Alexey Orishko <alexey.orishko@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocdc_ncm: fix MTU and max_datagram_size handling
Alexey Orishko [Wed, 14 Mar 2012 11:26:11 +0000 (11:26 +0000)]
cdc_ncm: fix MTU and max_datagram_size handling

Changes/fixes:
- inform device if max_datagram_size was changed by host
- max_datagram_size can't be bigger MTU in ETH func descr
- fix constants definitions to enable running CAIF service over NCM

Tested on Intel/ARM.

Reviewed-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Tested-by: Dmitry Tarnyagin <Dmitry.Tarnyagin@stericsson.com>
Signed-off-by: Alexey Orishko <alexey.orishko@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocdc_ncm: reduce driver latency in the data path
Alexey Orishko [Wed, 14 Mar 2012 11:26:10 +0000 (11:26 +0000)]
cdc_ncm: reduce driver latency in the data path

Changes:
- use high resolution timer
  latency became approx. 10-15 times less than before
- use taklet for sending remaining data in tx path
Tested on Intel/ARM.

Reviewed-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Tested-by: Dmitry Tarnyagin <Dmitry.Tarnyagin@stericsson.com>
Signed-off-by: Alexey Orishko <alexey.orishko@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agowimax/i2400m: fix erroneous NETDEV_TX_BUSY use
Eric Dumazet [Wed, 14 Mar 2012 09:21:44 +0000 (09:21 +0000)]
wimax/i2400m: fix erroneous NETDEV_TX_BUSY use

A driver start_xmit() method cannot free skb and return NETDEV_TX_BUSY,
since caller is going to reuse freed skb.

In fact netif_tx_stop_queue() / netif_stop_queue() is needed before
returning NETDEV_TX_BUSY or you can trigger a ksoftirqd fatal loop.

In case of memory allocation error, only safe way is to drop the packet
and return NETDEV_TX_OK

Also increments tx_dropped counter

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet/hyperv: fix erroneous NETDEV_TX_BUSY use
Eric Dumazet [Wed, 14 Mar 2012 08:53:34 +0000 (08:53 +0000)]
net/hyperv: fix erroneous NETDEV_TX_BUSY use

A driver start_xmit() method cannot free skb and return NETDEV_TX_BUSY,
since caller is going to reuse freed skb.

This is mostly a revert of commit bf769375c (staging: hv: fix the return
status of netvsc_start_xmit())

In fact netif_tx_stop_queue() / netif_stop_queue() is needed before
returning NETDEV_TX_BUSY or you can trigger a ksoftirqd fatal loop.

In case of memory allocation error, only safe way is to drop the packet
and return NETDEV_TX_OK

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: "K. Y. Srinivasan" <kys@microsoft.com>
Cc: Haiyang Zhang <haiyangz@microsoft.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet/usbnet: reserve headroom on rx skbs
Eric Dumazet [Wed, 14 Mar 2012 06:56:25 +0000 (06:56 +0000)]
net/usbnet: reserve headroom on rx skbs

network drivers should reserve some headroom on incoming skbs so that we
dont need expensive reallocations, eg forwarding packets in tunnels.

This NET_SKB_PAD padding is done in various helpers, like
__netdev_alloc_skb_ip_align() in this patch, combining NET_SKB_PAD and
NET_IP_ALIGN magic.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Oliver Neukum <oneukum@suse.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: fix memory leak in bnx2x_init_firmware()
Michal Schmidt [Thu, 15 Mar 2012 14:08:29 +0000 (14:08 +0000)]
bnx2x: fix memory leak in bnx2x_init_firmware()

When cycling the interface down and up, bnx2x_init_firmware() knows that
the firmware is already loaded, but nevertheless it allocates certain
arrays anew (init_data, init_ops, init_ops_offsets, iro_arr). The old
arrays are leaked.

Fix the leaks by returning early if the firmware was already loaded.
Because if the firmware is loaded, so are the arrays.

Signed-off-by: Michal Schmidt <mschmidt@redhat.com>
Acked-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: fix a crash on corrupt firmware file
Michal Schmidt [Thu, 15 Mar 2012 14:08:28 +0000 (14:08 +0000)]
bnx2x: fix a crash on corrupt firmware file

If the requested firmware is deemed corrupt and then released, reset the
pointer to NULL in order to avoid double-freeing it in
bnx2x_release_firmware() or dereferencing it in bnx2x_init_firmware().

Signed-off-by: Michal Schmidt <mschmidt@redhat.com>
Acked-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agosch_sfq: revert dont put new flow at the end of flows
Eric Dumazet [Tue, 13 Mar 2012 18:04:25 +0000 (18:04 +0000)]
sch_sfq: revert dont put new flow at the end of flows

This reverts commit d47a0ac7b6 (sch_sfq: dont put new flow at the end of
flows)

As Jesper found out, patch sounded great but has bad side effects.

In stress situation, pushing new flows in front of the queue can prevent
old flows doing any progress. Packets can stay in SFQ queue for
unlimited amount of time.

It's possible to add heuristics to limit this problem, but this would
add complexity outside of SFQ scope.

A more sensible answer to Dave Taht concerns (who reported the issued I
tried to solve in original commit) is probably to use a qdisc hierarchy
so that high prio packets dont enter a potentially crowded SFQ qdisc.

Reported-by: Jesper Dangaard Brouer <jdb@comx.dk>
Cc: Dave Taht <dave.taht@gmail.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoipv6: fix icmp6_dst_alloc()
Eric Dumazet [Wed, 14 Mar 2012 21:13:11 +0000 (21:13 +0000)]
ipv6: fix icmp6_dst_alloc()

commit 87a115783 ( ipv6: Move xfrm_lookup() call down into
icmp6_dst_alloc().) forgot to convert one error path, leading
to crashes in mld_sendpack()

Many thanks to Dave Jones for providing a very complete bug report.

Reported-by: Dave Jones <davej@redhat.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoasix: asix_rx_fixup surgery to reduce skb truesizes
Eric Dumazet [Wed, 14 Mar 2012 20:18:32 +0000 (20:18 +0000)]
asix: asix_rx_fixup surgery to reduce skb truesizes

asix_rx_fixup() is complex, and does some unnecessary memory copies (at
least on x86 where NET_IP_ALIGN is 0)

Also, it tends to provide skbs with a big truesize (4096+256 with
MTU=1500) to upper stack, so incoming trafic consume a lot of memory and
I noticed early packet drops because we hit socket rcvbuf too fast.

Switch to a different strategy, using copybreak so that we provide nice
skbs to upper stack (including the NET_SKB_PAD to avoid future head
reallocations in some paths)

With this patch, I no longer see packets drops or tcp collapses on
various tcp workload with a AX88772 adapter.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Aurelien Jacobs <aurel@gnuage.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Trond Wuellner <trond@chromium.org>
Cc: Grant Grundler <grundler@chromium.org>
Cc: Paul Stewart <pstew@chromium.org>
Reviewed-by: Grant Grundler <grundler@chromium.org>
Reviewed-by: Grant Grundler <grundler@chromium.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMAINTAINERS: Add Serge as maintainer of capabilities
James Morris [Fri, 16 Mar 2012 01:05:48 +0000 (12:05 +1100)]
MAINTAINERS: Add Serge as maintainer of capabilities

Add Serge as maintainer of capabilities, per suggestion on LWN:
http://lwn.net/Articles/486306/

Signed-off-by: James Morris <james.l.morris@oracle.com>
12 years agoMerge branch 'akpm' (Andrew's patch-bomb)
Linus Torvalds [Fri, 16 Mar 2012 00:16:22 +0000 (17:16 -0700)]
Merge branch 'akpm' (Andrew's patch-bomb)

Merge patches from Andrew Morton:
 "Nine patches - some bug fixes and some MAINTAINERS fiddling."

* emailed from Andrew Morton <akpm@linux-foundation.org>:
  drivers/video/backlight/s6e63m0.c: fix corruption storing gamma mode
  MAINTAINERS: add entry for exynos mipi display drivers
  MAINTAINERS: fix link to Gustavo Padovans tree
  MAINTAINERS: add Johan to Bluetooth maintainers
  MAINTAINERS: Gustavo has moved
  prctl: use CAP_SYS_RESOURCE for PR_SET_MM option
  rapidio/tsi721: fix bug in register offset definitions
  MAINTAINERS: update ST's Mailing list for SPEAr
  memcg: free mem_cgroup by RCU to fix oops

12 years agoMerge branch 'i2c-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvar...
Linus Torvalds [Fri, 16 Mar 2012 00:14:35 +0000 (17:14 -0700)]
Merge branch 'i2c-for-linus' of git://git./linux/kernel/git/jdelvare/staging

Pull i2c subsystem fixes from Jean Delvare.

* 'i2c-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
  i2c-algo-bit: Fix spurious SCL timeouts under heavy load
  i2c-core: Comment says "transmitted" but means "received"

12 years agoMerge tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck...
Linus Torvalds [Fri, 16 Mar 2012 00:13:39 +0000 (17:13 -0700)]
Merge tag 'hwmon-for-linus' of git://git./linux/kernel/git/groeck/linux-staging

Pull hwmon fixes from Guenter Roeck.

* tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
  hwmon: (zl6100) Enable interval between chip accesses for all chips
  hwmon: (w83627ehf) Describe undocumented pwm attributes
  hwmon: (w83627ehf) Fix temp2 source for W83627UHG
  hwmon: (w83627ehf) Fix memory leak in probe function
  hwmon: (w83627ehf) Fix writing into fan_stop_time for NCT6775F/NCT6776F

12 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Fri, 16 Mar 2012 00:07:25 +0000 (17:07 -0700)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull drm exynos/intel updates from Dave Airlie:
 "Two minor updates from Jesse for Intel SNB fixes, and a few fixes from
  Samsung for exynos.  The pull req has Alan's commit in it since Intel
  based their tree on my tree at that time, but it all seems fine wrt
  merging."

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm exynos: use drm_fb_helper_set_par directly
  drm/exynos: Fix fb_videomode <-> drm_mode_modeinfo conversion
  drm/exynos: fix runtime_pm fimd device state on probe
  drm/exynos: use correct 'exynos-drm' name for platform device
  drm/i915: support 32 bit BGR formats in sprite planes
  drm/i915: fix color order for BGR formats on SNB
  drm/gma500: Fix Cedarview boot failures in 3.3-rc

12 years agoMerge branch 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab...
Linus Torvalds [Fri, 16 Mar 2012 00:06:05 +0000 (17:06 -0700)]
Merge branch 'v4l_for_linus' of git://git./linux/kernel/git/mchehab/linux-media

Pull media fixes from Mauro Carvalho Chehab:
 "For 4 fixes for 3.3 (all trivial):
       - uvc video driver: fixes a division by zero;
       - davinci: add module.h to fix compilation;
       - smsusb: fix the delivery system setting;
       - smsdvb: the get_frontend implementation there is broken.

  The smsdvb patch has 127 lines, but it is trivial: instead of
  returning a cache of the set_frontend (with is wrong, as it doesn't
  have the updated values for the data, and the implementation there is
  buggy), it copies the information of the detected DVB parameters from
  the smsdvb private structures into the corresponding DVBv5 struct
  fields."

* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
  [media] smsdvb: fix get_frontend
  [media] smsusb: fix the default delivery system setting
  [media] media: davinci: added module.h to resolve unresolved macros
  [media] [FOR,v3.3] uvcvideo: Avoid division by 0 in timestamp calculation