kernel/kernel-generic.git
12 years agoinet_diag: Switch the _dump to work with new header
Pavel Emelyanov [Tue, 6 Dec 2011 07:58:58 +0000 (07:58 +0000)]
inet_diag: Switch the _dump to work with new header

Make inet_diag_dumo work with given header instead of calculating
one from the nl message.

The SOCK_DIAG_BY_FAMILY just passes skb's one through, the compat code
converts the old header to new one.

Also fix the bytecode calculation to find one at proper offset.

Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoinet_diag: Switch the _get_exact to work with new header
Pavel Emelyanov [Tue, 6 Dec 2011 07:58:39 +0000 (07:58 +0000)]
inet_diag: Switch the _get_exact to work with new header

Make inet_diag_get_exact work with given header instead of calculating
one from the nl message.

The SOCK_DIAG_BY_FAMILY just passes skb's one through, the compat code
converts the old header to new one.

Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoinet_diag: Introduce new inet_diag_req header
Pavel Emelyanov [Tue, 6 Dec 2011 07:58:21 +0000 (07:58 +0000)]
inet_diag: Introduce new inet_diag_req header

This one coinsides with the sock_diag_req in the beginning and
contains only used fields from its previous analogue.

The existing code is patched to use the _compat version of it
for now.

Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agosock_diag: Initial skeleton
Pavel Emelyanov [Tue, 6 Dec 2011 07:58:03 +0000 (07:58 +0000)]
sock_diag: Initial skeleton

When receiving the SOCK_DIAG_BY_FAMILY message we have to find the
handler for provided family and pass the nl message to it.

This patch describes an infrastructure to work with such nandlers
and implements stubs for AF_INET(6) ones.

Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoinet_diag: Switch from _GETSOCK to IPPROTO_ numbers
Pavel Emelyanov [Tue, 6 Dec 2011 08:05:24 +0000 (08:05 +0000)]
inet_diag: Switch from _GETSOCK to IPPROTO_ numbers

Sorry, but the vger didn't let this message go to the list. Re-sending it with
less spam-filter-prone subject.

When dumping the AF_INET/AF_INET6 sockets user will also specify the protocol,
so prepare the protocol diag handlers to work with IPPROTO_ constants.

Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoinet_diag: Move byte-code finding up the call-stack
Pavel Emelyanov [Tue, 6 Dec 2011 07:57:26 +0000 (07:57 +0000)]
inet_diag: Move byte-code finding up the call-stack

Current code calculates it at fixed offset. This offset will change, so
move the BC calculation upper to make the further patching simpler.

Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agosock_diag: Introduce new message type
Pavel Emelyanov [Tue, 6 Dec 2011 07:57:06 +0000 (07:57 +0000)]
sock_diag: Introduce new message type

This type will run the family+protocol based socket dumping.
Also prepare the stub function for it.

Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoinet_diag: Partly rename inet_ to sock_
Pavel Emelyanov [Tue, 6 Dec 2011 07:56:43 +0000 (07:56 +0000)]
inet_diag: Partly rename inet_ to sock_

The ultimate goal is to get the sock_diag module, that works in
family+protocol terms. Currently this is suitable to do on the
inet_diag basis, so rename parts of the code. It will be moved
to sock_diag.c later.

Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocaif-spi: Bugfix for dump upon device removal
Erwan Bracq [Tue, 6 Dec 2011 07:25:05 +0000 (07:25 +0000)]
caif-spi: Bugfix for dump upon device removal

Fix dump upon device removal, by moving deinitialization from
platform-device-remove to network-interface-uninit.

Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge branch 'for-davem' of git://git.kernel.org/pub/scm/linux/kernel/git/linville...
David S. Miller [Tue, 6 Dec 2011 18:31:19 +0000 (13:31 -0500)]
Merge branch 'for-davem' of git://git./linux/kernel/git/linville/wireless-next

12 years agobnx2x: fix crash while ethtool -t
Dmitry Kravkov [Tue, 6 Dec 2011 02:05:12 +0000 (02:05 +0000)]
bnx2x: fix crash while ethtool -t

commit 2df1a70aaf70e8dff11b89b938a5f317556ee640  "bnx2x: Support
for byte queue limits" has introduced an asymmetry in usage of
netdev_tx_completed_queue and netdev_tx_sent_queue. Missing
call to netdev_tx_sent_queue causes the crash during ethtool -t.

The patch adds the missing call.

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetback: fix typo in comment
Wei Liu [Tue, 6 Dec 2011 02:04:50 +0000 (02:04 +0000)]
netback: fix typo in comment

"variables a used" should be "variables are used".

Signed-off-by: Wei Liu <wei.liu2@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetback: remove redundant assignment
Wei Liu [Tue, 6 Dec 2011 00:08:20 +0000 (00:08 +0000)]
netback: remove redundant assignment

New value for netbk->mmap_pages[pending_idx] is assigned in
xen_netbk_alloc_page(), no need for a second assignment which
exposes internal to the outside world.

Signed-off-by: Wei Liu <wei.liu2@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x, cnic: support DRV_INFO upon FW request
Barak Witkowski [Mon, 5 Dec 2011 22:41:50 +0000 (22:41 +0000)]
bnx2x, cnic: support DRV_INFO upon FW request

Add support to send driver capabilities, settings and statistics to
management firmware.

[ Redone using many local variables, removed many unnecessary inlines,
  and put #defines at the left margin suggested by Joe Perches ]

Signed-off-by: Barak Witkowski <barak@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: Eddie Wai <eddie.wai@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: support classification config query
Ariel Elior [Mon, 5 Dec 2011 21:52:24 +0000 (21:52 +0000)]
bnx2x: support classification config query

To support copying MAC addresses to firmware query structure.

[ Fixed up style and formatting errors noted by DaveM and Joe Perches ]

Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Barak Witkowski <barak@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: add fcoe statistics
Barak Witkowski [Mon, 5 Dec 2011 21:52:23 +0000 (21:52 +0000)]
bnx2x: add fcoe statistics

Add FCoE statistics support for FCoE capable devices.

Signed-off-by: Barak Witkowski <barak@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: add PFC statistics
Barak Witkowski [Mon, 5 Dec 2011 21:52:22 +0000 (21:52 +0000)]
bnx2x: add PFC statistics

Add Priority flow control counters for ethtool -S.

Signed-off-by: Barak Witkowski <barak@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoopenvswitch: small potential memory leak in ovs_vport_alloc()
Dan Carpenter [Mon, 5 Dec 2011 20:27:07 +0000 (20:27 +0000)]
openvswitch: small potential memory leak in ovs_vport_alloc()

We're unlikely to hit this leak, but the static checkers complain if we
don't take care of it.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Jesse Gross <jesse@nicira.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/linville/wirel...
John W. Linville [Tue, 6 Dec 2011 15:47:12 +0000 (10:47 -0500)]
Merge branch 'master' of git://git./linux/kernel/git/linville/wireless-next into for-davem

12 years agoipv4: arp: Cleanup in arp.c
Igor Maravic [Mon, 5 Dec 2011 02:25:57 +0000 (02:25 +0000)]
ipv4: arp: Cleanup in arp.c

Use "IS_ENABLED(CONFIG_FOO)" macro instead of
"defined(CONFIG_FOO) || defined(CONFIG_FOO_MODULE)"

Signed-off-by: Igor Maravic <igorm@etf.rs>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agotcp: remove TCP_OFF and TCP_PAGE macros
Eric Dumazet [Mon, 5 Dec 2011 01:07:15 +0000 (01:07 +0000)]
tcp: remove TCP_OFF and TCP_PAGE macros

As mentioned by Joe Perches, TCP_OFF() and TCP_PAGE() macros are
useless.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobql: fix CONFIG_XPS=n build
Eric Dumazet [Sun, 4 Dec 2011 12:38:00 +0000 (12:38 +0000)]
bql: fix CONFIG_XPS=n build

netdev_queue_release() should be called even if CONFIG_XPS=n
to properly release device reference.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agotcp: fix tcp_trim_head()
Eric Dumazet [Sun, 4 Dec 2011 08:51:08 +0000 (08:51 +0000)]
tcp: fix tcp_trim_head()

commit f07d960df3 (tcp: avoid frag allocation for small frames)
breaked assumption in tcp stack that skb is either linear (skb->data_len
== 0), or fully fragged (skb->data_len == skb->len)

tcp_trim_head() made this assumption, we must fix it.

Thanks to Vijay for providing a very detailed explanation.

Reported-by: Vijay Subramanian <subramanian.vijay@gmail.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocaif: Stash away hijacked skb destructor and call it later
sjur.brandeland@stericsson.com [Sun, 4 Dec 2011 11:22:55 +0000 (11:22 +0000)]
caif: Stash away hijacked skb destructor and call it later

This patch adds functionality for avoiding orphaning SKB too early.
The original skb is stashed away and the original destructor is called
from the hi-jacked flow-on callback. If CAIF interface goes down and a
hi-jacked SKB exists, the original skb->destructor is restored.

Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocaif: Add support for flow-control on device's tx-queue
sjur.brandeland@stericsson.com [Sun, 4 Dec 2011 11:22:54 +0000 (11:22 +0000)]
caif: Add support for flow-control on device's tx-queue

Flow control is implemented by inspecting the qdisc queue length
in order to detect potential overflow on the TX queue. When a threshold
is reached flow-off is sent upwards in the CAIF stack. At the same time
the skb->destructor is hi-jacked by orphaning the SKB and the original
destructor is replaced with a "flow-on" callback. When the "hi-jacked"
SKB is consumed the queue should be empty, and the "flow-on" callback
is called and xon is sent upwards in the CAIF stack.

Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocaif: Add support for CAIF over CDC NCM USB interface
sjur.brandeland@stericsson.com [Sun, 4 Dec 2011 11:22:53 +0000 (11:22 +0000)]
caif: Add support for CAIF over CDC NCM USB interface

NCM 1.0 does not support anything but Ethernet framing, hence
CAIF payload will be put into Ethernet frames.

Discovery is based on fixed USB vendor 0x04cc (ST-Ericsson),
product-id 0x230f (NCM). In this variant only CAIF payload is sent over
the NCM interface.

The CAIF stack (cfusbl.c) will when USB interface register first check if
we got a CDC NCM USB interface with the right VID, PID.
It will then read the device's Ethernet address and create a 'template'
Ethernet TX header, using a broadcast address as the destination address,
and EthType 0x88b5 (802.1 Local Experimental - vendor specific).

A protocol handler for 0x88b5 is setup for reception of CAIF frames from
the CDC NCM USB interface.

Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoif_ether.h: Add IEEE 802.1 Local Experimental Ethertype 1.
sjur.brandeland@stericsson.com [Sun, 4 Dec 2011 11:22:52 +0000 (11:22 +0000)]
if_ether.h: Add IEEE 802.1 Local Experimental Ethertype 1.

Add EthType 0x88b5.
This Ethertype value is available for public use for prototype and
vendor-specific protocol development,as defined in Amendment 802a
to IEEE Std 802.

Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.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, 5 Dec 2011 23:27:28 +0000 (18:27 -0500)]
Merge branch 'master' of git://git./linux/kernel/git/jkirsher/net-next

12 years agoinfiniband: ipoib: Sanitize neighbour handling in ipoib_main.c
David Miller [Fri, 2 Dec 2011 16:52:44 +0000 (16:52 +0000)]
infiniband: ipoib: Sanitize neighbour handling in ipoib_main.c

Reduce the number of dst_get_neighbour_noref() calls within a single
call chain.  Primarily by passing the neighbour pointer down to the
helper functions.

Handle dst_get_neighbour_noref() returning NULL in ipoib_start_xmit()
by incrementing the dropped counter and freeing the packet.  We don't
want it to fall through into the ARP/RARP/multicast handling, since
that should only happen when skb_dst() is NULL.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Roland Dreier <roland@purestorage.com>
12 years agocxgb4i: Handle dst_get_neighbour_noref() returning NULL.
David Miller [Fri, 2 Dec 2011 16:52:39 +0000 (16:52 +0000)]
cxgb4i: Handle dst_get_neighbour_noref() returning NULL.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Roland Dreier <roland@purestorage.com>
12 years agolibcxgbi: Handle dst_get_neighbour_noref() returning NULL.
David Miller [Fri, 2 Dec 2011 16:52:35 +0000 (16:52 +0000)]
libcxgbi: Handle dst_get_neighbour_noref() returning NULL.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Roland Dreier <roland@purestorage.com>
12 years agoinfiniband: cxgb4: Consolidate 3 copies of the same operation into 1 helper function.
David Miller [Fri, 2 Dec 2011 16:52:31 +0000 (16:52 +0000)]
infiniband: cxgb4: Consolidate 3 copies of the same operation into 1 helper function.

Three pieces of code do the same thing, create a l2t entry and then
import this information into the c4iw_ep object.

Create a helper function and call it from these 3 locations instead.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Roland Dreier <roland@purestorage.com>
12 years agoinfiniband: nes: Use dst's neighbour entry.
David Miller [Fri, 2 Dec 2011 16:52:27 +0000 (16:52 +0000)]
infiniband: nes: Use dst's neighbour entry.

Do this instead of performing a by-hand lookup.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Roland Dreier <roland@purestorage.com>
12 years agocxgb3: Handle NULL dst neighbour in cxgb3_offload.c
David Miller [Fri, 2 Dec 2011 16:52:22 +0000 (16:52 +0000)]
cxgb3: Handle NULL dst neighbour in cxgb3_offload.c

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocxgb3: Rework t3_l2t_get to take a dst_entry instead of a neighbour.
David Miller [Fri, 2 Dec 2011 16:52:18 +0000 (16:52 +0000)]
cxgb3: Rework t3_l2t_get to take a dst_entry instead of a neighbour.

This way we consolidate the RCU locking down into the place where it
actually matters, and also we can make the code handle
dst_get_neighbour_noref() returning NULL properly.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoinfiniband: addr: Consolidate code to fetch neighbour hardware address from dst.
David Miller [Fri, 2 Dec 2011 16:52:14 +0000 (16:52 +0000)]
infiniband: addr: Consolidate code to fetch neighbour hardware address from dst.

IPV4 should do exactly what the IPV6 code does here, which is
use the neighbour obtained via the dst entry.

And now that the two code paths do the same thing, use a common
helper function to perform the operation.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Roland Dreier <roland@purestorage.com>
12 years agonet: Rename dst_get_neighbour{, _raw} to dst_get_neighbour_noref{, _raw}.
David Miller [Fri, 2 Dec 2011 16:52:08 +0000 (16:52 +0000)]
net: Rename dst_get_neighbour{, _raw} to dst_get_neighbour_noref{, _raw}.

To reflect the fact that a refrence is not obtained to the
resulting neighbour entry.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Roland Dreier <roland@purestorage.com>
12 years agoixgbe: Remove function prototype for non-existent function
Greg Rose [Thu, 17 Nov 2011 01:48:24 +0000 (01:48 +0000)]
ixgbe: Remove function prototype for non-existent function

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: DCB: IEEE transitions may fail to reprogram hardware.
John Fastabend [Sat, 12 Nov 2011 01:12:22 +0000 (01:12 +0000)]
ixgbe: DCB: IEEE transitions may fail to reprogram hardware.

Transitioning through an IEEE DCBX version from a CEE DCBX
and back (CEE->IEEE->CEE) may leave IEEE attributes programmed
in the hardware. DCB uses a bit field in the set routines to
determine which attributes PG, PFC, APP need to be reprogrammed.
This is needed because user flow allows queueing a series
of changes and then reprogramming the hardware with the
entire set in one operation.

When transitioning from IEEE DCBX mode back into CEE DCBX
mode the PG and PFC bits need to be set so the possibly

different CEE attributes get programmed into the device.

This patch fixes broken logic that was evaluating to 0
and never setting any bits. Further this removes some
checks for num_tc in set routines. This logic only worked
when the number of traffic classes and user priorities
were equal. This is no longer the case for X540 devices.
Besides we can trust user input in this case if the
device is incorrectly configured the DCB bandwidths will
be incorrectly mapped but no OOPs, BUG, or hardware
failure will occur.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoixgbe: DCBnl set_all, order of operations fix
John Fastabend [Sat, 12 Nov 2011 01:12:16 +0000 (01:12 +0000)]
ixgbe: DCBnl set_all, order of operations fix

The order of operations is important in DCBnl set_all(). When FCoE
is configured it uses the up2tc map to learn which queues to configure
the hardware offloads on. Therefore we need to setup the map before
configuring FCoE.

This is only seen when the both up2tc mappings and APP info are
configured simultaneously.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoigb: Update DMA Coalescing threshold calculation.
Matthew Vick [Thu, 17 Nov 2011 08:33:12 +0000 (08:33 +0000)]
igb: Update DMA Coalescing threshold calculation.

This patch updates the DMA Coalescing feature parameters to account for
larger MTUs. Previously, sufficient space may not have been allocated in
the receive buffer, causing packet drop.

Signed-off-by: Matthew Vick <matthew.vick@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoe1000e: hitting BUG_ON() from napi_enable
Bruce Allan [Tue, 29 Nov 2011 08:09:19 +0000 (08:09 +0000)]
e1000e: hitting BUG_ON() from napi_enable

Based on a patch from Mike McElroy created against the out-of-tree e1000e
driver:

Hitting the BUG_ON in napi_enable(). Code inspection shows that this can
only be triggered by calling napi_enable() twice without an intervening
napi_disable().

I saw the following sequence of events in the stack trace:

1) We simulated a cable pull using an Extreme switch.
2) e1000_tx_timeout() was entered.
3) e1000_reset_task() was called. Saw the message from e_err() in the
console log.
4) e1000_reinit_locked was called. This function calls e1000_down() and
e1000_up(). These functions call napi_disable() and napi_enable()
respectively.
5) Then on another thread, a monitor task saw carrier was down and executed
'ip set link down' and 'ip set link up' commands.
6) Saw the '_E1000_RESETTING'warning fron the e1000_close function.
7) Either the e1000_open() executed between the e1000_down() and e1000_up()
calls in step 4 or the e1000_open() call executed after the e1000_up()
call.  In either case, napi_enable() is called twice which triggers the
BUG_ON.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Cc: Mike McElroy <mike.mcelroy@stratus.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agoe1000e: Avoid wrong check on TX hang
Jeff Kirsher [Fri, 18 Nov 2011 14:25:00 +0000 (14:25 +0000)]
e1000e: Avoid wrong check on TX hang

Based on the original patch submitted my Michael Wang
<wangyun@linux.vnet.ibm.com>.
Descriptors may not be write-back while checking TX hang with flag
FLAG2_DMA_BURST on.
So when we detect hang, we just flush the descriptor and detect
again for once.

-v2 change 1 to true and 0 to false and remove extra ()

CC: Michael Wang <wangyun@linux.vnet.ibm.com>
CC: Flavio Leitner <fbl@redhat.com>
Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
12 years agotcp: tcp_sendmsg() page recycling
Eric Dumazet [Sun, 4 Dec 2011 07:05:17 +0000 (07:05 +0000)]
tcp: tcp_sendmsg() page recycling

If our TCP_PAGE(sk) is not shared (page_count() == 1), we can set page
offset to 0.

This permits better filling of the pages on small to medium tcp writes.

"tbench 16" results on my dev server (2x4x2 machine) :

Before : 3072 MB/s
After  : 3146 MB/s  (2.4 % gain)

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agotcp: take care of misalignments
Eric Dumazet [Sat, 3 Dec 2011 21:39:53 +0000 (21:39 +0000)]
tcp: take care of misalignments

We discovered that TCP stack could retransmit misaligned skbs if a
malicious peer acknowledged sub MSS frame. This currently can happen
only if output interface is non SG enabled : If SG is enabled, tcp
builds headless skbs (all payload is included in fragments), so the tcp
trimming process only removes parts of skb fragments, header stay
aligned.

Some arches cant handle misalignments, so force a head reallocation and
shrink headroom to MAX_TCP_HEADER.

Dont care about misaligments on x86 and PPC (or other arches setting
NET_IP_ALIGN to 0)

This patch introduces __pskb_copy() which can specify the headroom of
new head, and pskb_copy() becomes a wrapper on top of __pskb_copy()

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agosfc: Use kcalloc instead of kzalloc to allocate array
Thomas Meyer [Fri, 2 Dec 2011 12:36:13 +0000 (12:36 +0000)]
sfc: Use kcalloc instead of kzalloc to allocate array

The advantage of kcalloc is, that will prevent integer overflows which could
result from the multiplication of number of elements and size and it is also
a bit nicer to read.

The semantic patch that makes this change is available
in https://lkml.org/lkml/2011/11/25/107

Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoll_temac: Use kcalloc instead of kzalloc to allocate array
Thomas Meyer [Fri, 2 Dec 2011 12:35:43 +0000 (12:35 +0000)]
ll_temac: Use kcalloc instead of kzalloc to allocate array

The advantage of kcalloc is, that will prevent integer overflows which could
result from the multiplication of number of elements and size and it is also
a bit nicer to read.

The semantic patch that makes this change is available
in https://lkml.org/lkml/2011/11/25/107

Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoenic: Use kcalloc instead of kzalloc to allocate array
Thomas Meyer [Tue, 29 Nov 2011 11:08:00 +0000 (11:08 +0000)]
enic: Use kcalloc instead of kzalloc to allocate array

The advantage of kcalloc is, that will prevent integer overflows which could
result from the multiplication of number of elements and size and it is also
a bit nicer to read.

The semantic patch that makes this change is available
in https://lkml.org/lkml/2011/11/25/107

Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobnx2x: Use kcalloc instead of kzalloc to allocate array
Thomas Meyer [Tue, 29 Nov 2011 11:08:00 +0000 (11:08 +0000)]
bnx2x: Use kcalloc instead of kzalloc to allocate array

The advantage of kcalloc is, that will prevent integer overflows which could
result from the multiplication of number of elements and size and it is also
a bit nicer to read.

The semantic patch that makes this change is available
in https://lkml.org/lkml/2011/11/25/107

Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agotcp: drop SYN+FIN messages
Eric Dumazet [Fri, 2 Dec 2011 23:41:42 +0000 (23:41 +0000)]
tcp: drop SYN+FIN messages

Denys Fedoryshchenko reported that SYN+FIN attacks were bringing his
linux machines to their limits.

Dont call conn_request() if the TCP flags includes SYN flag

Reported-by: Denys Fedoryshchenko <denys@visp.net.lb>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge branch 'for-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/jesse...
David S. Miller [Sun, 4 Dec 2011 03:53:31 +0000 (22:53 -0500)]
Merge branch 'for-upstream' of git://git./linux/kernel/git/jesse/openvswitch

12 years agoipv6: Kill ndisc_get_neigh() inline helper.
David S. Miller [Sat, 3 Dec 2011 23:29:30 +0000 (18:29 -0500)]
ipv6: Kill ndisc_get_neigh() inline helper.

It's only used in net/ipv6/route.c and the NULL device check is
superfluous for all of the existing call sites.

Just expand the __ndisc_lookup_errno() call at each location.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoipv6: Various cleanups in route.c
David S. Miller [Sat, 3 Dec 2011 23:02:47 +0000 (18:02 -0500)]
ipv6: Various cleanups in route.c

1) x == NULL --> !x
2) x != NULL --> x
3) (x&BIT) --> (x & BIT)
4) (BIT1|BIT2) --> (BIT1 | BIT2)
5) proper argument and struct member alignment

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoipv6: Various cleanups in ip6_route.c
David S. Miller [Sat, 3 Dec 2011 22:50:45 +0000 (17:50 -0500)]
ipv6: Various cleanups in ip6_route.c

1) x == NULL --> !x
2) x != NULL --> x
3) if() --> if ()
4) while() --> while ()
5) (x & BIT) == 0 --> !(x & BIT)
6) (x&BIT) --> (x & BIT)
7) x=y --> x = y
8) (BIT1|BIT2) --> (BIT1 | BIT2)
9) if ((x & BIT)) --> if (x & BIT)
10) proper argument and struct member alignment

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet: Add Open vSwitch kernel components.
Jesse Gross [Wed, 26 Oct 2011 02:26:31 +0000 (19:26 -0700)]
net: Add Open vSwitch kernel components.

Open vSwitch is a multilayer Ethernet switch targeted at virtualized
environments.  In addition to supporting a variety of features
expected in a traditional hardware switch, it enables fine-grained
programmatic extension and flow-based control of the network.
This control is useful in a wide variety of applications but is
particularly important in multi-server virtualization deployments,
which are often characterized by highly dynamic endpoints and the need
to maintain logical abstractions for multiple tenants.

The Open vSwitch datapath provides an in-kernel fast path for packet
forwarding.  It is complemented by a userspace daemon, ovs-vswitchd,
which is able to accept configuration from a variety of sources and
translate it into packet processing rules.

See http://openvswitch.org for more information and userspace
utilities.

Signed-off-by: Jesse Gross <jesse@nicira.com>
12 years agoipv6: Add fragment reporting to ipv6_skip_exthdr().
Jesse Gross [Thu, 1 Dec 2011 01:05:51 +0000 (17:05 -0800)]
ipv6: Add fragment reporting to ipv6_skip_exthdr().

While parsing through IPv6 extension headers, fragment headers are
skipped making them invisible to the caller.  This reports the
fragment offset of the last header in order to make it possible to
determine whether the packet is fragmented and, if so whether it is
a first or last fragment.

Signed-off-by: Jesse Gross <jesse@nicira.com>
12 years agovlan: Move vlan_set_encap_proto() to vlan header file
Pravin B Shelar [Fri, 18 Nov 2011 21:15:54 +0000 (13:15 -0800)]
vlan: Move vlan_set_encap_proto() to vlan header file

Open vSwitch needs this function for vlan handling.

Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
12 years agogenetlink: Add rcu_dereference_genl and genl_dereference.
Jesse Gross [Sun, 20 Nov 2011 00:21:37 +0000 (16:21 -0800)]
genetlink: Add rcu_dereference_genl and genl_dereference.

This adds rcu_dereference_genl and genl_dereference, which are genl
variants of the RTNL functions to enforce proper locking with lockdep
and sparse.

Signed-off-by: Jesse Gross <jesse@nicira.com>
12 years agogenetlink: Add lockdep_genl_is_held().
Pravin B Shelar [Fri, 11 Nov 2011 03:14:51 +0000 (19:14 -0800)]
genetlink: Add lockdep_genl_is_held().

Open vSwitch uses genl_mutex locking to protect datapath
data-structures like flow-table, flow-actions. Following patch adds
lockdep_genl_is_held() which is used for rcu annotation to prove
locking.

Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
12 years agogenetlink: Add genl_notify()
Pravin B Shelar [Fri, 11 Nov 2011 03:14:37 +0000 (19:14 -0800)]
genetlink: Add genl_notify()

Open vSwitch uses Generic Netlink interface for communication
between userspace and kernel module. genl_notify() is used
for sending notification back to userspace.

genl_notify() is analogous to rtnl_notify() but uses genl_sock
instead of rtnl.

Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
12 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless
John W. Linville [Fri, 2 Dec 2011 20:44:03 +0000 (15:44 -0500)]
Merge branch 'master' of git://git./linux/kernel/git/linville/wireless

Conflicts:
drivers/net/wireless/iwlwifi/iwl-agn.c
drivers/net/wireless/libertas/cfg.c

12 years agoiwlwifi: change the default behavior of watchdog timer
Wey-Yi Guy [Fri, 2 Dec 2011 16:19:19 +0000 (08:19 -0800)]
iwlwifi: change the default behavior of watchdog timer

The current default watchdog timer is enabled, but we are seeing issues on
legacy devices. So change the default setting of watchdog timer to per
device based. But user still can use the "wd_disable" module parameter
to overwrite the system setting

Cc: stable@vger.kernel.org #3.0+
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
12 years agoiwlwifi: do not re-configure HT40 after associated
Wey-Yi Guy [Fri, 2 Dec 2011 16:19:18 +0000 (08:19 -0800)]
iwlwifi: do not re-configure HT40 after associated

The ht40 setting should not change after association unless channel switch

This fix a problem we are seeing which cause uCode assert because driver
sending invalid information and make uCode confuse

Here is the firmware assert message:
kernel: iwlagn 0000:03:00.0: Microcode SW error detected.  Restarting 0x82000000.
kernel: iwlagn 0000:03:00.0: Loaded firmware version: 17.168.5.3 build 42301
kernel: iwlagn 0000:03:00.0: Start IWL Error Log Dump:
kernel: iwlagn 0000:03:00.0: Status: 0x000512E4, count: 6
kernel: iwlagn 0000:03:00.0: 0x00002078 | ADVANCED_SYSASSERT
kernel: iwlagn 0000:03:00.0: 0x00009514 | uPc
kernel: iwlagn 0000:03:00.0: 0x00009496 | branchlink1
kernel: iwlagn 0000:03:00.0: 0x00009496 | branchlink2
kernel: iwlagn 0000:03:00.0: 0x0000D1F2 | interruptlink1
kernel: iwlagn 0000:03:00.0: 0x00000000 | interruptlink2
kernel: iwlagn 0000:03:00.0: 0x01008035 | data1
kernel: iwlagn 0000:03:00.0: 0x0000C90F | data2
kernel: iwlagn 0000:03:00.0: 0x000005A7 | line
kernel: iwlagn 0000:03:00.0: 0x5080B520 | beacon time
kernel: iwlagn 0000:03:00.0: 0xCC515AE0 | tsf low
kernel: iwlagn 0000:03:00.0: 0x00000003 | tsf hi
kernel: iwlagn 0000:03:00.0: 0x00000000 | time gp1
kernel: iwlagn 0000:03:00.0: 0x29703BF0 | time gp2
kernel: iwlagn 0000:03:00.0: 0x00000000 | time gp3
kernel: iwlagn 0000:03:00.0: 0x000111A8 | uCode version
kernel: iwlagn 0000:03:00.0: 0x000000B0 | hw version
kernel: iwlagn 0000:03:00.0: 0x00480303 | board version
kernel: iwlagn 0000:03:00.0: 0x09E8004E | hcmd
kernel: iwlagn 0000:03:00.0: CSR values:
kernel: iwlagn 0000:03:00.0: (2nd byte of CSR_INT_COALESCING is CSR_INT_PERIODIC_REG)
kernel: iwlagn 0000:03:00.0:        CSR_HW_IF_CONFIG_REG: 0X00480303
kernel: iwlagn 0000:03:00.0:          CSR_INT_COALESCING: 0X0000ff40
kernel: iwlagn 0000:03:00.0:                     CSR_INT: 0X00000000
kernel: iwlagn 0000:03:00.0:                CSR_INT_MASK: 0X00000000
kernel: iwlagn 0000:03:00.0:           CSR_FH_INT_STATUS: 0X00000000
kernel: iwlagn 0000:03:00.0:                 CSR_GPIO_IN: 0X00000030
kernel: iwlagn 0000:03:00.0:                   CSR_RESET: 0X00000000
kernel: iwlagn 0000:03:00.0:                CSR_GP_CNTRL: 0X080403c5
kernel: iwlagn 0000:03:00.0:                  CSR_HW_REV: 0X000000b0
kernel: iwlagn 0000:03:00.0:              CSR_EEPROM_REG: 0X07d60ffd
kernel: iwlagn 0000:03:00.0:               CSR_EEPROM_GP: 0X90000001
kernel: iwlagn 0000:03:00.0:              CSR_OTP_GP_REG: 0X00030001
kernel: iwlagn 0000:03:00.0:                 CSR_GIO_REG: 0X00080044
kernel: iwlagn 0000:03:00.0:            CSR_GP_UCODE_REG: 0X000093bb
kernel: iwlagn 0000:03:00.0:           CSR_GP_DRIVER_REG: 0X00000000
kernel: iwlagn 0000:03:00.0:           CSR_UCODE_DRV_GP1: 0X00000000
kernel: iwlagn 0000:03:00.0:           CSR_UCODE_DRV_GP2: 0X00000000
kernel: iwlagn 0000:03:00.0:                 CSR_LED_REG: 0X00000078
kernel: iwlagn 0000:03:00.0:        CSR_DRAM_INT_TBL_REG: 0X88214dd2
kernel: iwlagn 0000:03:00.0:        CSR_GIO_CHICKEN_BITS: 0X27800200
kernel: iwlagn 0000:03:00.0:             CSR_ANA_PLL_CFG: 0X00000000
kernel: iwlagn 0000:03:00.0:           CSR_HW_REV_WA_REG: 0X0001001a
kernel: iwlagn 0000:03:00.0:        CSR_DBG_HPET_MEM_REG: 0Xffff0010
kernel: iwlagn 0000:03:00.0: FH register values:
kernel: iwlagn 0000:03:00.0:         FH_RSCSR_CHNL0_STTS_WPTR_REG: 0X21316d00
kernel: iwlagn 0000:03:00.0:        FH_RSCSR_CHNL0_RBDCB_BASE_REG: 0X021479c0
kernel: iwlagn 0000:03:00.0:                  FH_RSCSR_CHNL0_WPTR: 0X00000060
kernel: iwlagn 0000:03:00.0:         FH_MEM_RCSR_CHNL0_CONFIG_REG: 0X80819104
kernel: iwlagn 0000:03:00.0:          FH_MEM_RSSR_SHARED_CTRL_REG: 0X000000fc
kernel: iwlagn 0000:03:00.0:            FH_MEM_RSSR_RX_STATUS_REG: 0X07030000
kernel: iwlagn 0000:03:00.0:    FH_MEM_RSSR_RX_ENABLE_ERR_IRQ2DRV: 0X00000000
kernel: iwlagn 0000:03:00.0:                FH_TSSR_TX_STATUS_REG: 0X07ff0001
kernel: iwlagn 0000:03:00.0:                 FH_TSSR_TX_ERROR_REG: 0X00000000
kernel: iwlagn 0000:03:00.0: Start IWL Event Log Dump: display last 20 entries
kernel: ------------[ cut here ]------------
WARNING: at net/mac80211/util.c:1208 ieee80211_reconfig+0x1f1/0x407()
kernel: Hardware name: 4290W4H
kernel: Pid: 1896, comm: kworker/0:0 Not tainted 3.1.0 #2
kernel: Call Trace:
kernel:  [<ffffffff81036558>] ? warn_slowpath_common+0x73/0x87
kernel:  [<ffffffff813b8966>] ? ieee80211_reconfig+0x1f1/0x407
kernel:  [<ffffffff8139e8dc>] ? ieee80211_recalc_smps_work+0x32/0x32
kernel:  [<ffffffff8139e95a>] ? ieee80211_restart_work+0x7e/0x87
kernel:  [<ffffffff810472fa>] ? process_one_work+0x1c8/0x2e3
kernel:  [<ffffffff810480c9>] ? worker_thread+0x17a/0x23a
kernel:  [<ffffffff81047f4f>] ? manage_workers.clone.18+0x15b/0x15b
kernel:  [<ffffffff81047f4f>] ? manage_workers.clone.18+0x15b/0x15b
kernel:  [<ffffffff8104ba97>] ? kthread+0x7a/0x82
kernel:  [<ffffffff813d21b4>] ? kernel_thread_helper+0x4/0x10
kernel:  [<ffffffff8104ba1d>] ? kthread_flush_work_fn+0x11/0x11
kernel:  [<ffffffff813d21b0>] ? gs_change+0xb/0xb

Cc: <stable@kernel.org> 3.1+
Reported-by: Udo Steinberg <udo@hypervisor.org>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
12 years agoiwlagn: fix HW crypto for TX-only keys
Johannes Berg [Fri, 2 Dec 2011 16:19:17 +0000 (08:19 -0800)]
iwlagn: fix HW crypto for TX-only keys

Group keys in IBSS or AP mode are not programmed
into the device since we give the key to it with
every TX packet. However, we do need mac80211 to
create the MMIC & PN in all cases. Move the code
around to set the key flags all the time. We set
them even when the key is removed again but that
is obviously harmless.

Cc: stable@vger.kernel.org
Reported-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
12 years agoatm: clip: Remove code commented out since eternity.
David S. Miller [Fri, 2 Dec 2011 19:27:11 +0000 (14:27 -0500)]
atm: clip: Remove code commented out since eternity.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
David S. Miller [Fri, 2 Dec 2011 18:49:21 +0000 (13:49 -0500)]
Merge git://git./linux/kernel/git/davem/net

12 years agoforcedeath: Fix bql support for forcedeath
Igor Maravic [Thu, 1 Dec 2011 23:48:20 +0000 (23:48 +0000)]
forcedeath: Fix bql support for forcedeath

Moved netdev_completed_queue() out of while loop in function nv_tx_done_optimized().
Because this function was in while loop,
BUG_ON(count > dql->num_queued - dql->num_completed)
was hit in dql_completed().

Signed-off-by: Igor Maravic <igorm@etf.rs>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoniu: Fix typo in comment.
David S. Miller [Fri, 2 Dec 2011 17:37:33 +0000 (12:37 -0500)]
niu: Fix typo in comment.

Reported-by: Thomas Jarosch <thomas.jarosch@intra2net.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Linus Torvalds [Fri, 2 Dec 2011 04:09:08 +0000 (20:09 -0800)]
Merge git://git./linux/kernel/git/davem/net

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (73 commits)
  netfilter: Remove ADVANCED dependency from NF_CONNTRACK_NETBIOS_NS
  ipv4: flush route cache after change accept_local
  sch_red: fix red_change
  Revert "udp: remove redundant variable"
  bridge: master device stuck in no-carrier state forever when in user-stp mode
  ipv4: Perform peer validation on cached route lookup.
  net/core: fix rollback handler in register_netdevice_notifier
  sch_red: fix red_calc_qavg_from_idle_time
  bonding: only use primary address for ARP
  ipv4: fix lockdep splat in rt_cache_seq_show
  sch_teql: fix lockdep splat
  net: fec: Select the FEC driver by default for i.MX SoCs
  isdn: avoid copying too long drvid
  isdn: make sure strings are null terminated
  netlabel: Fix build problems when IPv6 is not enabled
  sctp: better integer overflow check in sctp_auth_create_key()
  sctp: integer overflow in sctp_auth_create_key()
  ipv6: Set mcast_hops to IPV6_DEFAULT_MCASTHOPS when -1 was given.
  net: Fix corruption in /proc/*/net/dev_mcast
  mac80211: fix race between the AGG SM and the Tx data path
  ...

12 years agonetfilter: Remove ADVANCED dependency from NF_CONNTRACK_NETBIOS_NS
David S. Miller [Fri, 2 Dec 2011 03:19:01 +0000 (22:19 -0500)]
netfilter: Remove ADVANCED dependency from NF_CONNTRACK_NETBIOS_NS

firewalld in Fedora 16 needs this.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoniu: Add support for byte queue limits.
David S. Miller [Fri, 2 Dec 2011 03:15:47 +0000 (22:15 -0500)]
niu: Add support for byte queue limits.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoniu: Remove redundant PHY ID test.
David S. Miller [Fri, 2 Dec 2011 02:59:07 +0000 (21:59 -0500)]
niu: Remove redundant PHY ID test.

Reported-by: Thomas Jarosch <thomas.jarosch@intra2net.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoipv4: flush route cache after change accept_local
Peter Pan(潘卫平) [Thu, 1 Dec 2011 15:47:06 +0000 (15:47 +0000)]
ipv4: flush route cache after change accept_local

After reset ipv4_devconf->data[IPV4_DEVCONF_ACCEPT_LOCAL] to 0,
we should flush route cache, or it will continue receive packets with local
source address, which should be dropped.

Signed-off-by: Weiping Pan <panweiping3@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agosch_red: fix red_change
Eric Dumazet [Thu, 1 Dec 2011 11:06:34 +0000 (11:06 +0000)]
sch_red: fix red_change

Le mercredi 30 novembre 2011 à 14:36 -0800, Stephen Hemminger a écrit :

> (Almost) nobody uses RED because they can't figure it out.
> According to Wikipedia, VJ says that:
>  "there are not one, but two bugs in classic RED."

RED is useful for high throughput routers, I doubt many linux machines
act as such devices.

I was considering adding Adaptative RED (Sally Floyd, Ramakrishna
Gummadi, Scott Shender), August 2001

In this version, maxp is dynamic (from 1% to 50%), and user only have to
setup min_th (target average queue size)
(max_th and wq (burst in linux RED) are automatically setup)

By the way it seems we have a small bug in red_change()

if (skb_queue_empty(&sch->q))
red_end_of_idle_period(&q->parms);

First, if queue is empty, we should call
red_start_of_idle_period(&q->parms);

Second, since we dont use anymore sch->q, but q->qdisc, the test is
meaningless.

Oh well...

[PATCH] sch_red: fix red_change()

Now RED is classful, we must check q->qdisc->q.qlen, and if queue is empty,
we start an idle period, not end it.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoLinux 3.2-rc4
Linus Torvalds [Thu, 1 Dec 2011 22:56:01 +0000 (14:56 -0800)]
Linux 3.2-rc4

12 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec...
Linus Torvalds [Thu, 1 Dec 2011 22:55:34 +0000 (14:55 -0800)]
Merge branch 'upstream-linus' of git://git./linux/kernel/git/jlbec/ocfs2

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/ocfs2: (31 commits)
  ocfs2: avoid unaligned access to dqc_bitmap
  ocfs2: Use filemap_write_and_wait() instead of write_inode_now()
  ocfs2: honor O_(D)SYNC flag in fallocate
  ocfs2: Add a missing journal credit in ocfs2_link_credits() -v2
  ocfs2: send correct UUID to cleancache initialization
  ocfs2: Commit transactions in error cases -v2
  ocfs2: make direntry invalid when deleting it
  fs/ocfs2/dlm/dlmlock.c: free kmem_cache_zalloc'd data using kmem_cache_free
  ocfs2: Avoid livelock in ocfs2_readpage()
  ocfs2: serialize unaligned aio
  ocfs2: Implement llseek()
  ocfs2: Fix ocfs2_page_mkwrite()
  ocfs2: Add comment about orphan scanning
  ocfs2: Clean up messages in the fs
  ocfs2/cluster: Cluster up now includes network connections too
  ocfs2/cluster: Add new function o2net_fill_node_map()
  ocfs2/cluster: Fix output in file elapsed_time_in_ms
  ocfs2/dlm: dlmlock_remote() needs to account for remastery
  ocfs2/dlm: Take inflight reference count for remotely mastered resources too
  ocfs2/dlm: Cleanup dlm_wait_for_node_death() and dlm_wait_for_node_recovery()
  ...

12 years agoocfs2: avoid unaligned access to dqc_bitmap
Akinobu Mita [Tue, 15 Nov 2011 22:56:34 +0000 (14:56 -0800)]
ocfs2: avoid unaligned access to dqc_bitmap

The dqc_bitmap field of struct ocfs2_local_disk_chunk is 32-bit aligned,
but not 64-bit aligned.  The dqc_bitmap is accessed by ocfs2_set_bit(),
ocfs2_clear_bit(), ocfs2_test_bit(), or ocfs2_find_next_zero_bit().  These
are wrapper macros for ext2_*_bit() which need to take an unsigned long
aligned address (though some architectures are able to handle unaligned
address correctly)

So some 64bit architectures may not be able to access the dqc_bitmap
correctly.

This avoids such unaligned access by using another wrapper functions for
ext2_*_bit().  The code is taken from fs/ext4/mballoc.c which also need to
handle unaligned bitmap access.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: Joel Becker <jlbec@evilplan.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Joel Becker <jlbec@evilplan.org>
12 years agoMerge branch 'fixes' of http://ftp.arm.linux.org.uk/pub/linux/arm/kernel/git-cur...
Linus Torvalds [Thu, 1 Dec 2011 19:53:54 +0000 (11:53 -0800)]
Merge branch 'fixes' of ftp.arm.linux.org.uk/pub/linux/arm/kernel/git-cur/linux-2.6-arm

* 'fixes' of http://ftp.arm.linux.org.uk/pub/linux/arm/kernel/git-cur/linux-2.6-arm:
  ARM: 7182/1: ARM cpu topology: fix warning
  ARM: 7181/1: Restrict kprobes probing SWP instructions to ARMv5 and below
  ARM: 7180/1: Change kprobes testcase with unpredictable STRD instruction
  ARM: 7177/1: GIC: avoid skipping non-existent PPIs in irq_start calculation
  ARM: 7176/1: cpu_pm: register GIC PM notifier only once
  ARM: 7175/1: add subname parameter to mfp_set_groupg callers
  ARM: 7174/1: Fix build error in kprobes test code on Thumb2 kernels
  ARM: 7172/1: dma: Drop GFP_COMP for DMA memory allocations
  ARM: 7171/1: unwind: add unwind directives to bitops assembly macros
  ARM: 7170/2: fix compilation breakage in entry-armv.S
  ARM: 7168/1: use cache type functions for arch_get_unmapped_area
  ARM: perf: check that we have a platform device when reserving PMU
  ARM: 7166/1: Use PMD_SHIFT instead of PGDIR_SHIFT in dma-consistent.c
  ARM: 7165/2: PL330: Fix typo in _prepare_ccr()
  ARM: 7163/2: PL330: Only register usable channels
  ARM: 7162/1: errata: tidy up Kconfig options for PL310 errata workarounds
  ARM: 7161/1: errata: no automatic store buffer drain
  ARM: perf: initialise used_mask for fake PMU during validation
  ARM: PMU: remove pmu_init declaration
  ARM: PMU: re-export release_pmu symbol to modules

12 years agodccp: Fix compile warning in probe code.
David S. Miller [Thu, 1 Dec 2011 19:45:49 +0000 (14:45 -0500)]
dccp: Fix compile warning in probe code.

Commit 1386be55e32a3c5d8ef4a2b243c530a7b664c02c ("dccp: fix
auto-loading of dccp(_probe)") fixed a bug but created a new
compiler warning:

net/dccp/probe.c: In function ‘dccpprobe_init’:
net/dccp/probe.c:166:2: warning: the omitted middle operand in ?: will always be ‘true’, suggest explicit middle operand [-Wparentheses]

try_then_request_module() is built for situations where the
"existence" test is some lookup function that returns a non-NULL
object on success, and with a reference count of some kind held.

Here we're looking for a success return of zero from the jprobe
registry.

Instead of fighting the way try_then_request_module() works, simply
open code what we want to happen in a local helper function.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet: Make ndo_neigh_destroy return void.
David S. Miller [Thu, 1 Dec 2011 19:16:04 +0000 (14:16 -0500)]
net: Make ndo_neigh_destroy return void.

The return value isn't used.

Suggested by Ben Hucthings.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoRevert "udp: remove redundant variable"
David S. Miller [Thu, 1 Dec 2011 19:12:55 +0000 (14:12 -0500)]
Revert "udp: remove redundant variable"

This reverts commit 81d54ec8479a2c695760da81f05b5a9fb2dbe40a.

If we take the "try_again" goto, due to a checksum error,
the 'len' has already been truncated.  So we won't compute
the same values as the original code did.

Reported-by: paul bilke <fsmail@conspiracy.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agobridge: master device stuck in no-carrier state forever when in user-stp mode
Vitalii Demianets [Fri, 25 Nov 2011 00:16:37 +0000 (00:16 +0000)]
bridge: master device stuck in no-carrier state forever when in user-stp mode

When in user-stp mode, bridge master do not follow state of its slaves, so
after the following sequence of events it can stuck forever in no-carrier
state:
1) turn stp off
2) put all slaves down - master device will follow their state and also go in
no-carrier state
3) turn stp on with bridge-stp script returning 0 (go to the user-stp mode)
Now bridge master won't follow slaves' state and will never reach running
state.

This patch solves the problem by making user-stp and kernel-stp behavior
similar regarding master following slaves' states.

Signed-off-by: Vitalii Demianets <vitas@nppfactor.kiev.ua>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoipv4: Perform peer validation on cached route lookup.
David S. Miller [Thu, 1 Dec 2011 18:38:59 +0000 (13:38 -0500)]
ipv4: Perform peer validation on cached route lookup.

Otherwise we won't notice the peer GENID change.

Reported-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoipv4: use a 64bit load/store in output path
Eric Dumazet [Wed, 30 Nov 2011 19:00:53 +0000 (19:00 +0000)]
ipv4: use a 64bit load/store in output path

gcc compiler is smart enough to use a single load/store if we
memcpy(dptr, sptr, 8) on x86_64, regardless of
CONFIG_CC_OPTIMIZE_FOR_SIZE

In IP header, daddr immediately follows saddr, this wont change in the
future. We only need to make sure our flowi4 (saddr,daddr) fields wont
break the rule.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agodccp: Evaluate ip_hdr() only once in dccp_v4_route_skb().
David S. Miller [Thu, 1 Dec 2011 18:28:34 +0000 (13:28 -0500)]
dccp: Evaluate ip_hdr() only once in dccp_v4_route_skb().

This also works around a bogus gcc warning generated by an
upcoming patch from Eric Dumazet that rearranges the layout
of struct flowi4.

Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agopowerpc: tqm8548/tqm8xx: add and update CAN device nodes
Wolfgang Grandegger [Wed, 30 Nov 2011 23:41:21 +0000 (23:41 +0000)]
powerpc: tqm8548/tqm8xx: add and update CAN device nodes

This patch enables or updates support for the CC770 and AN82527
CAN controller on the TQM8548 and TQM8xx boards.

CC: devicetree-discuss@lists.ozlabs.org
CC: linuxppc-dev@ozlabs.org
CC: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Wolfgang Grandegger <wg@grandegger.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocan: cc770: add platform bus driver for the CC770 and AN82527
Wolfgang Grandegger [Wed, 30 Nov 2011 23:41:20 +0000 (23:41 +0000)]
can: cc770: add platform bus driver for the CC770 and AN82527

This driver works with both, static platform data and device tree
bindings. It has been tested on a TQM855L board with two AN82527
CAN controllers on the local bus.

CC: Devicetree-discuss@lists.ozlabs.org
CC: linuxppc-dev@ozlabs.org
CC: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Wolfgang Grandegger <wg@grandegger.com>
Acked-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocan: cc770: add legacy ISA bus driver for the CC770 and AN82527
Wolfgang Grandegger [Wed, 30 Nov 2011 23:41:19 +0000 (23:41 +0000)]
can: cc770: add legacy ISA bus driver for the CC770 and AN82527

This patch adds support for legacy Bosch CC770 and Intel AN82527 CAN
controllers on the ISA or PC-104 bus. The I/O port or memory address
and the IRQ number must be specified via module parameters:

  insmod cc770_isa.ko port=0x310,0x380 irq=7,11

for ISA devices using I/O ports or:

  insmod cc770_isa.ko mem=0xd1000,0xd1000 irq=7,11

for memory mapped ISA devices.

Indirect access via address and data port is supported as well:

  insmod cc770_isa.ko port=0x310,0x380 indirect=1 irq=7,11

Furthermore, the following mode parameter can be defined:

  clk: External oscillator clock frequency (default=16000000 [16 MHz])
  cir: CPU interface register (default=0x40 [DSC])
  bcr: Bus configuration register (default=0x40 [CBY])
  cor: Clockout register (default=0x00)

Note: for clk, cir, bcr and cor, the first argument re-defines the
default for all other devices, e.g.:

  insmod cc770_isa.ko mem=0xd1000,0xd1000 irq=7,11 clk=24000000

is equivalent to

  insmod cc770_isa.ko mem=0xd1000,0xd1000 irq=7,11 clk=24000000,24000000

Signed-off-by: Wolfgang Grandegger <wg@grandegger.com>
Acked-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocan: cc770: add driver core for the Bosch CC770 and Intel AN82527
Wolfgang Grandegger [Wed, 30 Nov 2011 23:41:18 +0000 (23:41 +0000)]
can: cc770: add driver core for the Bosch CC770 and Intel AN82527

Signed-off-by: Wolfgang Grandegger <wg@grandegger.com>
Acked-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoicplus: mdio_write(), remove unnecessary for loop
Patrick Kelle [Thu, 1 Dec 2011 17:54:46 +0000 (12:54 -0500)]
icplus: mdio_write(), remove unnecessary for loop

At this point the variable j is always set to 7 and the code within
the loop has to run only once anyway.

As suggested by David Miller:
"You can simply this even further since p[7] is what is used here,
and this means len is one, the inner loop therefore executes only
once, and the p[7].field value is not used (it's zero in the table)
and the write to it is completely thrown away."

Signed-off-by: Patrick Kelle <patrick.kelle81@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonet: net_device flags is an unsigned int
Eric Dumazet [Wed, 30 Nov 2011 21:42:26 +0000 (21:42 +0000)]
net: net_device flags is an unsigned int

commit b00055aacdb ([NET] core: add RFC2863 operstate) changed
net_device flags from unsigned short to unsigned int.

Some core functions still assume its an unsigned short.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agodsa: Include linux/if_ether.h to fix build error
Axel Lin [Wed, 30 Nov 2011 22:07:18 +0000 (22:07 +0000)]
dsa: Include linux/if_ether.h to fix build error

Include linux/if_ether.h to fix below build errors:

  CC      arch/arm/mach-kirkwood/common.o
In file included from arch/arm/mach-kirkwood/common.c:19:
include/net/dsa.h: In function 'dsa_uses_dsa_tags':
include/net/dsa.h:192: error: 'ETH_P_DSA' undeclared (first use in this function)
include/net/dsa.h:192: error: (Each undeclared identifier is reported only once
include/net/dsa.h:192: error: for each function it appears in.)
include/net/dsa.h: In function 'dsa_uses_trailer_tags':
include/net/dsa.h:197: error: 'ETH_P_TRAILER' undeclared (first use in this function)
make[1]: *** [arch/arm/mach-kirkwood/common.o] Error 1
make: *** [arch/arm/mach-kirkwood] Error 2

Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agonetem: fix build error on 32bit arches
Eric Dumazet [Wed, 30 Nov 2011 23:32:14 +0000 (23:32 +0000)]
netem: fix build error on 32bit arches

ERROR: "__udivdi3" [net/sched/sch_netem.ko] undefined!

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Hagen Paul Pfeifer <hagen@jauu.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux...
Linus Torvalds [Thu, 1 Dec 2011 16:28:53 +0000 (08:28 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mason/linux-btrfs

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  Btrfs: fix meta data raid-repair merge problem
  Btrfs: skip allocation attempt from empty cluster
  Btrfs: skip block groups without enough space for a cluster
  Btrfs: start search for new cluster at the beginning
  Btrfs: reset cluster's max_size when creating bitmap
  Btrfs: initialize new bitmaps' list
  Btrfs: fix oops when calling statfs on readonly device
  Btrfs: Don't error on resizing FS to same size
  Btrfs: fix deadlock on metadata reservation when evicting a inode
  Fix URL of btrfs-progs git repository in docs
  btrfs scrub: handle -ENOMEM from init_ipath()

12 years agoRevert "mac80211: clear sta.drv_priv on reconfiguration"
John W. Linville [Thu, 1 Dec 2011 15:44:17 +0000 (10:44 -0500)]
Revert "mac80211: clear sta.drv_priv on reconfiguration"

This reverts commit f785d83a19bca326f79d127a413e35769afc0105.

This was provoking WARNINGs from the iwlegacy drivers.

Signed-off-by: John W. Linville <linville@tuxdriver.com>
12 years agoBtrfs: fix meta data raid-repair merge problem
Jan Schmidt [Thu, 1 Dec 2011 14:30:36 +0000 (09:30 -0500)]
Btrfs: fix meta data raid-repair merge problem

Commit 4a54c8c16 introduced raid-repair, killing the individual
readpage_io_failed_hook entries from inode.c and disk-io.c. Commit
4bb31e92 introduced new readahead code, adding a readpage_io_failed_hook to
disk-io.c.

The raid-repair commit had logic to disable raid-repair, if
readpage_io_failed_hook is set. Thus, the readahead commit effectively
disabled raid-repair for meta data.

This commit changes the logic to always attempt raid-repair when needed and
call the readpage_io_failed_hook in case raid-repair fails. This is much
more straight forward and should have been like that from the beginning.

Signed-off-by: Jan Schmidt <list.btrfs@jan-o-sch.net>
Reported-by: Stefan Behrens <sbehrens@giantdisaster.de>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agonet/core: fix rollback handler in register_netdevice_notifier
RongQing.Li [Thu, 1 Dec 2011 04:43:07 +0000 (23:43 -0500)]
net/core: fix rollback handler in register_netdevice_notifier

Within nested statements, the break statement terminates only the
do, for, switch, or while statement that immediately encloses it,
So replace the break with goto.

Signed-off-by: RongQing.Li <roy.qing.li@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocaif: Remove unused attributes from struct cflayer
sjur.brandeland@stericsson.com [Wed, 30 Nov 2011 13:02:32 +0000 (13:02 +0000)]
caif: Remove unused attributes from struct cflayer

Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocaif: Remove unused enum and parameter in cfserl
sjur.brandeland@stericsson.com [Wed, 30 Nov 2011 09:22:48 +0000 (09:22 +0000)]
caif: Remove unused enum and parameter in cfserl

Remove unused enum cfcnfg_phy_type and the parameter to cfserl_create.

Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
12 years agocaif: Restructure how link caif link layer enroll
sjur.brandeland@stericsson.com [Wed, 30 Nov 2011 09:22:47 +0000 (09:22 +0000)]
caif: Restructure how link caif link layer enroll

Enrolling CAIF link layers are refactored.

Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>